Recent Developments in Vision Research (1966)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

80 BIBLIOGRAPHY12 * Abelsdorff, G. Zur Erforschung des Helligkeits- u. Farbsinnes bei Menschen u. Thieren. (Verb. Physiol. Ges. Berlin 23. Feb.) Arch. Anat. Physiol., (Physiol.) 1900, 561-562. (a) * Abelsdorff, G. Ueber die Moglichkeit eines objektiven Nachweises der Farbenblindheit. Arch. Augenheilk., 1900, 41, 155-162. (b) * Abelsdorff, G. Die Anderungen der Pupillenweite durch verschieden- farbige Belichtung. Z. Psychol. Sinnesorg.. 1900, 22, 81-95. (c) * Abelsdorff, G., & Feilchenfeld, H. Uber die Abhangigkeit der Pupillar- reaction von Ort und Ausdehnung der gereizten Netzhautflache. Z. Psychol. Sinnesorg., 1904, 34, 111-131. (a) * Abelsdorff, G., & Feilchenfeld, H. Erwiderung auf die vorstehenden Bemerkungen von Dr. H. Wolff. Z. Psychol. Physiol. Sinnesorg., 1904, 36, 98. (b) " * Adamiik, E. Uber die Innervation der Augenbewegungen. Centralbl. f.d. med. Wiss., 1870, 65-67. * Adamuk, E., & Woinow, M. Uber die Pupillenveraderung bei der Accomodation. v. Graefe's Arch. Ophthal., 1871, 17,1, 158-168. / Alpern, M. Photopupillary kinetics in normal man. Fed. Proc., 1961 (March), 20. * Alpern, M., & Benson, D. J. Directional sensitivity of the pupilo- motor photoreceptors. Amer. J. Optom., 1953, 30, 569-580- Symbols: * - The original text was read in its entirety; # = abstracts or quotations by other authors (authors quoted are indicated); / = not read by reviewer July 1963. All citations, unless otherwise indicated, are completely referenced within this Bibliography under the name of the individual author(s). When this usual publication information was lacking, source was verified and is identifiable as stated in Index catalog list. All dissertations of foreign source listed herein (except those of American universities) are available from the National Library of Medicine, Bethesda, Md.

81 * Alpern, M., & Campbell, F. W. The behavior of the pupil during dark adaptation. Proc. Physiol. Soc.. 1962, 28-29, 1-2. (a) * Alpern, M., & Campbell, F. W. The spectral sensitivity of the con- sensual light reflex. J. Physiol., 1962, 164, 478-507. (b) * Alpern, M., Kitai, S.. & Isaacson, J. D. The dark adaptation process of the pupillomotor photoreceptors. Amer. J. Ophthal., 1959, 48, Pt. 2, 583-593. * Alpern, M., Mason, G. L., & Jardinico, R. E. Vergence and accom- modation. V. Pupil size changes associated with changes in accom- modative vergence. Amer. J. Ophthal., 1961, 52, 762-767. * Angelucci,—., & Aubert, — . Beobachtungen iiber die zur Akkomodation des Auges und die zur akkomodativen Kriimmungsveranderung der vorderen Linsenflache erfordenlichen Zeiten. Pfliiger's Arch. f.d. ges. Physiol., 1880, 22, 69-86. " '~~" * v. Arlt, Jr., F. Beitrag zur Kenntniss der Zeitverhaltnisse bei den Bewegungen der Iris. v. Graefe's Arch. Ophthal., 1869, 15, 294-317. * Aubert, H. Physiologische Optik. In Graefe-Saemisch Handb. ges. Augenheilk. (1st ed.), 1876, Vol. 2. Pp. 453, 457, 470. * Bach, L. Das Verhalten der Pupillen bei der Konvergenz und Akkomo- dation. Zeit. f. Augenheilk., 1904, 12, 725-729. * Bach, L. Pupillenlehre. Anatomie, Physiologie und Pathologie. Methodik der Untersuchung. Berlin: Karger, 1908. * Baker, F. H. Pupil response to short-duration pulses. MIT: Res. Lab. Electronics, Quart. Prog. Rep., 1962(April), No. 65, 251-257. * Bar any, E., &Hallden, U. Phasic inhibition of the pupil during retinal rivalry. J. Neurophysiol.. 1948, 11, 25-30. * Barbieri, A. El campo visual luminoso y el reflejo pupilar. La peri- metria objetiva. Su importancia clinica. Un nuovo perimetro pupiloscopico. Arch. Oftalm., Buenos Aires, 1929, 4, 618-638. * Barkan, O. Differential pupilloscopy. Arch. Ophthal., 1922, 51, 29-39. * Bartels, —. Zur Methode der Pupillenuntersuchung bei Gasglilhlicht. Z. Augenheilk, 1904, 11, 445-449.

82 * Bartley, S. H. A factor in visual fatigue . Psychosom. Med., 1942, 4, 369-375. """" * Bartley, S- H. Some parallels between pupillary "reflexes" and brightness discrimination. J. exper. Psychol., 1943, 32, 110-122. ^ * Easler, A. Uber die Pupillarreaktion bei verschiedenfarbiger Belichtung. Nach gemeinsam mit Frau S. Hofer (Leipzig) ausgefuhrten Untersuchungen. Pfluger's Arch, ges. Physiol., 1905, 108, 87-104. " * Becker, W. Statistiche Untersuchungen der Pupillenweiten in Abhangigkeit von der Lichtintensitat. Z. Biol., 1957, 109, 81-85. * Behr, C. Zur topischen Diagnose der Hemianopsie. v. Graefe's Arch. Ophthal., 1909, 70, 340-402. ~ " * Behr, C. Zur Physiologie und Pathologie des Lichtreflexes der Pupille. v. Graefe's Arch. Ophthal., 1913, 86, 468-513. * Behr, C. Die Lehre von den Pupillenbewegungen. In Graefe-Saemisch Handbuch der ges. Augenheilk. Vol. II, 3rd ed . Berlin: Springer, " "" * Bender, M. B., & Weinstein, E. A. Functional representation in the oculomotor and trochlear nuclei. Arch. Neurol. Psychiat., 1943, 49, 98-106. * Biffis, A. Ricerche topografiche sulla sensibilita retinica, par il riflesso fotomotore della pup illa. Ann. Ottal., 1934, 62, 665-694. * Blanchard, J. The brightness sensibility of the retina. Phys . Rev . , 1918, Ser. 2, 11, 81-99. ~ * Bleichert, A. Die Lichtkontraktion der Pupille als Regelvorgang. v. Graefe's Arch. Ophthal., 1957, 159, 396-410. * Bleichert, A., & Wagner, R. Uber den Frequenzgang der Pupillen- reaktion auf Licht. Z. Biol., 1957, 109, 281-296. * Bomer, M. Uber die praktische Verwendbarkeit des Sanderschen Pupilloskopes. Klin. Mbl. Augenheilk., 1933, 90 (N.F. 55), 207-210. * Bordier, H. De l'acuite" visuelle . Mem. Soc . phys. nat. Bordeaux, 1892-1893, (Seance 15 juin 1893) 4s 4, 1-156. ~

83 * Borsotti, I. Indagini pupillografiche su alcuni particolari aspetti fisiologici del riflesso fotomotore mediante cinematografia a luce ordinaria. (Nota prev.) Boll. Soc. med.-chir., Pavia, 1939, 53, 943-961. * Borthen, L. Die topisch-diagnostichen Verhaltnisse bei einseitiger isolirter reflectorischer Pupillenstarre. Klin. Mbl. Augenheilk., 1892,30,121-134. * Bouma, H. Size of the static pupil as a function of wavelength and luminosity of the light incident on the human eye. Nature, 1962, 193, 690-691. * Braun, G. Ein neues Hemikinesimeter. Klin. Mbl. Augenheilk., 1931, 87, 441-450. * Braun, G. Ein neues Hemikinesimeter. Klin. Mbl. Augenheilk., 1932, 88, 61-62. ^ "" " * Broca, A., & Laporte, F. Etude des principales sources de lumiere au point de vue de 1 "Hygiene de 1'oeil. (Extr. Bull. Soc. int. electriciens, juin) Ann, d'oculist, 1908, 151, 190-218. f * Broca, A., Jouaust, de La Gorce, & Laporte, F. Etude des nouvelles lampes electriques luminescentes. Action sur 1'oeil du rouge extreme et de l'ultraviolet. (Extr. Bull. Soc. int. electriciens, fev.) Ann, d'oculist, 1914, 151, 273-285. * Brown, R. H., & Page, H. E. Pupil dilation and dark adaptation. J. exper. Psychol.. 1939, 25, 347-360. * Budge, J. Ueber die Bewegungen der Iris, fur Physiologen und Arzte. Braunschweig: F. Viewegu. Sohn, 1855. * Bujadoux, A. & Gourevitch, F. Les troubles du pre-Argyll-Robertson (De la reflexome'trie pupillaire). Presse med ., 1937, 45,550. * Bujadoux, A., & Kofman, T. Un nouveau reflexometre pupillaire. Bull. Soc. franc,. d'Ophthal., Paris, 1933, 46, 243-249. (46Congr., 26-29 juin, 1933) * Bumke, O- Beitrage zur Kenntnis der Irisbewegungen. I. Der gal- vanische Lichtreflex(Vorlaufige Mitteilung). Centralbl. f. Nerven- heilk. & Psychiat., 1903, 62, 447-451. ' * Bumke, O. Die Pupillenstorungen bei Geistes- und Nervenkrankheiten. Jena: Fischer, 1904.

84 / Burian, H., & Schubert, G. Das Wesen der Naheinstellungsreaktion der Pupillen. v. Graefe's Arch. Ophthal., 1936, 136, 377-386. * Burke, D. W. A comparison of pupillary and visual thresholds. Dissertation, Univer. Minn., 1963. * Campbell, F. W., & Alpern, M. Pupillomotor spectral sensitivity curve and color of the fundus. J. optical Soc. Amer., 1962, 52, 1084. " "™ * Campbell, F. W., & White side, T. C. D. Induced pupillary oscilla- tions. Brit. J. Ophthal., 1950, 34, 180-189. * Caspary, H., & Goeritz, C. Die Synergie von Akkomodation und Pupillenreaktion. Pfliiger's Arch. f.d. ges. Physiol., 1922, 193, 225-230. "~ ~~~~ * Chaveau, A. Sur le mechanisme des mouvements de l'iris. J. de 1'Anat. et de la Physiol., 1888, 24, 193-200. Also, C.R. Soc. Biol., 1888, 14, 352. * Clynes, M. Computer dynamic analysis of the pupil light reflex: a unidirectional rate sensitive sensor. London: 3rd Int. Conf. med. Electronics Rep., 1960, 356-358. * Clynes, M. Unidirectional rate sensitivity: a biocybernetic law of reflex and humoral systems as physiologic channels of control and communication. (Pavlovian Conf. Monogr. on Higher Nervous Activity) Ann. N.Y. Acad. Sci., 1961, 92:946-969. * Coccius, E. A. De Pupillenreactie by Accomodatie en Convergentie. Dissertation, Univer. Leiden, 1899. Cited by H. Vervoort, 1900; A. Fuchs, 1903; K. Weiler, 1910. * Cohn, H. Ueber das Photographieren des Auges. Centralbl. f. prakt. Augenheilk, 1888, 12, 65-67. ~ * Corrado, M. Influenza dell1 estensione dello stimulo luminoso sul • potere riflessogeno pupillomotore della retina (curva riflessometrica pupillomotoria). Ann. Ottal. din. Ocul., 1947, 73, 171-182. * Crawford, B. H. Dependence of pupil size uppn external light stimulus under static and variable conditions. Proc. Roy. Soc., Ser. B., 1936-1937, 121, 376-395. ~~~" * Cramer, A. Het accomodatie-vermogen der Ogen. Harlem: de erven Loosjes, 1853. Cited by H. Vervoort, 1900; F. C. Donders, 1888; G. Braun, 1931.

85 * Cuppers, C. Die fortlaufende Registrierung der direkten und der konsensuellen Pupillenreaktion. v. Graefe's Arch. Ophthal., 1954, 155, 588-616. ~ ~~~ * Cuppers, C., & Graff, Th. Uber ein neues Gerat zur stetigen Beobachtung und Aufzeichnung des normalen und pathologischen Pupillenreflexes (Ein Beitrag zur Frage der objektiven Perimetrie). Klin. Mbl. Augenheilk.. 1951, 110, 189-194. * Cuppers, C., & Wagner, E. Zur pharmacologischen Beeinflussung der Netzhautfunktion. Klin. Mbl. Augenheilk., 1951, 118, 208-230. * De Groot, S.G., &Gebhard, J. W. Pupil size as determined by adapt- ing luminance. J. opt. Soc. Amer., 1952, 42, 492-495. * De Launay, J. A note on the photo-pupil reflex. J. opt. Soc. Amer., 1949, 39, 364-367. * De Ruiter, G. C. P. Onderzoekingen over de werking van Atropa Belladonna op de iris (Trans., Diss. Phys. med. de actione Atropae | Belladonna in iridem. Dissertation, Univer. Utrecht, 1853). Nederl. Lancet. 1853-1854, 3, 433-472. Cited by F. C. Donders, 1888; G. Braun, 1931. * Dolenek, A., Kristek, A., Nemek, J., & Komenda, S- Uber Veranderungen der Pupillenreaktion nach erfolgreicher Amblyopie- behandlung. Klin. Mbl. Augenheilk. 1962, 141, 353-357. * Donders, F. C. On the anomalies of accommodation and refraction of the eye. The new Sydenham Soc., London, 1864, Vol. 22, 573- 583. * Donders, F. C. Reflexiebeweging der beide Pupillen bij het invallen van Licht aan eene Zijde. Nederl. Arch. voor Geneesk., 1865, 106. Cited by H. Vervoort, 1900; F. von Arlt. Jr., 1869; K. Weiler, 1910. * Donders, F. C. Die Anomalien der Refraktion und Akkommodation des Auges. (2nd ed.) Wien: W. Braumiiller, 1888. ~ * Drischel, H. Uber die Dynamik des Lichtreflexes der menschlichen Pupille . I. Mitt. Der normale Reflexablauf nach kurzdauernder Belichtung und seine Variabilitat. Pfluger's Arch. ges. Physiol., 1957, 264, 145-168. (a) ~ ' * Drischel, H. Untersuchungen iiber die Dynamik des Lichtreflexes der menschlichen Pupille. II. Mitt. Dynamische Typologie des normalen Lichtreflexablaufes unter dem Einfluss vegetativer Pharmaca. Pfliiger's Arch. ges. Physiol.. 1957, 264, 269-290. (b)

86 * Drischel, H. Experimentelle Untersuchungen iiber den Lichtreflex der menschlichen Pupille nach kurzdauerndem Lichtreiz . Klin. Mbl. Augenheilk., 1957, 131, 740-755. (c). * Drouin, A. Note pour demontrer qu'il n'y a pas de rapport direct entre l'etat d'accommodation de 1'oeil et le diametre de la pupille. C. R. Soc. Biol.. 1876, T3, 6th ser., 199-205. * Dubois-Poulsen, F., & Loisillier, M. Un appareil utilisant les convertisseurs d'infrarouges pour 1'etude de la pupille dans 1'obscurite absolue. (Soc. d'Ophtal. Paris 18. dec.) Bull. Soc. Ophtal., France, 1954-1955, 150-154. Also, Zbl. Ophthal., 1955-56, 66, 137-138. * Du Bois-Reymond, Cl., & Greeff, R. Uber Pupillenstudien. (Berl. Ophthal. Ges. 15, Juni, 1893) Centralbl. Augenheilk., 1894, 18, 171-173. * Dupont, M. Instruments pour provoquer et mesurer le reflexe pupil- laire. (Soc. Neurol. 15 Mai) Rev, gen. Ophtal., 1902, 21, 515-516. (a) * Dupont, M. Excitateur de la pupille pour la recherche du reflexe lumineux. C. R. Soc. Biol., 1902, 54, 1366-1368. (b) * Eckhard, C. Beitrage zur Geschichte der Experimentalphysiologie der .motorischen Nerven des Auges. Vol. 11, Giessen: Emil Roth, 1885. Pp. 115-218. * Elsberg, C. A., & Spotnitz, H. Relation between area and intensity of light and size of pupil, with formulas for pupillary reactions. Bull. neurol. Inst.. N.Y., 1938, 7, 160-169. * Engel, S- Widerstandspupilloskop. Ein neuer Apparat zur Messung der motorischen und optischen Unterschiedsempfindlichkeit. Arch. Augenheilk., 1930, 103, 657-664. ~ * Engelking, E. Der Schwellenwert der Pupillenreaktion und seine Beziehungen zum Problem der pupillomotorischen Aufnahmeorgane . Z. Sinnephysiol., 1919, 50, 319-337. * Engelking, E. Vergleichende Untersuchungen iiber die Pupillenreaktion bei der angeborenen totalen Farbenblindheit. Klin. Mbl. Augen- heilk., 1922, 69, 177-188. * Feldmann, J. B. Improved illuminator and pupillometer. Arch. Ophthal., 1933, 9, 974-976.

87 * Ferree, C. E., & Rand, G. Instrument for measuring breadth of the pupil. Amer. J. Psychol., 1937, 38, 292-293. * Ferree, C- E., & Rand, G. Relation of size of pupil to intensity of light and speed of vision and other studies. J. exper. Psychol., 1932, 15, 37-55. * Ferree, C. E., Rand, G., & Harris, E. T. Intensity of light and area of illuminated field as interacting factors in size of pupil. J. exper. Psychol., 1933, 16, 408-422. * Fick, A. Uber die chromatische Abweichung des menschlichen Auges. v. Graefe's Arch. Ophthal., 1856, 2,2, 70-76. * Flamant, Francoise. Variation du diametre de la pupille de 1'oeil en fonction de labrillance. Rev, d'opt. theoret. et instr., 1948, 27, 751-758. * Fragstein, — ., & Kempner, — . Pupillenreaktionspriifer . Klin. Mbl. Augenheilk., 1899, 37, 243. * Friberger, R. Om matning af pupillens vidd . Dissertation, Univer . Uppsala, 1903. Cited in Jber. Leist. Ophthal., 1903, 34, 68-70. * Friedlander, R., & Kempner, —. Beitrag zur Kenntnis der hemianopi- schen Pupillenstarre. Neurol. Zentralbl., 1904, 23, 2-8. * Fry, G. A. The relation of pupil size to accommodation and con- vergence. Amer. J. Optom., 1945, 22, 451-465. * Fry, G. A., & Allen, M. J. Effect of flashes of light on night visual acuity. Part II. Ann Arbor, Mich.: AF-NRC Vision Comm. Rep., 1953. * Frydrychovicz, G., & Harms, H. Das pupillomotorische Perimeter. (Ber. 53. deutsch. Ophthal. Ges. Dresden, August 5-7) Zbl. Ophthal.. 1940, 45, 326; 650. (a) * Frydrychovicz, G., & Harms, H. Objektive Perimetrie. (Ber. 53. Verh. deutsch. Ophthal. Ges. Dresden, August 5-7) Zbl. Ophthal.. 1940, 45, 614-615. (b) * Frydrychovicz, G., & Harms, H. Ergebnisse pupillomotorischer Untersuchungen an Gesunden und Kranken. (Ber. 53. Verh. deutsch. Ophthal. Ges. Dresden, August 5-7) Zbl. Ophthal.. 1940, 45, 615. (c) * Fuchs, A. Die Messung der Pupillengrosse und Zeitbestimmung der Lichtreaktion der Pupillen bei einzelnen Psychosen und Nerven- krankheiten. Jahrbiicher f. Psychiat.. 1903, 24, 326-458.

88 * Fugate, J. M. A masking technique for isolating the pupillary response to focused light. J. opt. Soc . Amer., 1954, 44, 771-779. * Fugate, J. M., & Fry, G. A. Relation of changes in pupil size to visual discom/ort. 111. Eng.. 1956, 51, 537-549. * Galenus (Galen), C. (131-200, approx.). Opus de usu partium corporis humani etc. Transl. by Daremberg, Ch.: Oeuvres anatomiques, physiplogiques et medicales de Galien, traduites sur les textes im- primes et manuscripts, accompagnees de sommnaires, de notes, de planches et d'une table de matieres, precedees d'une introduction ou etude biographique, literaire et scientifique sur Galien. Paris: J.-B. Bailliere, 1854, 1856, 2 vols. * Garten, S- Beitra'ge zur Kenntnis des zeitlichen Ablaufes der Pupillen- reaktion nach Verdunkelung. Pflilger's Arch. ges. Physiol., 1897, 68,68-94. * Gasteiger, H. Klinische Beobachtungen iiber die Ausdehnung des pupillomotorisch wirksamen Bezirkes der Netzhaut. Arch. Augen- heilk.. 1934, 108, 553-558. "~= * Gifford, R., & Mayer, L. L. Clinical use of Sander pupilloscope . Arch. Ophthal., 1931, 6, 63-69. * Gorham, J. The pupil-photometer. Proc. Roy. Soc., London, 1884, 37, 425-426. ~' * Gradle, H. S., & Ackerman, W. Reaction time of normal pupil; second communication. AMA Transactions, Sect. Ophthal., 1932, 342-348. ~ * Gradle, H. S., & Eisendraht, E. B. Die Reaktionszeit der normalen Pupille. Vorlaufige Mitt. Mbl. Augenheilk.. 1923, 71, 311. * von Graefe, A. Das Verhalten der Pupillen am Hund bei der Accomoda- tion in der Na*he (Observation by A. Miiller) v. Graefe's Arch. Ophthal., 1854, 1,1, 440. ~~ ~~~ * von Graefe, A. Pathologisches zur Accommodationslehre. v. Graefe's Arch. Ophthal., 1856, 2,2, 299-319. ~ * Groethuysen, G. Uber die Beziehungen zwischen motorischer und optischer Unterschiedsempfindlichkeit bei normalen und krankhaften Zustanden des Sehorgans. Arch. Augenheilk., 1921, 87, 152-188; 88; 83-115- ~ * Gualdi, V. Contributo sperimentale allo studio della reazione pupillare alla convergenza (Nota preventiva). Atti Congr. Soc. Ital. diOftal., Roma, 1930. Cited by V. Gualdi, 1931. (a) ~~ ~~~

89 * Gualdi, V. La reazione della pupilla alla convergenza (Revisione critica degli studi sull'argomento e stato attuale della conoscenza scientifica del fenomeno: reazione pupillare alla accomodazione- convergenza). Lettura oftal., 1930, 7, 205-226. * Gualdi, V. Richerche sperimentali e osservazioni cliniche sul corn- portamento e sul meccanismo della reazione pupillare alla con- vergenza. Boll, d'oculistica, Florence, 1931, 10, 806-855. (b) / Gundlach, R. H. The speed of pupillary contraction in response to light in pigeons, cats, and humans. J. genet. Psychol., 1934, 44, 250-253. '~ '" ~ * Haessler, H. F. Near reaction of the pupil in the dark. Arch. Ophthal.. 1937, 18, 796-801. * Harms, H. Grundlagen, Methodtk und Bedeutung der Pupillenperimetrie fiir die Physiologie und Pathologie des Sehorgans. v. Graefe's Arch. Ophthal., 1949, 149, 1-68. * Harms, H. Hemianopische Pupillenstarre. Klin. Mbl. Augenheilk.. 1951, 118, 133-147. (a) -—~ * Harms, H. Objektive Kontrolle von Gesichtsfeldstb'rungen. 57. Zus. dtsch. Ophthal. Ges., Heidelberg, 1951, 245-251. (b) * Harms, H. Neue Methoden der Perimetrie. In: Zeitfragen der Augen- heilk. Leipzig: Georg Thieme, 1954. ' "~ * Hartinger, H. Pupillenweiten. Z. Ophthal. Opt.. 1937, 25, 1-7. * Hecht, S., & Pirenne, M. H. Sensibility of the nocturnal long-eared owl in the spectrum. J. gen. Physiol., 1939-1940, 23, 709-717. * Heddaeus. E. Uber hemiopische Pupillenreaction. Dtsch. med. Wschr.. 1893, 19, 748-751. * Heine, L. Physiologisch-anatomische Untersuchungen uber die Accomodation des Vogelauges. v. Graefe's Arch. Ophthal., 1898, 45,469-496. * Helmbold, —. Zur Priifung der Pupillarreaktion. Med. Klinik, 1911, No. 47. Cited in Neurol. Zbl., 1912, 428 (Tobias). * Hensen, V., & VSlckers, C. Uber den Ursprung der Accomodations- nerven, nebst Bemerkungen uber die Function der Wurzeln des Nervus oculomotorius. v. Graefe's Arch. Ophthal., 1878, 24, 1-26.

90 i Hering, E. Die Lehre vom binocularen Sehen. Leipzig: W. Engelmann, n.d. Cited by V. Gualdi, 1931; E. Wlotzka, 1905; W. Lohmann, 1908. Also in C. A. Raddle: Spatial sense and movements of the .eye .* Trans., n.d., no title of original text. * Hess, C. Untersuchungen iiber die Ausdehnung des pupillomotorisch wirksamen Bezirkes der Netzhaut und uber die pupillomotorischen Aufnahmeorgane. Arch. Augenheilk., 1907, 58, 182-205. * Hess, C. Untersuchungen iiber das Sehen und uber die Pupillenreaction von Tag- und von Nachtvogeln. Arch. Augenheilk., 1908, 59, 143-167. (a) * Hess, C. Zur Physiologie und Pathologie des Pupillenspiels. Arch. Augenheilk., 1908, 60, 327-389. (b) * Hess, C. Neue Untersuchungen iiber den Lichtsinn bei wirbellosen Tieren. Pfliiger's Arch. ges. Physiol., 1910, 136, 282-367. * Hess, C. Vorfiihrung eines Apparates fur Pupillenuntersuchungen. (34. Vers. Rhein.-Westfal. Augenarzte Kb'ln) Zbl. ges. Ophthal., 1914-1915, 1-2, 434. (a) * Hess, C. Neue Versuche iiber Lichtreaktionen bei Tieren und Pflanzen. Miinch. med. Woch., 1914, 61, No. 27, 1489-1492. (b) * Hess, C. Messende Untersuchungen zur vergleichenden Physiologie des Pupillenspiels. v. Graefe's Arch. Ophthal., 1915, 90, 382- 393. * Hess, C. Das Differentialpupilloskop. Eine Methode zur messenden Bestimmung von Storungen des Pupillenspiels. Arch. Augenheilk., 1916, 80, 213-228. * Hess, C., & Heine, L. Arbeiten aus dem Gebiet der Akkomodations- lehre. IV. Experimentelle Untersuchungen iiber den Einfluss der Akkomodation auf den intraoculaten Druck, nebst Beitragen zur Kenntnis der Akkomodation bei Saugetieren. v. Graefe's Arch.. Ophthal., 1898, 46, 2, 243-276. * Hesse, R. Studien iiber die hemiopische Pupillenreaktion und die Ausdehnung des pupillomotorischen Bezirkes der Netzhaut. Klin. Mbl. Augenheilk.. 1909, 47(N.F.7), 33-55. * Hesse, R. Uber die Verengerung der Pupille beim Nahesehen. Klin. Mbl. Augenheilk., 1912, 50 (N.F. 13), 740-745. * Himly, K. Ophthalmologische Beobachtungen und Untersuchungen: 1801; 47. Cited by J. Budge, 1855; C. Eckhard, 1885.

91 * Hirschberg, J. Geschichte der Augenheilkunde. In Graefe-Saemisch Handbuch der Augenheilk. Berlin: 1908-1918. 15vols. * Holladay, L. L. The fundamentals of glare and visibility. J. opt. Soc. Amer.. 1926, 12, 271-319. * Hopkinson, R. G. Glare discomfort and pupil diameter. J . opt. Soc. Amer., 1956, 46, 649-656. * Hu'bner, A. H. Beschreibung eines Apparates zur Untersuchung der Pupillen nebst Bemerkungen iiber einige Pupillenreaktionen. Monatsschr. Psychiat., 1907, 22, 15-20. * Isakowitz, J. Einige Bemerkungen zu der Arbeit von Dr. Robert Hesse "Uber die Verengerung der Pupille beim Nahesehen. " Klin. Mschr. Augenheilk., 1912, 50,2, 228-230. * Jampel, R. S- A study of convergence, divergence, pupillary reac- tions and accommodation from faradic stimulation of the macaque brain. J. comp. Neurol., 1960, 115, 371-397. * Jampel, R. S- Representation of the near-response on the cerebral cortex of the macaque. Amer. J. Ophthal., 1959, 48, Pt. 2, 573-582. * Kadlecova, V. The pupillomotorial adaptation of the eye. Ophtbalmo- logica, (Basel) 1960, 140, 379-387. (a) / Kadlecova, V. (Objective adaptometry by control of the pupillary reaction.) In Czech. Cs. Oftalmol.) 1958, 14, 85-91. Cited in Ophthal. Lit., 1958. / Kadlecova, V. (The influence of pre-adaptation on pupillomotor adapta- tation.) In Czech. Cs. Oftalmol., 1960, 16, 241-246. Cited in Ophthal. Lit., 1960. (b) / Kadlecova, V. (Second bend in the course of the sensory and pupil- lomotor adaptation curve.) In Czech. Sbor Jek.. 1960, 62, 189- 192. Cited in, Ophthal. Lit., 1960. (c) * Kadlecova, V. (The effect of dazzling on sensory and pupillomotor adaptation of the eye.) In Czech. Cs. Qftal., 1961, 17, 410-412. Summary by author in English. * Kadlecova, V., & Peleska, M. (Infrascopic examination of the diameter of the pupil jii darkness_adaptation; infrascopic pupil- lometry.) In Czech. 6s. Oftal. ,"~1955, 11, 4-5.

92 * Kadlecova, V., & Peleska, M. (The pupillary diameter in light and dark in various age categories.) In Czech. Cs. Oftal., 1957, 13, 278-282. (See author's summary in English.) Cited in Zbl. ges. Ophthal., 1958, 73, 255. * Kanngiesser, F. Licht und Konvergenzmiosis, zwischen der Atropin- und Eserinwirkung auf Iris und Ciliarmuskel, nebst Bemerkungen u'ber die Form der Pupille . Dissertation, Univer. Marburg, 1908. Wiesbaden: Bergmann, 1908. Also, Arch. Augenheilk., 1909, 63, 78-88. "~~ * Kappauf, W. E. The size of the cat's pupil as a function of stimulus brightness. Psychol. Bull., 1938, 35, 719. * Kappauf, W. E. Variation in the size of the cat's pupil as a function of stimulus brightness. J. comp. Psychol., 1943, 36, 125-230. * Kawabata, H. Electroretinogram and pupillogram. Comparison of the two methods, based on simultaneous recording of retinal potentials and pupillary movements. Paper presented at Assoc. res. Ophthal. Eastern Section. Amer. J. Ophthal., 1961. (Abstract) / Keeler, C. E. Le reflexe irien a la lumiere chez la souris a retine sans batonnets. C. R. Soc. Biol., 1927, 96, 1, 10. (Ref.: Jber. Ophthal., 1927) * Kempner, —. Neues Instrument f.d. Priifungd. hemianopischen Pupillenreaktion. (Ber. Verh. 9. Int. Ophthal. Congr. Utrecht.) Beilageheft, Zeitschr. f. Augenheilk., 1899, 2,23. * Kestenbaum, A., & Eidelberg, L. Konvergenzreaktion der Pupille und Naheeinstellung. v. Graefe's Arch. Ophthal., 1928-1929, 121, 166-212. """ * Kleefeld, G. Pupillometrie physiologique et pathologique. Une nouvelle methode de mesuration du reflexe photomoteur de la pupille. Ann. d'oculist, 1921, 158, 4-43; 262-304. * Knoll, H. A. Pupillary changes associated with accommodation and convergence. Amer. J. Optom., 1949, 26, 346-356. * Koch, R. Ein einfaches Pupilloskop. Munch. med. Wschr. (Feldarzt. Beilage II), 1916, 1607-1608. * Kofman, T., & Bujadoux, A. Le reflexometre pupillaire (presentation de 1'appareil). Les resultats de la reflexometrie dans l'etude du reflexe photomoteur normal. C. R. Soc. Biol., Paris, 1922, 186, 1165-1166; 1166-1167.

93 * Kofman, T., & Bujadoux, A. Recherches sur la reflexometrie pupil - laire normale et pathologique. Ann, d'oculist, 1923, 160, 943- 961. "~ * Kreuchel, V. Over de werking'van muscarine op de accommodatie en de pupil. Ned. Gasthuis voor Ooglijdes, Utrecht, 1874, 36-52. Cf. also, Uber_die Wirkung des Muscarins auf Akkomodation und Pupille., v. Graefe's Arch. Ophthal., 1874, 20, 135-150. * Krusius, F. F. Uber ein Unocularpupillometer. Arch. Augenheilk., 1907, 57, 97-100. * Krusius, F. F. Uber ein Biocular-Pupillometer. Neurol. Zentralbl., 1908, 154-157. " * Lambert, J. H. Photometria sive de mesura et gradibus luminus, colorum et umbrae, 1760J In E . Audig (Ed. & Trans.) Photo- metria. Leipzig: W. Engelmann, 1892. Ch. 2. Experimentelle und theoretische Bestimmung der Beziehungen zwischen der Offnung der Pupille, der subjektiven Helligkeit und der Grosse der Lichtquellen. * Landolt, M. Suggestions on methodical examination of the pupils. Amer. J. Ophthal., 1920, 3, 81-87. * Lans, L. J. Diametre pupillaire. (Xe Congr. int. d'ophtal.) Rev. gen, d'Ophthal.. 1899, 18, 475-476. '" * Lans, L. J. Ueber Pupillenweite. Arch. Anat. Physiol. (Physiol.), 1900, 79-102. ~ * Laurens, H. The relative physiological value of spectral lights. III. The pupillomotor effects of wave-lengths of equal energy content. Amer. J. Physiol., 1923, 64, 97-119. * Le Conte, J . On some phenomena of binocular vision. Amer . J . Sci. & Art, 1869, 47, 68-77. Cited by G. A. Fry, 1945; E. Marg & M. W. Morgan, Jr., 1949. * Lehrfeld, L. Quantitative pupillary light reflex. Amer. J. Ophthal.. 1928, 11, 897-898. / de Leonibus, F. Richerche di pupillometria nella visione del vincino. Ann. Ottalm.. 1949, 75, 231-238. * Levatin, P. Pupillary escape in disease of the retina or optic nerve. Arch. Ophthal., 1959, 62, 768-779.

94 * Listing, J. B. Beitrag zur Physiologischen Optik. Besonderer Abdruck aus d. Gottinger Studien, 1845, 27,28. Cited by C. Eckhard, 1885; A. Drouin, 1876; H. Aubert, 1876. * Lodato, G. Pupilloscopy and adaptometry. Arch. di Ottal., 1934, 41, 306. Also, Amer. J. Ophthal.. 1935, 18, 572. (Abstract, H. D. Scarney) * Loewenfeld, Irene E. Uber die Aufzeichnung des Pupillenspiels und die nervose Steuerung dar Pupillenreflexe . v. Graefe's Arch. Ophthaq., 1956, 157, 628-655. ~ * Loewenfeld, Irene E. Mechanisms of reflex dilation of the pupil. Historical review and experimental analysis. Doc. Ophthal., 1958, 12, 185-448. ~ * Loewenfeld, Irene E. The iris as pharmacologic indicator. I. Effect of physostigmine and of pilocarpine on pupillary movements in normal man. Arch. Ophthal., 1963, 70, 42-51. * Lohmann, W. Uber die Frage: Konvergenz- oder Akkomodations- Verengerung der Pupille bei der Naheinstellung? Vers. d. Ophthal. Ges. Heidelberg, 1908, 264-277. * Lowenstein, O. Der Psychische Restitutionseffekt. Das Prinzip der psychisch bedingten Wiederherstellung der ermiideten, der erschopften und der erkrank);en Funktion. Basel: Benno Schwabe, 1937. ~~ * Lowenstein, O. Les troubles du reflexe pupillaire a la lumiere dans les affections syphilitiques du systeme nerveux central" Paris: Doin & Cie, 1939. " ~ * Lowenstein, O. The Argyll Robertson pupillary syndrome. Mechanism and localization. Amer. J. Ophthal., 1956, 42, 2, 105-121. * Lowenstein, O. Electronic pupillography. Why, how and when? Eye, ear, nose & throat monthly, 1959, 38, 549-550. * Lowenstein, O., & Givner, I. Pupillary reflex to darkness. Arch. Ophthal., 1943, 30, 603-609. * Lowenstein, O., & Loewenfeld, Irene E. Mutual role of sympathetic and parasympathetic in shaping of the pupillary reflex to light. Arch. neurol. Psychiat., 1950, 64, 341-377. * Lowenstein, O., & Loewenfeld, Irene E. Types of central autonomic innervation and fatigue. Arch. neurol. Psychiat.. 1951, 66, 580- 599. "

95 * Lowenstein, O., & Loewenfeld, Irene E. Disintegration of central autonomic regulation during fatigue and its reintegration by psycho- sensory controlling mechanisms. J. nerv. ment. Dis., 1952, 115, 1-21; 121-145. * Lowenstein, O., & Loewenfeld, IreneE. Electronic pupillography. A new instrument and some clinical applications. Arch. Ophthal., 1958, 59, 352-363. * Lowenstein, O., & Loewenfeld, Irene E. Scotopic and photopic thresh- olds of the pupillary light reflex in normal man. Amer. J. Ophthal., 1959, 48,2, 87-98. (a) " * Lowenstein, O., & Loewenfeld, Irene E. Influence of retinal adapta- tion upon the pupillary reflex to light in normal man: Part I: Effect of adaptation to bright light on the pupillary threshold. Amer. J. Ophthal., 1959, 48,2, 536-549. (b) * Lowenstein, O., & Loewenfeld, Irene E. Influence of retinal adapta- tion upon the pupillary reflex to light in normal man. Part II: Effect of adaptation to dim illumination upon pupillary reflexes elicited by bright light. Amer. J. Ophthal.. 1961, 51,4, 644-654. * Lowenstein, O., & Loewenfeld, .Irene E. The sleep-waking cycle and pupillary activity. Ann. N.Y. Acad. Sci., 1964, 117: 142-156. * Lowenstein, O., Feinberg, R., & Loewenfeld, Irene E. Pupillary movements during acute and chronic fatigue. A new test for the objective evaluation of tiredness. Invest. Ophthal., 1962, 2,2, 138-157. * Lowenstein, O., Kawabata, H., & Loewenfeld, Irene E. The pupil as indicator of retinal activity. Amer. J. Ophthal. 1964, 57, 569-596. * Luckiesch, M., & Moss, F. K. Size of pupil as possible index of ocular fatigue. Amer. J. Ophthal., Ser. 3, 1933, 16, 393-396. * Luckiesch, M., & Moss, F. K. A correlation between pupillary area and retinal sensibility. Amer. J . Ophthal., Ser. 3, 1934, 17, 598-601. (a) '" * Luckiesch, M ., & Moss, F. K. Area and brightness of stimulus related to the pupillary light reflex. J. opt. Soc. Amer., 1934, 24, 130- 134. (b) * Lythgoe, R. G. Med. Res. Comm. Special Report #173: Reports of the Committee upon the physiology of vision. 1932.

* Machemer, H. Beitrage zur Physiologie und Pathologie der Pupille. Klin. Mbl. Augenheilk., 1933, 91, 302-316. * Machemer, H. Beitrage zur Physiologie und Pathologie der Pupille. II. Mitt. Uber den Ablauf des normalen Lichtreflexes. Klin. Mbl. Augenheilk.. 1935, 94 (N.F. 59), 305-319. * Machemer, H. Beitrage zur Physiologie und Pathologie der Pupille. III. Mitt. Die Latenzzeit des Lichtreflexes der normalen mensch- lichen Pupille. Klin. Mbl. Augenheilk., 1941, 106, 385-389. * Magitot, A. L'apparition precoce du reflexe photomoteur au cours du developpement foetal. Ann, d'oculist, 1909, 141, 161-181. * Magitot, A. L'iris. Paris: Doin, 1921. * Magitot, A. Physiologie oculaire clinique. Paris: Masson, 1946. * Marg, E., & Morgan, M. W. Jr. The pupillary near reflex. The relation of pupillary diameter to accommodation and the various components of convergence. Amer . J . Optom., 1949, 26, 183- 198. * Marg, E., & Morgan, M. W. Jr. The pupillary fusion reflex. Arch. Ophthal., 1950, 43, 871-878. (a) * Marg, E., & Morgan, M.W. Jr. Further investigation of the pupillary near reflex. The effect of accommodation, fusional convergence and the proximity factor on pupillary diameter. Amer. J. Optom.., 1950, 27, 217-225. (b) / Marina, A. Uber die Pupillenreaktion bei der Konvergenz. Neurol. Zbl.. 1902, 21, 980. * Marina, A. Uber die Contraction des Sphincter iridis bei der Conver- genz und die Seitenbewegungen der Bulbi. Zeitschr. f. Nerven- heilk., 1903, 24, 274-284. ~ ~ * Mazzucconi, M. La riflessometria pupillare con 1'appareccho Kofman & Bujadoux modificato- Ann. di Ottalm.. 1925, 53, 808-813. * Mehrtens, H. G., & Barkan, O. The use of the pupilloscope in neurology. Cal. St. J. Med., 1923, 21, 13-14. * Meineri, L. Pupilloscopio differnziale di Hess. Giorn. di Med. milit., 1927, 75, 593-606. " * Moderow, F. Das Verhalten der Pupillen bei der Konvergenz und Akkomodation. Dissertation, Univer. Marburg, 1905. Leipzig: A. Hoffman, 1905.

97 * Modonesi, F. Contributo alla seminologia pupillare con presentazione ed illustrazione di un nuovo apparecchio "II Pupilloscopio" e i suoi vantaggi. Bull, della Sci. med., Bologna, 1929, Vol. 101, 133- 155. * Moon, P., & Spencer, D. E. On the Stiles-Crawford Effect. J. opt. Soc. Amer.. 1944, 34, 319-329. * Morone, G. Richerche pupillografiche nell'abbagliamento. Boll. d'Oculist., 1948, 27, 639-652. $ Naquin, H. Argyll Robertson pupil following Herpes Zoster Ophthal - micus. With remarks on the efferent pupillary pathways. Amer. J. Ophthal., 1954, 38, Pt. 1, 23-31. * Nayrac, P., & Franchomme, J. Un procede de mesure numerique du reflexe photomoteur. Echo me"d. du Nord, 1937, 7, 405-413. * Nicolai, G. F. Erfahrung mit Pupillenmessung. Rev. Asoc. med. Argent.. 1929, 221-267. Cited in Zbl. OphthalTT" 1930, 22, 375. (Abstract) * Olbers, —. De mutationibus oculi internis. Dissertation, Univer. Gottingen, 1780. Cited by H. Aubert, 1876; M. Landolt, 1920; A. Fuchs, 1903. * Ovio, G- Sui movimenti pupillari. (15thCongr. dell'Ass. Oftalm. Ital.) Annali di Ottalm., 1898,28,89. (Abstract) Cited in Jber. Ophthal., 1898, 29. (Abstract, Schenk) * Ovio, G. Movimenti pupillari, intensita luminosa, accomodazione. Annali di Ottalm., 1905, 34, 102-146.' * Ovio, G. Ueber den Pupillenerweiterungskoeffizienten (19th Vers. ital. Ges. Ophthal., Parma, 1-4 Oct.) Zeitschr. Augenheilk., 1908, 19, 77-78. (Abstract) * Petersen, P. Die Pupillographie und das Pupillogramm. Eine methodologische Studie. Acta Physiol. Scand., 1956, 37, Suppl. 125. * Phillips, L. R. Some factors producing individual differences in dark adaptation. Proc. Roy. Soc., London, Ser. B, 1939, 127, 405-424. Cited by N. M. J. Schweitzer, 1955. * Pick, —. Konvergenz- oder Akkomodationsbewegungen. (Vereind. Augenarzte v. Ost- u. West-Preussen, Sitz. 28. Febr., Kb'nigs- berg i. Pr.) Zeitschr. f. Augenheilk., 1920, 44, 329. (Abstract)

98 * Piltz, J. Ein neuer Apparat zum Photographieren der Pupillen- bewegungen. Neurol. Zentralbl., 1904, 17, 18, 801-811; 853-857. * Plateau, —. L'Institut, J. ge"n. des Soc. et Traveaux sci. de la France et de 1'Etranger, 1835, No. 103. Cited by F. Moderow, 1905; E. Wlotzka, 1905; V. Gualdi, 1931. * Plempius, V. F. Ophthalmographia, sive tractatio de oculo. 3rd ed. Lovanii: Nempaei, 1659. Cited by C. Eckhard, 1885; J. Budge, 1855; J. Hirschberg, 1908-1918. * Polimanti, O. Sulla Valenza motoria della pupilla. Arch. di Ottlam., 1907, 14, 85-93, also, Arch. ital. de Biol., 1907, 47, 400. * Poursines, Y. Nouvel appareil pour la recherche du reflexe pupil - laire. L'isophote de Walter. Marseille med., 1932, 692, 320- 324. * Reeves, P. Rate of pupillary dilation and contraction. Psychol. Rev.. 1918, 25, 330-340. * Reeves, P. The response of the pupil to various intensities of light. J. opt. Soc. Amer., 1920, 4, 35-43. * Renard, G., & Massonnet-Naux, —. La synergie pupillaire a la con- vergence. Arch. d'Ophthal., 1951, N.S. 11, 137-145. * Sachs, M. Uber den Einfluss farbiger Lichter auf die Weite der Pupille. Pfluger's Arch. ges. Physiol., 1892, 52, 79-86. * Sachs, M. Eine Methode der objektiven Prufung des Farbensinnes. v. Graefe's Arch. Ophthal., 1893, 39, Ft. 3, 108-125. * Sachs, M. Zum Nachweis der hemiopischen Pupillarreaktion. Pfliiger's Arch. ges. Physiol., 1910, 136, 402-411. * Samojloff, A. J. Die Lokalpupillographie bei Sehbahnenlasionen. Zeit. arztl. Fortbildung, 1959. (Ber. 3. Tag. Augenarzte DDR, Dresden, 16-18 April) * Samojloff, A. J., Sokolova, O. N., & Shakhnovitch, A. R. (Pupil- lographic method in the study of the act of convergence.) Biofizika, 1961,6. In Russian. Biophysics, Royer &Roger, Inc. Amsterdam: Elsevier Publishing Co., 1961. (Trans, in English) * Samojloff, A. J., & Shakhnovitch, A. R. (A new method of local pupil- lography and its applications in physiology and in the clinic.) In Russian. Vestnik Akad. Med. Nauk., SSSR, 1958, 13, 47-58.

99 * Sander, E. Uber quantitative Messung der Pupillenreaktion und einen in der Praxis hiefiir geeigneten einfachen Apparat. Klin. Mbl. Augenheilk.. 1929, 83, 318-322. / Santamaria, — . Le oscillazioni pupillari consecutive al riflesso fotomotore. II Policlinicio, Ser. Prat., 1926, 145. * Sattler, C. H. Zwei Falle von chronischer progressiver Ophthalmo- plegie. Vorhandene Lichtreaktion bei fehlender Konvergenzreak- tion. Deutsch. med. Woch., 1920, 40, 793. * Sautter, H. Ein neues Gera't zum Nachweis der hemianopischen Pupillenstarre (Hemikinesimeter). Ber. D. Ophthal. Ges., 1950, 55, 380-382. * Schadow, —. Beitrage zur Physiologie der Irisbewegung. v. Graefe's Arch. Ophtal., 1882, 28,3, 183-200. * Schafer, G. Uber die Untersuchung auf Anisocorie ohne Pupillenstarre. Dissertation, Univer. Giessen, 1899. Altona: Hammerich & Lesser, 1899. * Schafer, G. Wie verhalten sich die Helmholtzschen Grundfarben zur Weite der Pupille? Z. Psychol. Physiol. Sinnesorg, 1903, 32, 416-419. ~ —- * Scheiner, C. Oculus, hoc est: fundamentum opticum, in quo ex accu- rata oculi anatome, abstrusarum experimentiarum sedula per- vestigatione, etc. Oeniponti, D. Agricolam, 1619. * Schirmer, O: Untersuchungen zur Physiologie der Pupillenweite. v. Graefe's Arch. Ophthal.. 1894, 40, 8-21. * Schirmer, O. Zur Methodik der Pupillenuntersuchung. Deutsch. med. Woch., 1902, 218-221. (a) * Schirmer, O- Noch einmal die Methodik der Pupillenuntersuchung. Deutsch. med. Woch., 1902, 13, 412. (b) * Schlesinger, E. Pupillometer. Med. Klin., 1907, 3, 205-206. * Schlesinger, E. Uber den Schwellenwert der Pupillenreaktion und die Ausdehnung des pupillomotorischen Bezirkes der Retina. Unter- suchungen auf Grund einer neuen Methodik. Deutsch. med. Woch., 1913, 39, No. 4, 163-165. * Schlesinger, E. Demonstration eines Pupillenmessapparates. (Berl. Ges. Psychiat. Nervenkrankh. 15. Juni) Neurol. Centralbl., 1914, 33, 863 (Summary by author)

100 * Schlesinger, E. Ein neuer Apparat zum Studium der Physiologie und Pathologie des Pupillenreflexes (Peripupillometer). Zeitschr . f. Augenheilk.. 1927, 62, 96-97. * Schmelzer, H. Eine einfache Untersuchungsmethode auf hemianopische Pupillenstarre mittels der Spaltlampe. Klin. Mbl. Augenheilk., 1931, 87(N.F. 52), 200-203. * Schober, H. Akkomodation, Konvergenz und Pupillenweite. Mschr. Feinmechanik & Optik, 1953, 70, 135-136. * Schubert, G., & Burian, H. Die Fusionsreaktion, eine bisher unbekannte Reaktion der Pupille. Pfluger's Arch. ges. Physiol., 1936, 238, 184-186. * Schweitzer, N. M. J. Threshold measurements on the light reflex of the pupil in the dark-adapted eye. Dissertation, Univer. Utrecht, 1955. S'Gravenhage: H. Junck, 1955. * Schweitzer, N. M. J. Threshold measurements on the light reflex of the pupil in the dark-adapted eye. Doc. Ophthal., 1956, 10, 1-78. .* Schweitzer, N. M. J., & Bouman, M. A. Threshold measurements on the light reflex of the pupil. Ophthalmologica, 1956, 132, 286-288. ^ " * Schweitzer, N. M. J., & Bouman, M. A. Differential threshold measurements on the light reflex of the human pupil. Ophthalmo- logica, 1958, 135, 298-300. * Shakhnovitch, A. R. (Cinematographic investigation of orienting and defensive pupillary reflexes.) In Russian. Akademia Nauk SSSR . Probl. physiol. Optics, 1958, 12, 175-180.(b) * Shakhnovitch, A. R. (Concerning the method of investigation of color vision in animals.) In Russian. Akademia Nauck (SSSR) Biophysics, 1959, 4, 367-371. Summary in English. * Shakhnovitch, A. R. (On the principle of scanning the pupil.) In Russian. Akademia Nauk SSSR, 1960, 5, 494-496. (a) * Shakhnovitch, A. R. (Hemianopic pupillary reactions in disorders of the optic tract at different levels.) In Russian. Vop . Neirokhir., 1960, 24, 20-24. (b) * Shakhnovitch, A. R., & Shakhnovitch, V. R. (Scanning pupillograph.) In Russian. Vop. Neiokhir., 1961, 25, 57-59. * Shakhnovitch, V. R. (A method of cinematographic recording of pupil reflexes and its diagnostic significance.) In Russian. Bjull. eksper.

101 Biol. i Med., 1957, 105-108. Summary in English. Cited in Zbl. ges Ophthal., 1958, 73, 254. * Shakhnovitch, V. R. (Cinematographic investigation of pupillary reac- tion to convergence.) In Russian. J. Physiol. USSR, 1958, 44, 170-172. (a) ~~ " * Silberkuhl, W. Untersuchungen iiber die physiologische Pupillenweite. v. Graefe's Arch. Ophthal., 1896, 42,3, 179-187. * Smythe, R. H. Veterinary ophthalmology. Baltimore: Williams & Wilkins, 195(T * Sommer, R. Verbesserter Pupillenmessapparat mit Messung von Reiz und Wirkung. (Vers. sudwestdtsch. Nervenarzt. Frankfurt a.M. 12.Nov.) Miinch. med. Woch., 1899, 1657. * Spring, K. H., & Stiles, W. S. Variation of pupil size with change in the angle at which the light stimulus strikes the retina. Brit. J. Ophthal., 1948, 32, 340-352. * Stark, L. Biological rhythms, noise and asymmetry in the pupil- retinal control system. Ann. N. Y. Acad. Sci., 1962, 98, 1096-1108. (a) * Stark, L. Environmental clamping of biological systems: pupil servo- mechanism. J. opt. Soc. Amer., 1962, 52, 925-930. (b) * Stark, L. The effect of drugs on the transfer function of the human pupil system. MIT: Res. Lab. Electronics, Quart. Prog. Rep., 1963, No. 68. * Stark, L., & Baker, F. Stability and oscillations in a neurological servomechanism. J. Neurophysiol., 1959, 22, 156-164. * Stark, L., Campbell, F. W., & Atwood, J. Pupil unrest: an example of noise in a biological servomechanism. Nature, 1958 (Sept.), No. 4639, 857-858. * Stark, L., & Cornsweet, T. N. Testing a servoanalytic hypothesis for pupil oscillations. Science, 1958 (March), Vol. 127, 588. * Stark, L., Redhead, J., & Payne, R. C- Asymmetrical behavior in pupil system. MIT: Res. Lab. Electronics, Quart. Prog. Rep., 1961, No. 61. * Stark, L., & Sherman, Ph. M. A servoanalytic study of consensual pupil reflex to light. J. Neurophysiol., 1957, 20, 17-25.

102 * Stark, L., Van der Tweel, H., & Redhead, J. Pulse response of the pupil. Actaphysiol. Pharmacol. Neerl., 1962, 11, 235-239. * Stegemann, J. Uber den Einfluss sinusformiger Leuchtdichtenanderun- gen auf die Pupillenweite. Pfluger's Arch. ges. Physiol.. 1957, 264, 113-122. * Steinhardt, J. Discrimination in the human eye. I. The relation of AI/I to intensity. J. gen. Physiol., 1936-1937, 20, 185-209. Cited by G. A. Fry & M. J. Allen, 1953. * Stern, H. J. Simple method for early diagnosis of abnormalities of pupillary reaction. Brit. J. Ophthal., 1944, 28, 275-276. * Stiles, W. S. The effect of glare on the brightness difference thresh- old. Brit. Dept. Sci. Ind. Res.: Illum, res. tech. Paper, 1929, No. 8. * Stiles, W. S. The scattering theory of the effect of glare on the brightness difference threshold. Proc. Roy. Soc. London B, 1930, 105, 131-146. ~: * Stiles, W. S., & Crawford, B. H. The luminous efficiency of rays entering the eye pupil at different points. Proc. Roy. Soc. London B, 1933, 112, 428-450. '~~~ * Stoewer, — . Ein Fall von Sehnervenatrophie bei Diabetes nebst Bemerkungen iiber Pupillenreaktion bei Durchleuchtung der Sklera. (Rhein. Westf. Augenarzt. Vers. Nov. 6) Klin. Mbl. Augenheilk., 1903, 41, 97-100. * v. Studnitz, G. Pupillarreaktion und Helligkeitswahrnehmung. Naturwiss., 1934, 61, 627-633. * Talbot, S. A. Pupillography and the pupillary transient. Dissertation, Harvard Univer., Cambridge, 1938. (a) * Talbot, S. A. Cinematography of the pupil. Amer . J . Physiol., 1938, 123, 200. (b) * Tange, R. A. Die normale Pupillenweite nach Bestimmungen in der Poloklinik. Dissertation, Univer. Amsterdam, 1901. Also, Arch. f. Augenheilk., 1903, 49-61. * Thomson, L. C. Binocular summation within the nervous pathways of the pupillary light reflex. J . Physiol., 1947, 106, 59-65. * Toulant, P. Le reflexe pupillaire a la lumiere ultraviolette. (39th Congr.)Soc. franc.. d'Ophtal., 1926, 155-159.

103 * v. Trautvetter, D. Uber den Nerv der Accomodation. v. Graefe's Arch. Ophthal., 1866, 12, 95-149. * Troelstra, A., Zuber, B. L., Simpson, J. I., & Stark, L. Pupil variation and disjunctive eye movements as a result of photic and accommodative stimulation. MIT: Res. Lab. Electronics, Quart. Prog. Rep., 1963, No. 69, 250-253. * Tschirren, B. Die fortlaufende Registrierung des Pupillenreflexes. Helv. Physiol. Acta, 1947, 5, C57-C58. * Ulbrich, H. Instrument zur Prufung der hemiopischen Pupillenreak- tion. (Wien. Ophthal. Ges. 9. Feb) Klin. Mbl. Augenheilk.. 1914, 536. Cited by C. Behr, 1924. * Van der Tweel, L. H., & van der Gon, D. The light reflex of the normal pupil of man. Acta Physiol. Parmacol. Neerl., 1959, 8, 52-88. (English version of thesis, L. H. Van der Tweel, 11 July 1956) * Venco, L. & Marucci, L. Rllievi pupillografici sul riflesso alla luce e dall'oscuramento in condizioni fisiologiche. Ann. diOttalm. Clin. Ocul., 1947, 73, 486-494. * Veraguth, — . Zur Prufung der Lichtreaktion der Pupille. Neurol. Zentralbl., 1905, 24, 338-343. * Veraguth, —. Zur Fage nach dem pupillomotorischen Feld der Retina. Neurol. Zentralbl., 1908, 27, 402-404. * Vervoort, H. Die Reaction der Pupille bei der Accomodation und der Convergenz und bei der Belichtung verschieden grosser Flachen der Retina mit einer constanten Lichtmenge. v. Graefe's Arch. Ophthal., 1900, 49, 2, 348-374.. ~ * v. Vintschgau, M. Zeitbestimmungen der Bewegung der eigenen Iris. Pfliiger's Arch, ges. Physiol., 1881, 26, 324-390. * v. Vintschgau, M. Weitere Beobachtungen u'ber die Bewegungen der eigenen Iris. Pfliiger's Arch. ges. Physiol., 1882, 27, 194-196. * Vogt, A. Ein neues Hemikinesimeter. (Ges. Schweitz. Augenairzt. 15, Schaffhausen) Klin. Mbl. Augenheilk., 1922, 69, 119. * Wagman, I. H., & Gulberg, J. E. The relationship between mono- chromatic light and pupil diameter. The low intensity visibility curve as measured by pupillary movements. Amer. J. Physiol., 1942, 137, 769-778. ''"

104 * Wagman, I. H., & Nathanson, L. M. Influence of intensity of white light upon pupil diameter of the human and of the rabbit. Proc. Soc. exper. Biol.. 1942, 49, 466-470. i Walker, C. B. Further observations on the hemiopic pupillary reac- tion obtained with a new clinical instrument. J. AMA, 1914, 63, 846-851. Cited by C . Behr, 1924. * Walter, A. Isophote, pour le recherche du reflexe pupillaire. Gaz . d'hop., 1933, 106, 76-77. * Weber, E. H. Tractatus de motu iridis. Lipsiae: Gluckii, 1821. * Weber, E. H. Additamenta ad tractatum de motu iridis. Pars I. Lipsiae: 1823. * Weber, E. H. Annotationes anatomicae et physiologicae. Programma collecta. Fas. III. Summa doctrinae de motu iridis. Lipsiae: 1851. Cited by J. Budge, 1855; C. Eckhard, 1885; V. Gualdi, 1931; J. Hirschberg, 1908-1918; A. Drouin, 1876; E. Marg & M. W. Morgan, Jr., 1949; F. E. Donders, 1888; H. Vervoort, 1900; G. A. Fry, 1945. * Weiler, K. Demonstration eines neuen Pupillenmessapparates. III. Jahresvers. (Verein Bay. Psychiat. Munchen, June 1905) Neurol. Centralbl., 1905, 24, 682-683. * Weiler, K. Untersuchungen der Pupille und der Irisbewegungen beim Menschen. Zeitschr. ges Neurol. Psychiat., 1910, II, 101-274. * Wernicke, C. Ueber hemiopische Pupille nreaktion. Fortschr . Med ., 1873, 1,2, 49-53. * Weve, H. Vorrichtung zur Untersuchung der hemiopischen Pupil- larreaktion. Klin. Mbl. Augenheilk., 1918, 61, 140. * Weve, H. Zur Physiologie des Lichtreflexes der Pupille. v. Graefe's Arch. Ophthal., 1919, 100, 137-156. * Wilbrand, H. Ueber Hemianopsie und ihr Verhaltniss zur topischen Diagnose der Gehirnkrankheiten. Berlin: Hirschwald, 1881. * Wilbrand, H., & Saenger, A. Die Neurologie des Auges. Vol. 7. Die Erkrankungen der Sehbahn vom Tractus bis in den Cortex. Die homonyme Hemianopsie, nebst ihren Beziehungen zu den anderen cerebralen Herderscheinungen. Wiesbaden: J. F. Bergmann, 1917. * Wirth, A. Comportamento del riflesso pupillare fotomotore nell' antagonismo binoculare dei colori. G. Ital. Oftal., 1952, 5, 22-29.

105 * Wlotzka, E. Die Synergie von Akkomodation und Pupillenreaktion. Pfliiger's Arch. ges. Physiol., 1905, 107, 174-182. • • n ^ n_ * Wolff, H. Uber Pupillenreaktionspriifung rnit Berucksichtigung der Refraktion des untersuchten Auges sowie iiber eine zentrale und periphere Pupillenreaktion, nebst Angabe eines neuen Instrumentes. Berl. klin. Wochenschr., 1900, No. 28, 613-615. * Wolff, H. Uber die Abnahme der Pupillenreflexempfindlichkeit der Netzhaut vom Zentrum nach der Peripherie. Z. Augenheilk., 1904, 12, 644-650. (a) ' * Wolff, H. Bemerkungen iiber die Arbeit "Uber die Abhangigkeit der Pupillarreaktion von Ort und Ausdehnung der gereizten Netahaut- f lache " von Dr. G. Abelsdorff u. Dr. H. Feilchenfeld inBd. 34 dieser Zeitschrift. Z. psych. Physiol., Sinnesorg, 1904, 36, 93-97. (b) * Wybar, K. C. Ocular manifestations of disseminated sclerosis. Proc. Roy. Soc. Med.,1952, 45, 315-320. * Young, F. A., & Biersdorf, W. R. Pupillary contraction and dilation in light and darkness. J. comp. physiol. Psychol., 1954-1955, 47, 264-268. ' ~~^ ~ * Zeldenrust, E. L. K. Uber die Chronaxie des Lichtreflexes der Pupille. Arch. Augenheilk., 1931, 104, 585-593.

106 THE FUSION REFLEX Kenneth N. Ogle Mayo Clinic and Mayo Foundation From earliest times, man has been puzzled over one of the amazing facts of vision: namely, that whereas he has two eyes he experiences generally only a single perception of space. Speculation as to the reason for this phenomenon had been made long before anything was known about the neuro-anatomical structures of the visual pathways. Most popular of these speculations was that one or the other of the uniocular images was alternately suppressed. This notion was thought to be reinforced to some extent by the experiments with retinal rivalry. Leaving aside these speculations, however, it is now usually said that the neural excitations from the images on the two retinas "fuse"— the process involved is called fusion, and takes place in the occipital cortex of the brain. The term fusion sometimes carries other connota- tions, but here it is used only to convey the idea that when both eyes enter the visual act with appropriate fixation of both eyes, the observer perceives singly. (This terminology will be used herein without implying any notions as to the exact nature of the process.) It is the neuro-anatomy of the visual processes, of course, which makes this fusion possible and which provides the basis for it, but this cannot be the cause of fusion. Fusion cannot be just a complete unifica- tion of the excitations from the two eyes alone, either, for one must assume that the integrity of both images is always maintained. To provide for this sensory aspect of fusion all movements of the two eyes must be executed and coordinated in the interest of maintaining single vision, and especially for fusion of the images of the fixation ob- ject. Eye movements made to preserve single vision are known as fusional movements. To be precise, therefore, the fusion reflex is actually a reflex concerned with the motor control of the eyes in order to preserve the interests of the fusion of the images of the object fixated. At the outset, to understand the phenomenon of the fusion of the images from the two eyes, one must be familiar with the principles of corresponding retinal points and of Hering's law of identical visual direc- tions, which define corresponding retinal points. Only through these can one have a similar topographical representation of visual space from the

107 optical images that fall on the two retinas. In this representation there must be a common point of fixation of the two eyes, the images of which fall on the foveas of the two eyes. The neural excitations from any pair of these points give rise to the same subjective visual direction. The study of the fusion processes in general has proven difficult. As in the study of any natural phenomenon, however, insight is gained into the phenomenon only with respect to those procedures with which it can be changed. Present knowledge of the fusion reflex has come through the following: 1 . forced changes in convergence of the eyes—horizontally and vertically, and changes in a cyclotorsion or "twisting" of the images about the anterior-posterior axes of the eyes; 2. studies of the above changes with modification of the visible patterns; 3. anomalies of fusion found in subjects with impaired binocular vision (mostly with strabismus); 4. studies of physiologically "perceived" double images. Consider the following simple and perhaps trite experiment (Fig. 1). While fixating a point on a chart across the room, suddenly interpose be- fore one eye an ophthalmic prism of, say, three prism diopters, base-out. Instantly, one sees two charts which appear to move toward one another, the speed of movement seeming to increase as the separation of the two decreases. Suddenly, the two charts appear to coa- lesce—only one chart is seen. A person watching during this time can readily verify that the eye behind the prism has turned inward —the convergence of the eyes has been increased. One eye still fixates the chart, but the other fixates the displaced image of the chart produced by the prism. This experiment demonstrates the phenomenon designated as the com- pulsion-to-fusion reflex. If the images are widely separated immediately after the prism is inserted, the observer may be able to exert a certain degree of vol- untary control over the movement of the two half-images. However, if the sepa- ration is small the fusional movements under ordinary conditions are involuntary and compulsive, that is, can seldom be changed at will. Fixation point R.E. Fig. 1. Simple experiment to demonstrate phenom- enon of compulsion to fusion.

108 When the images come together at the end of a fusional movement, the sensation of the observer is that the two seem to lock together. It is this sensation that probably gave rise to the term fusion. All subsequent fixation eye movements to different parts of the charts are precisely co- ordinated movements, and bifoveal single vision is maintained. Coordi- nated eye movements such as these are not, however, a property of fusion per se. When an occluder is placed before one eye so that binocular per- ception is prevented, the eye movements are still more or less coordi- nated, that is, said to be concomitant, although, because of the usual ocu- lomoter imbalance of the subject, both eyes are not directed to the same point on the chart. The fusion reflex, therefore, adds something new in that, in the interest of single vision, innervations to the extra-ocular muscles of the two eyes are directed to the same point, or nearly so. The phenomenon of the compulsion to fusion just demonstrated has led to many notions about fusion. Of particular interest is the strength of this compulsion for fusional eye movements, and even the speed with which the coordinated eye movements can lead to a resumption of single binocular vision (Brecher, 1954). One hopes for some kind of a measure of this reflex, for it probably varies among individuals and might serve a clinical purpose. Certainly the ophthalmologist often finds persons, many with strabismus, who have impediments to binocular vision of several types, and also who, under suitable conditions, can see both half-images. There are patients for whom the images in the two eyes cannot be made to appear superposed. As attempts are made to cause the image from one eye to pass over the image of the other eye in a stereoscope or haploscope by control of target positions, the two images first appear double-crossed and then double-uncrossed. The subject is unable to see both images at the same time in the same direction. In these instances there is almost a repulsion, or a negative compulsion reflex to fusional movements. There is also the baffling problem presented by the person with one eye habitually misdirected, the other eye being called the sighting eye. The misdirected eye sees the image of the object fixated by the sighting eye by peripheral vision, but oddly perceives that object in the same or nearly the same visual direction as does the fovea of the fixating eye. Usually, but not in all instances, the two eyes are relatively concomi- tantly coordinated. Generally, there is reason to believe that, despite the fact that visual direction of the squinting eye is similar to that for the sighting eye, true fusional movements cannot usually be demonstrated. Finally, there are subjects whose ability to maintain binocular vision is very tenuous. Small changes in convergences introduced by prisms of low deviating power before one eye cause a doubling of the images. Or small changes in ophthalmic lenses before the eyes may cause doubling. In these individuals fatigue may also readily give rise

109 to diplopia. Can it be said that the strength of the compulsion for fusional movements in these individuals is weak? The phenomenon found in the individuals with strabismus is evidence that a functioning of the normal neuro-anatomical structures of the visual processes to the occipital cortex is a prerequisite for true fusion and fusional movements. To find a measure of the reflex compulsion for fusional eye move- ments, an attempt is usually made to force a change in the normal point- ing of the eyes, that is, to force a change in the convergence of the eyes while keeping the stimulus to accommodation constant. These changes can be produced by artificial means, such as by ophthalmic prisms before the eye, or by the use of such instruments as the haploscope. Usually one seeks to find the greatest change in convergence (the magnitude of prism power) that can be produced just before diplopia or double images are perceived. The larger the range of this convergence-divergence change, the greater is said to be the strength of the reflex compulsion for fusional eye movements. However, in this procedure one must consider the role played by an increasingly embarrassed relationship between the accommodation and convergence of the two eyes, which may be a govern- ing factor. Perhaps, mechanical limitations may also be imposed by the properties of the extra-ocular muscles themselves. The essence of the sensory aspects of fusion lies in there being in the visual field similar contours, for the fusion compulsion strives to keep the cortical excitations arising from images of these contours unified by appropriate innervations to the extrinsic muscles of the two eyes. The importance of contours cannot be over-emphasized. In Fig. 2 are shown illustrations of certain kinds of contours that might be used in a stereo- scope . Where contours are identical, fusion—unitary perception—of both images easily occurs. One can verify that both eyes are involved in the visual act by the use of check marks (the dot and cross). In the third set, the line viewed by the- left eye is wider than that viewed by the right, but the contours of the edges are similar. When viewed in the haploscope it can be seen that the narrow line appears to "stick" to either the right edge or the left edge of the wider line; which edge it sticks to depends on the direction of the oculomotor imbalance, the heterophoria of the observer. Thus, one cannot say that the image of the thin line fuses with that of the wide, only that the images of one or the other of the similar contours fuse. If circles of unequal size are presented, again one edge of the smaller circle usually appears to "stick" to one edge of the larger. Which side of the larger again depends on the direction of the oculomotor im- balance under the conditions of observation.

110 Left Right Binocular Defective but partially sim- ilar contours may also be suffi- cient for fusion and for controlling the oculomotor adjustments of the eyes. However, if the contours are very dissimilar either the images appear to move randomly, or, as is more probable, to move to the phoria position, or there occurs the familiar phenomenon known as retinal rivalry. Even in these circumstances, there tends to be a marked persistence of the visibility of the contours from both eyes, or more often a • random alternation of perception of the dissimilar contours. Small wheel-like rotations of the eyes about their lines of fixation —anterior-posterior axes of the eyes—are called cyclotor- sional movements. These move- ments can and do occur constantly in the normal use of the two eyes so as to maintain single binocular vision. These movements then are called cyclofusional movements (Ogle & Ellerbrock, 1946). Cyclo- fusional movements occur more readily when the visible contours are oriented more or less vertically than when the contours are oriented nearly horizontally. It can be said that the fusional movements of the two eyes provide bifoveal fixation on the object of interest. However, one must distinguish between bifoveal fixation brought about by the fusion reflex and the so- called fixation reflex, which is a monocular phenomenon. Bifoveal fixa- tion with fusion is not the result of two monocular fixation reflexes acting at the same time, for it can be shown readily that images of contours that fall on the peripheral parts of the retinas of the two eyes are sufficient to control the pointing of the eyes through the fusing of these images (Burian, 1939). In the absence of foveal stimuli, prism vergences of the eyes wholly from peripheral stimuli can be demonstrated easily. Dis- placed peripheral contours, if sufficiently plentiful, can cause a doubling of non-displaced contours seen only foveally. In most demonstrations of the compulsion for fusional movements, a large number of contours in a single vertical plane are commonly used. Fig. 2. Types of contours that do and do not provide fusion.

1ll When a prism is placed before one eye large and equal disparities are introduced between the images of all the contours on the surface. The disparities of the images of all the contours, therefore, provide the in- nervations for the fusional eye movements. The character of the fusional movements in the sense of their compulsive attribute and the speed of the recovery will be influenced by the number and complexity of the contours on this screen. The strength of the fusion compulsion as measured by prisms certainly is less for a point of light in a dark room than for the details on a wall chart. In the binocular vision of one's normal surroundings, however, the point of fixation may be in the midst of a larger number of objects in the field of view at different distances (Fig. 3). The images in the two eyes of those objects more distant than the fixation point will be uncrossed and disparate; the images of those nearer than the fixation point will be crossed and disparate. One would expect these disparities also to pro- vide stimuli for fusional movements. That these movements do not occur is usually attributed to the role of attention given to the point of interest. If the attention is suddenly given to a second object, it is thought that the fusional movements of the eyes necessary to fix bifoveally this second object then act as a reflex movement. These movements have been said to be initiated by psycho- optical stimuli (Hofmann, 1925). Images are uncrossed Fixation point Image s are crossed P L.E. R.E. Fig. 3. When eyes fixate near ob- ject, images of other ob- jects at different distances in field of view will be dis- parate . The fusional movements as- sociated with the vertical diplopia caused by the introduction of a vertical disparity by ophthalmic prisms placed base-down or base- up before an eye are thought to be purely reflex movements. Cyclo- fusional movements that occur naturally when objects fixated are inclined to the visual plane are thought to be reflex movements also. Under these conditions it is thought that psycho-optical stimuli arise only from the horizontal dis- parities. The sensory aspects of fusion and the initiation of fusional move- ments for objects spatially distrib- uted involve another well-known phenomenon (Fig. 4). If two par- allel horizontal lines are presented in a stereoscope to each of the eyes.

112 Volkmanrte figures Left Right -H HI^ _________ i f fo Panxirris area. yo.- horizontal dimension - vertical dimension Haploscopic figures used to demonstrate Panum's areas of fusion in vertical and horizontal meridians. Fig. 4. it is found that the separation of the lines on one of the targets can be increased or decreased by a small amount before the moved line is perceived double, that is, before three lines are perceived. The magnitude of the change in separation is greater than the sum of the widths of the two lines, and, hence, cannot be a function only of visual resolution. The range in which this separation can be changed is a measure of Panum's fusional area in the vertical meridian. The differ- ence in separation just as dou- bling appears is usually ex- pressed as the angle subtended at the eye expressed in minutes of arc. Similarly, in the hori- zontal meridian a range can be found within which single binoc- ular vision is reported. This range is a measure of Panum's fusional area in the horizontal meridian. The experiment is more difficult to perform in the horizontal meridian because the moved line with respect to its mate in the other eye will appear to move in stereoscopic depth. Within the ranges of separation described here, one may say the images are fused in spite of the fact that, except for some one separation, the images must be dis- parate. In the horizontal meridian these disparate images can be not only the stimulus for stereoscopic depth perception but also potential stimuli for fusional movements. In normal surroundings, say, with a fixation on a near object (Fig. 5), because of Panum's fusional areas, only objects lying in a certain re- gion would have images in the two eyes that would fall within the region of binocular single vision. All other objects that lie outside these areas would have disparate images and should be seen double. However, gen- erally one is unaware of this doubling, unless attention is called to it, because of a kind of suppression, although it is just as probable that the subject ignores the confusion of images. In spite of the fact that one is not consciously aware of the doubling, the question is whether the dis- parities between the images do not still act as potential stimuli for fusional movements when attention is directed to any other object. It is also true that for disparities within certain limits that are larger than the horizontal dimension of Panum's area, stereoscopic depth of the images is readily perceived; this implies a similar type of interaction between the excitations from the half-images in the visual cortex (Ogle, 1952a; 1952b).

113 Region ofxs'ngle binocular vision / / y Measure of Pan urn's area Fig. 5. Schematic illustration of region of binocular single vision due to existence of Panum's areas. According to presently accepted notions, when the eyes first fixate a distant object and then a near object, a change in convergence of the eyes is necessary if there is to be bifoveal fixation. The question is: Where do the innervations arise for this change in convergence of the eyes, and how much of this convergence is due to the compulsion-to- fusion reflex? Except in subjects with large oculomotor imbalances, the part of the total convergence change due to fusional movements may be quite small. This magnitude would vary among individuals. The greater part of the change in convergence comes from the accommodative- convergence synkinesis, together, perhaps, with the somewhat controver- sial psychological influence of the awareness of distance of the objects fixated. The so-called fusional convergence then makes up only for the residual convergence needed for fusion of the object fixated. One would expect that if the subject were orthophoric, that is, had no motor imbal- ance for objects at the two distances, no fusional convergence whatever would be needed, though sensory fusion of the images would, of course, be maintained. The fusional convergence is needed, therefore, only where an oculomotor imbalance (or heterophoria) between the eyes exists. Why oculomotor imbalances exist at all is not clear. They may often be due in part to the so-called tonic factors, uncorrected refractive errors, low accommodative-convergence ranges, perhaps psychological factors, et cetera. When binocular vision is prevented by a suitable occluder before one eye, for a given point of fixation by the other, and thus a given stimu- lus to accommodation, the occluded eye turns to a position such that no

114 imbalance exists between the tensions of the extra-ocular muscles, or better, an equilibrium exists in those tensions. The angle through which the eye has turned from pointing to the fixation point can be measured, for example, by using a Maddox rod as the occluder, which at the same time forms a streak image of the fixation point. By also introducing prisms of various strengths before the eye, the image of this streak can be made to appear in the same visual direction as the fixation point. The strength of the prisms used then measures the angle—the phoria. This angular deviation is said to be a measure of the heterophoria, and it is assumed also to be a measure of the oculomotor imbalance when fusion is maintained. At the request of the chairman of this symposium, a method is briefly described here for obtaining a measure of the oculomotor imbal- ance when fusion of the images and normal convergence are maintained, and when both eyes are under the same stimulus to accommodation. This method rests on the fact that, because of the existence of Panum's fusional areas, it is not necessary that the centers of both foveas be directed exactly upon the point of fixation. Thus, if an oculomotor imbalance exists for a given fixation on a given target, it is possible for the eyes actually to overconverge (Fig. 6), or underconverge by a very small angle in the True point of convergence Distal limit •Target Proximal L.E. limit \\""Fixation A disparity R.E. Fig. 6. Schematic illustration of how, under influence of convergent oculomotor imbalance, eyes will actually overconverge by a small angle, and yet images of target fixated will be seen single.

115 direction of the oculomotor imbalance, namely, an esophoria, or an exophoria, respectively (Fig. 7). This small angular error in conver- gence is called fixation disparity, because, when first discovered in horopter experiments using the nonius technique, it was clear that the images of the fixation point were actually disparate when the experi- menter was heterophoric (Ogle, 1958; 1962). Distal True point S__^/\ ^--.Target of converg ence Proximal^-'"" / \\ "~-~- limit Fixation disparity L.E. R.E. Fig. 7. Schematic illustration of how, under influence of divergent oculomotor imbalance, eyes will actually underconverge by a small angle, and yet images of target fixated will be seen single. To perceive this small angle of overconvergence or underconver- gence, it is necessary only to present details near or at the fixation point, the images of which cannot be fused, and each of which, therefore, would be perceived in the primary uniocular visual direction of each eye separately (Fig. 8). Experiments have shown that if an oculomotor im- balance is present these dissimilar details will appear displaced in spite of fusion, whether these details are seen foveally with peripherally fused images, or extrafoveally with fusion for the images of a fixation object. In a recent study, the angle of fixation disparity has been demonstrated and measured objectively. To measure subjectively the small angle of fixation disparity, it is necessary only to design an instrument so that one of the two nonius lines can be displaced laterally with respect to the other, until the subject re- ports that the two nonius lines appear in the same visual direction (Fig. 9). The actual displacement is then expressed in minutes of arc corresponding to the angle subtended by the separation of the two lines at the eyes.

116 L.E. R.E. L Appearance Fig. 8. Type of target used to demonstrate fixation disparity. L.E. R.E. Appearance Fig. 9. Type of target, in which lower nonius line can be displaced horizontally, used to measure fixation disparity. Of special interest are the studies that show the manner in which fixation disparity changes as the oculomotor imbalance between the two eyes is artificially changed by prisms or by ophthalmic lenses. In either case, the relationship between the accommodation and the convergence is altered with prisms by changing the stimulus to convergence (Fig. 10), with lenses by changing the stimulus to accommodation (Fig. 11). The pattern of response of the fixation disparity measurements to these changes varies among subjects.

117 L.E. Fig. 10. For target at given observation distance, prisms base-in before eyes reduce angle of convergence of eyes, and thus alter status between accommodation and convergence. Figure 12 illustrates typical data for a subject measured for both distant vision and near vision (Ogle & Prangen, 1951). The abscissa of this graph corresponds to the prismatic deviation (in prism diopters) in- troduced: to the right of the origin prisms base-out, which necessitates an increased convergence, and to the left of the origin prisms base-in, which necessitates a decreased convergence of the eyes if fusion is to be maintained. On the ordinate are plotted the measured angles of fixation disparity: above the origin for an overconvergence of the eyes, i.e., an esodisparity (crossed), and below the origin for an underconvergence, an exodisparity (uncrossed) —all in minutes of arc. When prisms are placed before the eyes base-in to cause a decreased convergence an esodisparity is induced, as though the eyes resisted the forced decrease in conver- gence. When prisms are placed before the eyes base-out to necessitate an increased convergence an exodisparity is induced, as though the eyes resisted the forced increase in convergence. Using the technique of ob- taining measurements of fixation disparity by alternately placing prisms of increasing power base-in and then base-out before the eyes, by short exposures of the movable nonius test line, and by not prolonging the

118 L.E. R.E. Fig. 11 . For target at a given observation distance, ophthalmic lenses of minus power before eyes cause increase in stimulus to accommodation, again altering status between accommodation and convergence. period of time used for the measurement, one obtains data that can be fitted by a smooth curve. At certain limits of prismatic power the dis- parity becomes large, and diplopia results, usually dramatically, for any further increase in prismatic deviation. These limits correspond to the usual prism vergences. For distant vision the subject whose data are shown here is virtually orthophoric, but for near vision there is a large exodeviation of 16 min- utes of arc. The point where the curve crosses the abscissa indicates the prism power for which the oculomotor imbalance has been reduced to zero. This particular prism power would be comparable to the measure- ment of the heterophoria obtained with disassociated vision. In the near graph, then, the equivalent phoria is about 13 prism diopters base-in — an exophoria. The fixation disparity will also change when lenses are introduced before the eyes to alter the stimulus to accommodation, and thus to alter the accommodation-convergence relationship as shown in the central graph of Fig. 13- On this graph the abscissa corresponds to the lens power in diopters, minus lenses to the right (because there would then be an increased stimulus to accommodation), and plus lenses to the left (there would be a decreased stimulus to accommodation). The range of lens powers that can be used is usually more limited than for prisms. If one plots the prism power that produces the same fixation dis- parity as does a given change in the stimulus to accommodation, one

119 Fig. 12. Two sets of data for distant and near observation distances. Near set shows equivalent phoria of -13 prism diopters. : A :Prism dala Minus 1 15| Lenses 1. SXN |0 i, Nx N Plus Minus H +3 +2 -Z -3 -4 .enses 5 \ i Diopters c Derived; data ' <*». .^ 3.3 VD 13. Sets of data showing change in fixation disparity with both prisms and lenses, and derived graph giving relation between change in prism vergence and change in stimulus to accommodation.

120 almost invariably obtains derived data that fall on a straight line. The slope of this line is the accommodative-convergence/accommodation ratio (Ogle & Martens, 1957). It is not desirable to go further here with the phenomenon of fixation disparity. However, on the basis of these studies it may be assumed that the fixation disparity is a measure of the oculomotor imbalance while sensory fusion is maintained and both eyes respond to the same stimulus to accommodation. This measure, then, tells how the oculomotor imbal- ance changes when changes are made either with prisms or with ophthalmic lenses. It would be concomitantly a measure of the strength of the inner- vations that must be supplied to the extra-ocular muscles of the eyes to provide the fusional convergence necessary for binocular vision to be maintained on the object fixated • One might consider the slope of the tangent to the curve at any point as an inverse measure of the strength of the fusion compulsion reflex at that point. In conclusion, this paper has presented an over-simplified discus- sion of these topics: 1 . the difficulty of determining the exact nature of fusion; 2. the relationship between the fact of sensory fusion and the reflex nature of the innervations to the extra-ocular muscles necessary to main- tain sensory fusion; 3. the demonstration of the reflex compulsion for fusional move- ments that will maintain sensory fusion, but that in normal surroundings may be due to psycho-optical stimuli; 4. the role of sensory fusion of disparate retinal images brought about by Panum's areas; 5- the difficulties of obtaining a measure of the reflex compulsion to fusional movements to preserve sensory fusion; 6. the role of fusional convergence to overcome oculomotor imbalances; 7. finally, the phenomenon of fixation disparity used as a method of measuring, on the one hand, the degree .of oculomotor imbalance with fusion maintained, and, on the other hand, the strength of the innervations to the fusional processes to maintain sensory fusion.

121 References Brecher, G. A. Quantitative studies of binocular fusion. Amer. J. 3hth.. 1954, 38, 134-141. Burian, H. M. Fusional movements; role of peripheral retinal stimuli. Arch. Ophth., 1939, 21, 486-491. Hofmann, F. B. Physiologische optik (Raumsinn). In A. Graefe & T. Saemisch, Handbuch der gesamten Augenheilkunde. (2nd ed.) Berlin: Julius Springer, 1925. Vol. 3, Ch. 13. Ogle, K. N. Disparity limits of stereopsis. Arch. Ophth., 1952, 48, 50-60 (a). Ogle, K. N. On the limits of stereoscopic vision. J. exp. Psychol., 1952, 44, 253-259 (b). Ogle, K. N. Fixation disparity and oculomotor imbalance. Amer. Orthoptic J., 1958, 8, 21-36. Ogle, K. N. The optical space sense. In H. Davson (Ed.), The eye. New York: Academic Press, 1962, Vol. 4. Ogle, K. N., & Ellerbrock, V. J. Cyclofusional movements. Arch. Ophth., 1946, 36, 700-735- Ogle, K. N., & Martens, T. G. On the accommodative convergence and the proximal convergence. AMA Arch. Ophth., 1957, 702-715- Ogle, K. N., & Prangen, A. deH. Further considerations of fixation disparity and the binocular fusional processes. Amer. J. Ophth., 1951, 34(2), 57-72.

122 VESTIBULAR MECHANISMS AND VISION1 Earl F. Miller II and Ashton Graybiel U. S- Naval School of Aviation Medicine, Pensacola This report summarizes briefly the results of some investigations carried out at the U. S- Naval School of Aviation Medicine dealing with the vestibular and visual systems under the influence of unusual gravita- tional-inertial force environments (GIFE). The initial stimulus for these researches stemmed from the need to evaluate the role of the GIFE in causing disorientation in pilots; now, space flight has provided an added impetus. It was learned early that exposure to unusual force environ- ments greatly affected the flyer through the sensory receptors in his vestibular organs, and through them visual mechanisms, and that, in turn, vision and the visual environment also affected the responses to stimulation of the vestibular organs. Thus, exposure to angular acceler- ation, a physiological stimulus to the semicircular canals, resulted in apparent movement of an object which was fixed in relation to the observer. This has been termed the oculogyral illusion, and its form bears a definite relation to the pattern of angular acceleration. Unusual patterns evoked bizarre effects. The Coriolis phenomenon may be regarded as a special instance of the oculogyral illusion in response to simultaneous rotation of the head about two axes which generates a Coriolis acceleration. Ocular nystagmus may be an associated response to angular and Coriolis acceler- ations and, when prominent, contributes to the illusion. When a person is exposed to linear acceleration with a change in direction of the resultant force relative to himself, he not only feels that he is being tilted, but he also perceives an apparent displacement of ob- jects in the visual field which tend to accord with the new direction of the mass acceleration. The visual component has been termed the oculo- gravic illusion which has quite different characteristics from the oculo- gyral or Coriolis illusions. An associated phenomenon is ocular counter- rolling which is manifested to a greater degree than when a person simply tilts with respect to the gravitational upright. This research was conducted under the sponsorship of the Office of Life Science Programs, National Aeronautics and Space Administration (Grant R-47). Opinions and conclusions contained in this report are those of the authors and do not necessarily reflect the views or endorse- ment of the Navy Department.

123 Simultaneous exposure to linear and angular acceleration alters the responses in a characteristic manner, suggesting a close functional rela- tionship between the two vestibular organs. Positional nystagmus, and positional alcohol nystagmus may be dependent on this relationship. In carrying out the experiments, an attempt was made to simulate the unusual force environments in the laboratory, but in the case of weight- lessness it was necessary to go aloft. It was relatively easy to control vision and the visual environment, and it was possible to control the in- puts from the semicircular canals and the otolith apparatus by the use of subjects who had lost the function of the canals and, probably, also of the otoliths. The investigations sought to exploit these vestibulo-visual phe- nomena: (a) in testing the function of the semi-circular canals and the otolith organs; (b) in using these phenomena as indicators in the investiga- tion of different psychophysiological mechanisms including adaptation; and (c) in attempting to show how these mechanisms are affected under the unusual force environments to which man may be subjected in an aircraft or space vehicle. Otolith Organs and Vision Linear acceleration, considered the adequate stimulus to the macular end organs, can be varied to evoke changes in overt behavior by reposi- tioning the head (actually the otolith organs themselves) with respect to gravity, moving it in a circular path at a constant rate, or in a rectilinear path at an accelerating rate. Direction of the stimulus force is controlled by orienting the head with respect to the resultant gravitational-inertial force, which varies in magnitude as a function of velocity. Normal gravi- tational acceleration can be counteracted completely or partially within the earth's gravitational field by Keplerian trajectory flight maneuvers. The otolith organs of humans can be probed remotely by linear accelera- tion in a way analogous to the direct mechanical stimulation of these organs in animals. It is possible to apply forces of various magnitudes in specific directions relative to the anatomical spatial arrangement of the otolith organs within the skull; the mode of action and role in perception of the otolith organs can then be determined indirectly by measuring external changes such as occur in ocular counterrolling and egocentric visual local- ization. Subjects with known labyrinthine defects offer a means of evalu- ating the extent to which extra-labyrinthine factors are involved in these external signs of inner ear function. Counterrolling When certain experimental procedures are followed, the conjugate rolling movement of the eyes around their lines of sight opposite to the lateral inclination of the head is generally held to be a direct reflex origi- nating in the otolith organ. The distinct advantage of having an external indicator of otolith activity which is not under voluntary control has been

124 outweighed in the past by the great difficulty in obtaining precise measure- ments of the rolling movements. Throughout the long colorful history of counterrolling studies, several methods of measuration have been used (Miller, 1962). All have as a common basis the selection of anatomical landmarks on the eye to establish a reference plane containing the line of sight for specifying the rotation of the eye around its line of sight. The most accurate of these older methods yielded a measuring error which was large even in relation to the maximum amount of counterrolling, in some cases less than six degrees (°), that can be evoked by head incli- nation . A method involving photography of natural landmarks on the iris was devised to meet the requirement of greater precision in measure- ment. A solution to the problem of measuring very small amounts of movement of these landmarks was found in simple magnification. In this procedure, the film image of the entire eye is enlarged over 300 times the actual eye size by projection onto a distant screen. Measurement of angular torsional movement around the center of the pupil is then accom- plished by superimposing upon each test image in succession a second projected image of the subject's eye serving as a standard of comparison. More complete details of this measuring technique have been published (Miller, 1962). It is sufficient for this discussion to point out that a high degree of accuracy and reliability in measurements (= ± 5 minutes [mini of arc) is possible with this procedure. Normal subjects. Counterrolling measurements using the photo- graphic technique have been made on several normal individuals tilted in 25° steps up to ± 75° from the gravitational vertical (Miller & Graybiel, 1962a). Each of these subjects revealed a qualitatively similar counter- rolling response (Fig. 1) to head inclination, but quantitatively there were interindividual differences. There were also significant right-left differ- ences in some individuals but not in others. A more extensive study (Miller, 1962) was also conducted in which measurements were made at every 15° within the frontal, sagittal, and two intermediate planes. Maximal compensatory torsional eye movement was found in the frontal plane, somewhat less in the intermediate planes, and none at all in the sagittal plane. The absence of appreciable counterrolling when the head is tilted in its sagittal plane does not justify any inference that compensatory eye reflexes arising in the otolith organs do not exist when tilting in this plane. On the contrary, there is evidence that the eyes move reflexly in a vertical direction directly counter to the fore-and-aft tilting of the head. In the counterrolling experiments, innervation to elevate or de- press the eyes was compensated by counter-fixational innervation. Coun- terrolling was always found to occur (Fig. 2) opposite to the lateral com- ponent of head tilt and to increase fairly rapidly up to maximum at a head inclination between 60° and 90° . From this point on counterrolling de- creased but at a lesser rate than it increased, reaching about zero when

125 -50 -25 0 »25 «5O »75 HEAD (BODY) TILT IN DEGREES 75 ^50 ^250 "T25t50 HEAD (BODY) TILT IN DEGREES Fig. 1. Mean counterrolling values plotted as function of leftward and rightward tilt (left column: normals; right column: labyrinthine- defective (L-D) subjects; closed circles average values in minutes of arc; open circles: values for different trials at given body position).

126 420 360 300 240 IS0 IM 60 ° -60 -|20 o O -J60 -420 •480 -540 -600 -|80 -190 -120 • -60 -30 0 SO 60 90 I2O ISO 180 HEAD TILT (DEGREES) Fig. 2. Counterrolling (average values) as function of lateral head tilt. the head was positioned vertically downward. The curve in Fig. 2 repre- sents the average of several values obtained at each position for each head position. A considerable amount of variability among individual measurements far greater than the measuring error has been found in almost every subject tested, indicating that a certain amount of physio- logical unrest also exists with respect to the antero-posterior axis of the eye. This variability has been observed by several authors using various measuring techniques. By the use of counter rolling and known anatomical data a theory ("inward shearing") was developed in an attempt to explain the mechanism of otolith stimulation. It was proposed that each of the otolith organs (utricles and saccules) can be stimulated physiologically only by a shearing force applied inwardly, i.e., toward the median sagittal plane. This uni- directional response, furthermore, reaches its maximum when the direc- tion of the force (gravity) is parallel to the individual macular planes. In its active zone the response of each otolith, as indicated by counterrolling movements, appears to vary as a cosine square function of the angular displacement of the inward direction of the macula from the force of gravity. If the response were proportional to the gravity component in the inward direction of the macula, then the response would be proportional to the cosine of this displacement. But this simple relationship did not conform to the findings. The structure of the otolith organ apparently is such as to dampen the effect of gravity (G) in yielding the cosine square

127 function. A more complete discussion of these factors has been presented elsewhere (Miller, 1962). An important aspect of the general theory is the assumption that the saccules, like the utricles, act as gravireceptors and, when activated, contribute a smaller, yet significant effect upon counterrolling. There are several studies that would tend to support the supposition that the saccule functions as an equilibrial organ. Labyrinthine-defective subjects. The photographic technique is particularly useful in measuring smaller-than-normal amounts of counter- rolling that are found in individuals with disease or otherwise damaged vestibular organs. Investigators with less sensitive measuring devices had the difficult task of differentiating between a relatively large measur- ing error and a possible small residuum of otolith function. Information gained from precise measurements of otolith organ activity is needed to evaluate completely the inner ear organ triad and would complement the now routine audiometric and caloric irrigation tests of the other two auricular organs. If counterrolling, as has been assumed, is a specific indicator of otolith activity, then the level of function should be revealed in the character of the counterrolling response. In order to examine this theory, ten deaf subjects with bilateral loss of the semicircular canals were used as subjects for counterrolling measurements (Miller & Graybiel, 1962a). As can be seen in Fig. 1, these labyrinthine-defective (L-D) subjects did not disclose the characteristic pattern found in normal subjects in most instances. The magnitude of the response was in all cases less than in a comparable normal group. In some instances, there was no definite evidence of counterrolling; in others, it was limited to one direction of tilt; and in still others there was a small but regular dependence of counterroll with the successive increase in bodily tilt. The highly significant differences between the normal and L-D groups must have been due to loss of function of the auricular sensory organs. More specifically, since there is no evidence that the counterrolling re- flex is released by the organ of Corti, and insufficient evidence that it originates in the semicircular canals, but good evidence that it is released by the otoliths, it was concluded that the reduction in counterrolling in these cases was the result of injury to the otoliths. Interindividual dif- ferences in the L-D group are best explained by the presence of some residual otolith function. It was proposed that a single index—counter- rolling (CD—calculated as one-half of the difference between the greatest maximum roll associated with leftward and rightward tilt be used to de- scribe the functional status of the otolith organs for a given individual. Effect of change in gravitational inertial force environments. Counterrolling can also serve as an indicator of the effect of physiological deafferentation of the otolith organs brought about by eliminating or re- ducing the gravitational inertial force environment. The counterrolling response of six normal and six L-D subjects was measured at five dif- ferent tilt positions under zero G, 1/2 G, and standard G conditions. The average results of the normal and L-D groups are portrayed in Fig. 3. In the case of the normal subjects otolith activity as indicated

128 U <r (0 UJ I- Z 2 Z (9 o HE <r UJ O U NORMAL Jeo1 L-D N = 6 -50 -25 0 +25 HEAD (BODY) TILT IN DEGREES +5O Fig. 3. Counterrolling as function of magnitude of gravitational force (zero G, 1/2 G, 1 G) and head (body) position with respect to direction of force in normal and labyrinthine defective (L-D) subjects. by counterrolling response decreased in a regular fashion as the force was reduced. In the weightless condition tilting the normal individual would appear to have little effect upon the output of the otolith organs under the conditions of the test. The L-D subjects manifested a greatly reduced but similar pattern to the normals. This could be accounted for either as a residuum of otolith function in certain of the L-D subjects, or as an effect of stimulation to extra-labyrinthine source(s) of tonic in- nervation to the extraocular muscles. The former explanation seems more reasonable based on the results of the oculogravic illusion test and the care exercised to eliminate cervical, fixational, or binocular sources of cyclorotational eye movement. Even if these factors were involved, their importance is not great, as evidenced by the small maximum amount of counterrolling in the L-D group; thus, it seems more reasonable to assume that otolith function has been revealed. It is interesting to note

129 that, in the normal group of subjects, the 1/2 G curve falls somewhat short of the midway points between zero G and 1 G as might be predicted. The significance of this finding is not known. U) ILJ UJ IT 0 UJ Q 3 J y 5 0.2 04 06 0.8 Ui' \l ' 14 LATERAL F0RCE (G) 18 ' ID Fig. 4. Degree of counterroll as function of magnitude of force acting laterally on body (head). Information concerning the effect upon otolith activity (counter- rolling) of increased gravitational-inertial force is provided by another study (Woellner & Graybiel, 1959). As shown in Fig. 4, the counterroll } varies as a function of the magnitude of the lateral force, even beyond the standard level of 1 G. This demonstrates the important fact that the counterrolling response is normally stimulus bound and its limit is not reached even with a lateral force of 2.25 G. The function at this higher level, however, is no longer linear. Egocentric Visual Localization Aubert's phenomenon and its variants. Man's absolute localization of objects in space is made with respect to his egocentric frame of refer- ence. This frame of reference in turn is influenced by the.action and interaction of certain body mechanisms providing visual, vestibular.

130 tactual, proprioceptive, and other cues. When adequate visual cues are visible, they normally dominate all others so that judgments of in. principal axes of space are quite precise, stable, and easily rendered (Miller & Graybiel, 1963a). Removal of empirical visual cues, on the other hand, reveals considerable errors in perception, plus loss of stability and ease in localization, especially when the direction of the gravitational (inertial) force deviates appreciably from the longitudinal axis of the head (body). In the upright position, judgment of horizontally in normal individuals is not appreciably affected by the removal cues. When individuals are placed in a recumbent position, howeve -, and background cues are removed, suddenly they observe, after a brie, lag period, a gradual spontaneous rotation of the phenomenal horizon up to a maximum displacement typical for a given individual. Superim- posed upon these changes is the fluctuant movement of the apparent hori- zontal-rotary autokinesis (Miller &Graybiel, 1963a). The time cour of these illusions in four subjects is presented in Fig. 5. Fig. 5- Time course of perception of horizontality for each of four sub jects in upright (broken line) or recumbent (continuous line) posture with (unshaded strips) and without (shaded strips) visual reference cues. In addition to these relatively small fluctuations about the consider- able average deviation, the average level of deviation itself was found to vary from test period to test period as illustrated in Fig. 6 (heavy lines represent average of the several daily curves [thin lines] ). The variance in magnitude of deviation, however, does not appear to occur in a random manner but shows a regular dependence on the position of the head (body). It is, therefore, possible to describe a characteristic qualitative pattern of response, even though there are quantitative intertrial differences. In the three subjects tested at 10° intervals throughout the range of head tilts within ±90° from upright, the apparent visual horizontal tended to rotate in the same direction (E-phenomenon) as head (body) tilt in

131 -80 -60 -40 -20 0 +20 +4O +60 +80 HEAD (B0DY) TILT tN DEGREES Fig. 6. Apparent position of visual horizontal as function of head (body) tilt in degrees as tested in three subjects during each of several sessions. (Heavy lines represent average of individual session curves [thin lines] .)

132 moderate amounts from upright, reaching a peak bilaterally at approxi- mately 40° to 50° . Then it reversed direction, causing the magnitude of deviation to decrease until the subjective was coincident with the objec- tive horizontal. Tilting of the head beyond this point resulted in further displacement in the direction counter to head inclination (A-phenomenon). In all cases the deviation in the Aubert direction continued to increase on the average in proportion to head inclination beyond the "neutral" point, so that the greatest deviation occurred at the maximum positions of tilt (±90°) used in this study. Considerable information about these curious illusions has appeared in the literature since Aubert's original work, but quantitative data are scarce, especially in larger angles of head tilt. The underlying mecha- nisms are still not completely known. Counter rolling, a plausible expla- nation of the E-phenomenon involving the visual system only, has been cited by certain authors. If this be the sole or primary cause of this phenomenon, L-D subjects without this compensatory eye movement would be expected to manifest little or no E-phenomenon. A recent study (unpublished) provides quantitative proof that this theory is untenable. A group of L-D subjects compared to a similar group of normals revealed significantly more deviation in the E direction in certain moderate angles of tilt. An experiment (Miller & Graybiel, 1963b) was also conducted in which a group of subjects with known bilateral labyrinthine defects was compared with a group of normal persons with respect to ability to judge horizontality as a function of upright, recumbent, and inverted posture. The fact that the Aubert illusion and its variants have been reportedly observed by certain deaf subjects by no means proves that the labyrinths do not influence this perception. On the contrary, although similar quali- tative responses were found among all subjects, there were significant quantitative intergroup differences. When upright, the normals were able on the average to maintain their accuracy, while the L-D subjects deviated significantly in their settings to the apparent visual horizontal when empirical visual cues were removed- Both groups of subjects in the recumbent position perceived the Aubert illusion, but the magnitude of the illusion was considerably less in the normal group. When inverted, both groups were less accurate in their estimates in the dark, but no significant intergroup difference was found. In spite of the fact that there was some overlap in the group distributions of settings obtained in the upright and recumbent positions, indicating extra-labyrinthine factors were involved, the intergroup perceptual differences are best explained as an effect of the loss of vestibular function in the L-D subjects. It was concluded that the otolith organs in man act to increase his accuracy in egocentric visual localization, at least in the upright and recumbent positions. This conclusion is in alignment with an earlier finding (Miller & Graybiel, 1962b) in which a group of L-D subjects as compared to a group of normals perceived significantly greater amounts of autokinesis, another indicator of egocentric visual localization.

133 Oculogravic illusion—normal subjects. Man seated upright can in effect be tilted by generating, as in the human centrifuge, a centrifugal force which is vectorially added to the standard gravitational force; as a result, upright is perceived in the direction of the resultant gravitational- inertial force. Tilting the subject with respect to gravity differs in one important aspect from tilting the gravitational-inertial resultant force with respect to the subject. The magnitude of the resultant force is al- ways larger in the latter situation and bears a fixed but nonlinear relation to the angle phi. It has long been known that normal persons perceive the oculogravic illusion, and some of its characteristics have been systematically investi- gated (Graybiel, 1952). Recent evidence (Clark & Graybiel, 1962) would indicate that the oculogravic illusion, as was found for the Aubert illusion, is a function of a number of complex factors other than input from the otolith organs. For example, this illusion increases with a reduction in visual framework exposed prior to the rendering of a judgment of hori- zontality. In making this judgment, however, using the frame of reference there was no evidence of adaptation in subjects exposed to constant cen- tripetal force for four hours- Normal subjects have been found to judge the visual horizontal in a similar manner which is more or less in accord with the resultant force environment (Graybiel, 1952). The curve (Fig. 7) depicting the mean -70 -60 -50 -40 5 -30 HI !-20 CO Ld -10 SERIES I AND! COMBINED (9 NORMAL SUBJECTS) —=MEAN 10' 20• ANGLE <f> 30* 40° Fig. 7. Estimates of oculogravic illusion (OGI) by normal subjects. Single settings of luminous line .

134 values obtained by discrete settings to the apparent horizontal is typical of normal subjects. When continuous adjustments of the visual horizontal are made, a long delay is regularly observed between the time of peak acceleration and the time of the subject's peak response. This is perhaps due in part to a response lag of the peripheral sense organ, but is thought to be primarily due to delay in central nervous system mechanisms. A delay of similar character was seen in the static tilt studies. Oculogravic illusion—labyrinthine defective subjects. Failure to perceive the oculogravic illusion has been ascribed to loss of function of the otolith apparatus, but few studies have been carried out on L-D sub- jects, and few quantitative data are available to validate this claim. A study (Graybiel & Clark, 1962) was, therefore, conducted to determine the validity of the oculogravic illusion as a specific indicator of otolith function. A group of deaf subjects having complete functional loss of their semicircular canals but with unknown functional loss of the otolith organ was compared with a group of normal subjects in regard to the oculogravic illusion. In selecting naturally occurring experimental sub- jects with labyrinthine defects, the usual procedure is to screen a group of deaf persons, selecting those who also have lost the function of the semicircular canals. In doing this, it was not assumed that loss of all canal function was a valid indication of the complete loss of otolith func- tion. Indeed, evidence from counterrolling and the study just cited indi- cates that this assumption may be erroneous in some cases. The mean discrete settings to the apparent horizontal of the L-D subjects, in con- trast to the normal group, were not characteristic for all members of the group (Fig. 8). Indeed, the variability was so great that considera- tion of the results of individual subjects in this group was necessary. Differences among subjects in the L-D group are explicable on the assump- tion that in certain members there was a specific level of residual function of the otoliths, and in others it was lost completely. Unilateral labyrin- thine loss does not abolish the illusion (Graybiel & Niven, 1953). Semicircular Canals and Vision Nystagmus, induced by thermal stimulation or in response to angular or Coriolis acceleration, serves as an indicator of semicircular canal function, but its use in this regard is complicated by factors which may alter or abolish it. For example, it can be increased by mental activity. On the other hand, it can be reduced by introducing a visual fixation field, or by requiring a subject to repeat a particular pattern of vestibular stimulation. The reduction in nystagmic response through stimulus repetition may stem from a loss of arousal or drowsiness, but there is also evidence to show that nystagmus may be actively suppressed. Restriction of head movements to certain patterns diminished nystagmus during a 64-hour exposure to rotation at 5-4 RPM within a 15-foot diam- eter rotating room (Guedry & Graybiel, 1962). Efforts to maintain alert- ness by instructions did not restore nystagmus. Moreover, the nystagmus measured following the rotational period was opposite in direction to the

135 o n -70 -60 -50 -40 o -30 ui < (O UJ -20 -10 SERIES I AND H COMBINED (IOOFU SUBJECTS) — = MEAN l0' 20• ANGLE 30° 40° Fig. 8. Estimates of oculogravic illusion (OGI) by labyrin- thine-defective subjects. Single settings of luminous line. response produced by the same head movement during rotation, indicating that it was a conditioned response inasmuch as the stimulus was no longer Coriolis acceleration. Similar evidence of habituation was obtained from the subjective reports of the apparent motion (oculogyral illusion) of a target light in an otherwise dark room. For conditioning purposes, head movement was confined to one quadrant of the frontal plane. Subsequent tests were then made in this quadrant (practiced) and the opposite quad- rant (unpracticed). Dramatic reductions in response occurred in the practiced quadrant, but habituation was not transferred to the unpracticed quadrant (Fig. 9). Vestibular Organ Interaction and Vision It is obvious that, in many areas, research concerning the inter- action between vision and the vestibular organs has only begun. Differ- entiating the function of the otolith and cupula organs has occupied the interest of many investigators in the past, but the complete understanding of vestibular function seems dependent also on experimental programs which are directed at the physiological connections between the two types of vestibular organs and their combined influence upon vision.

136 ROTATION BEFORE BEGIN UNPRACTICED PRACTICED Fig. 9. Comparative magnitude and direction of Coriolis illusion associated with single head movements before, during, and after prolonged rotation at 5.4 RPM. Tests carried out at 7.5 RPM. Based on overwhelming evidence, there can be little doubt that the so-called vestibular nystagmus is released by the action of the semi- circular canals and, as in the examples cited above, is modified by the central nervous system. A basic question that now needs to be answered is whether under special circumstances the otolith organ may also either release it independently or contribute an essential element for its release by the canals . The fact that nystagmus can be elicited in certain indi- viduals by simply changing the position of the head is strong evidence that the otolith organ is involved in this ocular response. References Clark, B., & Graybiel, A. Visual perception of the horizontal during prolonged exposure to radial acceleration on a centrifuge. J. exp. Psychol.. 1962, 63, 294-301. ~~ Graybiel, A. Oculogravic illusion. Arch. Ophthal.. 1952, 48, 605-615. Graybiel, A., & Clark, B. Validity of the oculogravic illusion as a specific indicator of otolith function. Pensacola: USN Sch. av. Med. (BuMed) Rep., 1962, No. 67; & NASA Order No. R-37. (Proj. MR005.13-6001, Subtask 1)

137 Graybiel, A., & Niven, J. I. The absence of residual effects attributable to the otolith organs following unilateral labyrinthectomy in man. The laryngoscope, 1953, 63, 18-30. Guedry, F., & Graybiel, A. Compensatory nystagmus conditioned during adaptation to living in a rotating room. J. appl. Physiol., 1962, 17, 398-404. "~~ Miller, II., E. F. Counter rolling of the human eyes produced by head tilt with respect to gravity. Acta Otolaryng. Stockholm, 1962, 54, 479-501. ' Miller, II., E. F., & Graybiel, A. A comparison of ocular counterrolling movements between normal persons and deaf subjects with bilateral labyrinthine defects. Pensacola: USN Scn. av. Med. (BuMed) Rep., 1962, No. 68; & NASA Order No. R-47. (Proj. MR005.13-6001, Subtask 1) (a) Miller, II., E. F., & Graybiel, A. Comparison of autokinetic movement perceived by normal persons and deaf subjects with bilateral laby- rinthine defects. Aerospace Med., 1962, 33, 1077-1080. (b) Miller, II., E. F., & Graybiel, A. Rotary autokinesis and displacement of the visual horizontal associated with head (body) position. Aero- space Med., 1963, 34, 915-919. ~~~ Miller, II., E. F., & Graybiel, A. Role of the otolith organs in the perception of horizontality. Pensacola: USNSch. av. Med. (BuMed) Rep., 1963, No. 80; & NASA Order No. R-47. (Proj. MR005.13-6001, Subtask 1) (b) Woellner, R. C., & Graybiel, A. Counterrolling of the eyes and its dependence on the magnitude of gravitational or inertial force acting laterally on the body. J. appl. Physiol., 1959, 14, 632-634.

SOME RECENT ADVANCES IN INSTRUMENTATION AND PROCEDURES IN VISION RESEARCH Robert Boynton, Chairman

141 INTRODUCTORY REMARKS BY THE CHAIRMAN Robert M. Boynton Department of Psychology University of Rochester Progress in science depends to an important extent on the develop- ment of new experimental techniques. In the interdisciplinary subject of visual science, such developments sometimes come from the work of those directly concerned with vision, and sometimes from those whose interests are quite different. This fact is reflected in the four papers presented in this section. Two of the authors, Drs. Krauskopf and Cornsweet, are recognized as outstanding vision researchers whose many contributions are well known. The other two authors, Drs. Blough and Stark, are not primarily vision-oriented, but have worked intensively on visual problems since such problems were amenable to attack by their methods. The oldest of the techniques reported here, that of stabilized images, is only about ten years old. Since the first, nearly simultaneous, and independent reports of Riggs, Ratliff, Cornsweet, and Cornsweet and Ditchburn, in the early 1950's, many experiments have been devoted to this subject, whose importance seems to have been recognized instantly. The fact that the visual percept fades out completely under some condi- tions of retinal image stabilization serves to emphasize that the visual receptors tend to respond to change in illumination level, rather than to steady levels as such. Such effects are most easily observed with low contrast objects (or very small test objects) in peripheral vision, where the fading may be observed with careful fixation, even without the use of stabilizing optics (the so-called Troxler phenomenon). But with large, high-contrast, centrally-viewed stimuli, it is sometimes observed that fading does not occur, even with stabilizing optics, or that it is transitory, with vision returning unpredictably from time to time. Dr. Cornsweet believes such results are artifacts, usually associated with slippage be- tween the eye and the contact lens which supports a mirror that is part of the stabilizing optics used in many experiments. He reviews some of the various methods and comments particularly upon the problem of artifact, which is very important for any theoretical interpretation of the results. The other full-time vision expert, Dr. Krauskopf, reports on methods which had their start with the work of Campbell, Rushton,

142 Brindley, and Hagins in Cambridge, England, and Flamant in Paris in the middle 1950's. Images upon the fundus of the eye have, of course, been observed ophthalmoscopically for more than one hundred years but, until recently, these could not be easily measured due to the lack of sufficiently sensitive non-visual sensors. Photomultipliers have pro- vided the needed detectors The modern methods have been applied principally to two problems: (a) the bleaching of retinal photopigments, and (b) measurements of the quality of the retinal image. Dr. Krauskopf, from his expert knowledge, explains the difficult technical problems in- volved in making such important measurements. The use of animal subjects in sensory experiments is necessary because many physiological experiments can be performed on animals which are not possible on human subjects. Psychologists for many years have used animals in a variety of experiments in which visual discrimina- tions were involved The methods used have been generally very differ- ent from those one would use with human subjects, because of a feeling that it would be impossible to instruct animals to make the kinds of dis- criminations that human subjects make (and then only highly-trained ones) in the experiments of classical visual psychophysics. The develop- ment of animal psychophysics, which began with the work of Blough and Ratliff at Harvard in the middle 1950's, is predicated on the assumption that it is possible to do classical psychophysical experiments with animals, provided that one is clever enough to succeed in "telling" the animal what he is to do. The procedures that are proving effective are those of operant conditioning introduced originally by Professor B. F- Skinner. An important aspect of these procedures is that they are auto- mated: it takes a very long time to train an animal for some psycho- physical problems and the patience of the experimenter is much less strained if he can put the animal in the box, turn on the automatic con- trols, and read a book, or even leave the laboratory. It should be men- tioned that there is art as well as science involved in these training pro- cedures . Dr. Blough, in addition to being a first-rate scientist, is one of the real artists in this field- A few years ago it would have seemed ridiculous to predict that electrical engineers would some day come to have an intense interest in human sensory processes. Now this is a commonly accepted fact, and in a double sense. It is obvious that one's conceptualization of the visual system may be improved by applying progressively more sophisticated models borrowed from other areas, provided that the models really seem to fit, since much, if not most, theorizing about complex systems involves the use of models. It is perhaps less obvious that the study of biological systems may help to provide engineers with better solutions to their non- biological problems, but there is now a field of inquiry (bionics) dedicated to that proposition. To get work started in a new field often requires an unusual human being whose background and talents spread beyond the usual boundaries

143 set up by the various academic and applied disciplines. Dr. Stark is such a man—a physician turned electrical engineer who has been able to bring the biological and engineering approaches together. Although classical physiologists recognized long ago that there are many servo- systems and feedback loops in the nervous system, they did not use such terminology and they knew little about man-made servo devices, many of which are very recent. Meanwhile, the electrical control engineers have been developing some very sophisticated concepts, devices, and mathematical techniques for the analysis of servomechanisms. Dr. Stark describes how such methods may be applied to the study of such biological servosystems as the pupillary control mechanism, and the system which controls eye movements. It is to be expected that for many readers one or more of the four areas of methodology described here will be encountered for the first time, and it is to be hoped that this encounter will lead to further reading and to the development of new experiments and applications that might not otherwise have been forthcoming. For others, who may already have some knowledge about all four areas, up-to-date reviews by experts who are working with the methods should be useful.

144 VISUAL PSYCHOPHYSICS WITH ANIMALS Donald S. Blough Department of Psychology Brown University The past few years have seen renewed interest in animal psycho- physics. The impetus has come largely from the desire to obtain physiological and psychophysical data from the same or comparable species, rather than relying on lower animals for the one and human observers for the other. New behavioral techniques are making this convergence increasingly practicable. In many respects, the new methods are similar to classical tech- niques in which the animal chooses one of two stimuli by running or jumping to it, and is rewarded for "correct" responses. The new methods, however, are more efficient, and, in some cases, more sensitive. The animal stays in one place and responds rapidly to keys or levers; it re- mains unhandled for long periods and makes a high number of discrimina- tive responses per minute. Four illustrative techniques are cited here. All of them use pigeons as subjects, but, though the pigeon is a highly visual animal, there is reason to believe that the methods are applicable generally. Variations with monkeys, rats, and other species have been used successfully in visual and auditory work. Each procedure does two things: (a) it presents stimuli and records responses to these stimuli, and (b) it sets up and maintains the discriminative behavior through reinforcement techniques. To do these things some rather complex manipulations are necessary and they can be only roughly sketched here. (See references for more complete descriptions of some of the methods.) The first method is a tracking technique for determining the absolute detection threshold through time. The subject stays in a dark enclosure, and it responds to visual stimuli by operating two keys. The essentials of the situation as used with the pigeon are shown in Fig. 1. The bird is first trained to peck either key for food reward. Next, it is trained to peck Key A when the stimulus is visible, and Key B when the stimulus is off. If it performs this task correctly it will "track" its absolute thresh- old, for pecks on Key A automatically reduce the intensity of the stimulus, while pecks on Key B increase the intensity. The method is similar to, and derives from, Bekesy's technique (1947) for determining the human auditory threshold.

145 SHUTTER STIMULUS OPTICAL WEDGED—x RESPONSE/ KEYS/* Fig. 1 Central to the method is the procedure used to reward correct behavior. One cannot consistently reward with food all pecks on Key A given when the stimulus is visible and all pecks on B when the stimulus is not visible, since the threshold of visibility is not known in advance. Hence, a "chain" is set up. Pecks on Key A, instead of producing food reward, occasionally cause a shutter to close. (See Fig. 1.) Pecks on Key B are rewarded only when the shutter has been closed in this way. Thus, the pigeon works on Key A "to turn off the stimulus"; it works on Key B, when the stimulus is off, "to get food." These reward contin- gencies are intermittent; much of the time, the pigeon's responses only vary stimulus intensity, and during these intervals the threshold tracking data is collected. A complete description of this method may be found in Blough, 1958. Thresholds determined by this tracking me thod have yielded infor- mation on dark adaptation (Blough, 1956), spectral sensitivity (Blough, 1957a; Blough & Schrier, 1963), and drug effects (Blough, 1957b). The method has been applied to the determination of critical flicker frequency (CFF) in monkeys (Symmes, 1962), and in a number of auditory studies. A related technique has been applied to the determination of wave length difference thresholds. The pigeon subjects view a circular split field projected upon one of two response keys. The field is divided horizontally into half-circles of the same or slightly different wave lengths. If the fields are of the same wave length, the bird is rewarded for pecking a dark key next to the spot. The number of successive pecks required

146 for reward can be varied independently for each key, allowing the experi- menter to counteract position preferences. A correct choice of the dark key, when the wave lengths differ, causes the difference to increase 5 millimicrons (mji) on the next trial. The difference level to which the bird adjusts the stimulus provides an index of its difference limen. After considerable pretraining, birds were run with several basic wave lengths each day. Extensive data on one bird suggest that the shape of the wave length difference function over the range from 490 mji to 620 m^ is similar to the human function. The absolute level of the differ- ence limen—somewhat higher than the values usually cited for human subjects—has little meaning because it depends on experimental pa- rameters . An adjustment method has also been used to study brightness per- ception and contrast in the pigeon. Here the bird is "asked" to "report" which of two spots looks brighter. The spots appear on two keys side by side, and the bird is intermittently rewarded for pecking at the more intense spot. Pecking a given spot, however, causes the apparatus to dim that spot slightly and increase the intensity of the other. The trained bird switches from key to key, pecking at each spot in turn as it becomes the more intense. A strip-chart continuously records the bird's adjust- ments of the spots. Contrast effects were recorded in one study by putting a bright surround around one of the spots. In almost all cases, the birds pecked in such a manner as to increase the intensity of the spot on this bright field. This is, of course, the reaction of the human subject when asked to produce a brightness match in such a contrast situation. Carr and Guttman, at Duke University, have produced impressive data on CFF (unpublished), using a new variant of the two-choice pro- cedure. The pigeon faces two response keys. These keys are always identically illuminated at any given time. This stimulus illumination may vary from time to time with respect to intensity and flicker rate. If the stimulus (both keys) is flickering, the bird is rewarded for pecking the left key; if it is steady, the bird is rewarded for pecking the right key. Figure 2 shows the percentage of left-key responses typical of one bird, as a function of flicker rate. Carr and Guttman find a linear relation be- tween CFFs determined in this manner and intensity. They hope to use this relation to determine equal brightness functions for the pigeon, choosing a constant CFF as the criterion. A final new development is represented by the work of Herrnstein and van Sommers (1962). Using as their measure the response rate of an animal responding freely, they have obtained data suggestive of "inten- sity scaling" data from human experiments. The pigeon subjects were rewarded for pecking at different rates to each of several selected stimuli; their rates to other intensities, not specifically associated with reward, were recorded also. The results suggest a power law relation between pecking rate and stimulus intensity.

147 100 80 % RESPS. TO 60 LEFT KEY 40 20 550 m\i (Data of Carr 8 Guttman -unpublished ) \ 40 42 44 46 48 FLICKER RATE (CPS ) Fig. 2 The above summary indicates how new behavioral techniques may be used to attack with animal subjects some classical problems in vision. They show that the behavior of animal subjects can be closely controlled by stimuli —in some cases with a precision rivaling that achieved in ex- periments with human subjects. The new methods raise the efficiency of animal work by the use of intermittent reward, free responding with the opportunity for many "critical responses" per minute, and "feed-back" in the form of stimulus adjustment dependent on responding. The funda- mental training problem is the same as ever: to "tell" the animal what to discriminate. The solution to the training problem is to maximize reward for the desired discriminative behavior, and to minimize reward for unwanted behavior. References Bekesy, G. v. A new audiometer. Acta Oto-laryn., 1947, 35, 411; 422. Blough, D. S. Dark adaptation in the pigeon. J . comp. physiol. Psychol., 1956, 6, 49, 425-430. ** Blough, D. S- Spectral sensitivity in the pigeon. J. opt. Soc. Amer., 1957, 47, 827-833. (a) "

148 Blough, D. S- Effect of lysergic acid diethylamide on absolute visual threshold of the pigeon. Science. 1957, 126, 304-305. (b) Blough, D. S- A method of obtaining psychophysical thresholds from the pigeon. J . exp . anal. Behav., 1958, 1, 31-43. Blough, D. S., & Schrier, A. M. Scotopic spectral sensitivity in the rhesus monkey. Science, 1963, 139, 493-494. Herrnstein, R. J., & van Sommers, P. Method for sensory scaling with animals. Science. 1962, 135, 40-41. Symmes, D. Self-determination of critical flicker frequencies in monkeys. Science. 1962, 136, 714-715.

149 MEASUREMENTS OF LIGHT REFLECTED FROM THE RETINA John Krauskopf University of Maryland and Walter Reed Army Institute of Research When one looks into the eye using an ophthalmoscopic arrangement, reflections from the sclera, iris, cornea, lens surfaces, and retina may be observed. All of these reflections have been used to measure charac- teristics of the eye. This paper is concerned with three recent uses of light reflected from the retina: experiments on the spectral reflectivity of the retina; studies of photopigments in vivo; and measurements of retinal image formation. Retinal Reflectivity The first of these experiments may be considered as analogous to the test tube study of the absorption spectrum of pigments, but is greatly complicated by the fact that the experiments are performed on the living eye. In the test tube case, monochromatic beams, weak enough not to bleach the pigments, are passed through test solutions and solvent con- trols and allowed to fall on a phototube. Comparison of the phototube output for test and control samples at various wave lengths throughout the spectrum yields the absorption or density spectrum for the pigment. The spectrum thus measured is not necessarily related to visual function. Impurities may yield spurious absorption and the photopigment itself may absorb light in certain spectral regions with thermal but not photochemical effect and, therefore, has no visual effect. Alpern and Campbell (1962) were not directly interested in photo- chemistry. Rather, they were concerned with retinal reflection in the interpretation of the spectral sensitivities of the pupillary response. To this end, they compared the amount of light entering the eye at various wave lengths with that emerging after reflection through the pupil. These measurements were translated into retinal spectral reflectivities by cor- rection from the transmission properties of the ocular media. Both absolute and relative reflectivity data have been published. Interpretation of these experiments is difficult. It seems quite likely that the relative reflectivity of the retina may be measured with accuracy, but data to be presented later make it clear that it is quite difficult to measure the absolute reflectivity to a useful precision.

150 Measurement of Photopigments If there is interest in photopigments, it is obvious that retinal reflectivity per se will not produce the information required. A variety of impurities is present with an inhomogeneous mixture of photopigments. In the study of photopigments, which in the main has been conducted in England by Rushton and Weale and their collaborators, primary interest has centered on the difference spectrum. In the test tube case, the difference spectrum is measured by comparing the phototube output at various wave lengths before and after bleaching. If the solution contained but one photopigment, and the photo- products did not significantly absorb in the region of interest, the result might be simply interpreted as giving the effective absorption spectrum of the photopigment, but this is not usually the case Even in those cases where purified extracts are measured, an increase in absorption in cer- tain parts of the spectrum is found following bleaching. This means that the photoproducts do significantly absorb visible light, and the difference spectrum must be interpreted as yielding the absorption spectrum of the photopigment minus the absorption spectrum of the photoproducts. In the case of rhodopsin, the difference spectrum does agree rather well with the psychophysically measured scotopic spectral sensitivity curves over most of the spectrum, except in the short wave lengths where the photo- product absorbs. In extending the measurement of photopigments to the living eye, Rushton and Weale have followed somewhat different approaches. Weale (1959) has concentrated on getting data rapidly throughout the spectrum. Rushton's method, while a good deal slower in producing data, more closely fits the ideal of null measurement. This review is primarily concerned with Rushton's experiments, as they are more extensive and typify what can be achieved. A recent version of the apparatus used by Rushton (1958a) is shown in Fig. 1. The main beam of the apparatus is a double monochromator. Slits in the spectrum plane Q. . .Q allow the selection of different mono- chromatic test bands. On the assumption that little bleaching takes place in the far red end of the spectrum, a long wave length reference light is selected by use of the uppermost slit which is covered by a polaroid P^ oriented in one direction. Test lights may be selected by uncovering one of the other slits which are covered by a second polaroid ?2 oriented at right angles to Pj . Polaroid Po is rotated so that light alternately is passed by P^ and P« • W is a neutral wedge which is used to adjust the output of the photocell P.C. to a steady level when the two lights alter- nately reflected from the retina are equal. Sj and 82 are bleaching sources which may be introduced into the main beam. Suitable stops are present to eliminate stray light from the iris, sclera, and non-focal parts of the retina. The corneal reflex is eliminated in this arrangement by being deflected to one side, since the input beam enters one side of

151 Fig. 1. Rushton's reflection densitometer. the pupil and the recorded beam is taken out of the other side of the pupil. A suitable stop insures that the deflected beam does not enter the photomultiplier. An example of measurements of bleaching and regeneration of rhodopsin by Campbell and Rushton (1955) is given in Fig. 2. The ordi- nate is in centimeters (cm) of wedge displacement, which may be con- verted to pigment density if the wedge calibration is known. The three limbs of the bleaching curve (open circles) represent the effects of lights of different intensities in the ratios indicated. The time course of bleaching and regeneration (solid circles) are in accord with expectations from psychophysics. A good deal of other supporting evidence justifies the conclusion that a visual pigment was being measured. These results were obtained by recording with the light reflected from the peripheral retina. When attacking the cone pigments, a much more difficult situation is encountered. In order to escape contamination by rhodopsin, meas- urements must be made in the small, rod free, area of the fovea. Since efforts to extract cone pigments have been very disappointing, it is reasonable to assume that little cone pigment is present. Furthermore, even in the testtube, the situation becomes much more difficult if more than one photopigment is present in the solution. Rushton (1958a) there- fore, began the study of cone pigments with color blind (protanopic and

152 10 Time (min) 15 Fig. 2. Bleaching and regeneration curves for human peripheral retina. deuteranopic) subjects. The protanope is of particular interest, since it is generally agreed that he has but one pigment operating in the middle to long wave length portion of the spectrum. Figure 3 shows a regeneration curve for a protanopic subject (Rushton, 1958a). In general form it is like the rhodopsin curve, but, as expected, the time course is speeded up. Difference spectra recorded with two protanopic subjects are illustrated in Fig. 4. The line and circles are luminosity data. Considering the difficulties of the experi- ments, the difference spectra data (the boxes) show remarkably good agreement with the luminosity data. A puzzling feature of these data is the absence of any evidence of a blue receptor. Three possible explana- tions of this lack have been offered. One involves the notion of foveal blue blindness. Perhaps there are no blue receptors in the fovea. Secondly, the yellow macular pigment may absorb the blue light so strongly that too little remains to reveal the blue pigment. Finally, in agreement with the low contribution to luminosity of short wave length light, there may be little blue pigment present. In any case, to date no evidence of a blue pigment has been obtained. The deuteranope is a less satisfactory subject since more doubt exists with regard to the number and nature of the pigments in his eye.

153 DouHe 0-J6 014 oiz 0 10 cos 006 0-0 f 00) Fig. in. Dork. Tltae,nvUi 01 23 + 5-6719 10 3. Bleaching and regeneration in protanopic fovea. Spectlunv 5oo HY* 600 Fig. 4. Difference spectra for protanopic subjects. Willmer (1955) presents data suggesting that two classes of deuteranopes may exist: one missing a green receptor, and the other a fusion type in accordance with the Fick-Leber hypothesis. Rushton reported some data on one deuteranope which indicated two pigments, but recent psycho- physical data suggest that this is true for few if any deuteranopes (Speelman & Krauskopf, 1963).

154 The normal subject definitely presents the problem of measuring a mixed sample of photopigments. In test tube work, one can test a solution for homogeneity by bleaching with lights of different spectral composition. If but one photopigment is present the difference spectrum will be invari- ant under different bleaching lights. If it is not it may be possible to re- duce to insignificant concentrations all but one class of photopigments, and to demonstrate this by further homogeneity tests, which at the same time will yield the difference spectrum of the residual component (Dartnell, 1957). Results obtained by the method of partial bleaching on the normal subject are shown in Fig. 5. First, the retina was bleached with a long wave length light which had been shown to have no effect on the protanope. This produced the right hand difference spectrum. This was followed by further bleaching with white light which produced the second difference spectrum. Dewbk Density OoubU Deniitv -i 0-03 - OO2 - 001 600 650 Fig. 5- Difference spectra for deuteranope by method of partial bleaching. Almost all psychophysical spectral sensitivity curves are based on an equal response criterion. Thus, it is advantageous to use the same approach in collecting data which are to be compared with psychophysical data. For example, it is now common practice in electroretinogram work to find the stimulus required to evoke a criterion electrical output rather than to measure the output for equal energy stimuli. The photo- chemical equivalent is the action spectrum. In this case, a suitable measuring light is chosen, one for which the difference in density before and after bleaching is large. The bleaching effectiveness of various spectral lights is then assessed by finding that level of illumination re- quired to produce some criterion change in density for the test light. In the test tube case, colored impurities will distort these measurements; where the impurities absorb strongly they rob light from the photopig- ments and make it appear that more light is required to bleach. Again,

155 the case in which two or more photopigments are present greatly compli- cates matters. The partial bleaching technique can be used, but artifacts may be introduced if the photoproducts absorb significantly in the specral region of interest, since they would act in the same manner as other im- purities. Action spectra were measured by Rushton (1958b) by determining the bleaching effect of lights which were judged equal in brightness by his subjects. The test light was always the same, while the spectral compo- sition of the bleaching lights was varied. The expectation that lights which look equal in brightness should be the same in bleaching effect was sub- stantially confirmed in the case of rhodopsin and the cone pigment of the protanope. Fairly good agreement was also obtained between the differ- ence spectra and action spectra. A more complex analysis was involved in the determination of action spectra in the case of two pigments of the normal observer. In addition to the measurement of spectra, results have been pro- duced which are designed to provide evidence on the amount of pigments present, their photosensitivity, density, and apparent density, and the kinetics of light and dark reactions. In this interpretation of the data, Rushton has assumed that the measuring light passes through the photopigments twice, and that the density changes recorded are, therefore, double the changes in the photo- pigment density. For this to be correct it would be necessary for all the light to pass through the pigments. But it does not seem likely that such a perfect arrangement exists. Lewis (1956) has presented data on the rat retina which suggests that some of the light passes between the re- ceptors. Another requirement for Rushton's interpretation is perfect elimination of stray light from other ocular surfaces. In all probability, the quantitative interpretation of bleaching and regeneration will need revision as more is learned about the nature of reflection by the retina. In this review of the experiments no attempt has been made to go into the precise shape of the spectra obtained. There is still a good deal of uncertainty in these matters. Nor have the differences between the findings of Weale and Rushton been discussed. There exists fairly good agreement in regard to the general spectral character of the pigments, but the two groups do disagree on details. Considering the complexities of the measurement problem and the low levels of light this is hardly surprising. Important work remains to be done on these problems. Re- fining the measurements of the two cone pigments already identified and seeking the evasive blue pigment are two of the more important outstand- ing problems.

156 Retinal Image Formation The light reflected from the retina may also be analyzed to provide information about the quality of the optical system. Although this is not a new idea, there has been a recent outbreak of work on this problem. The pioneer work was done by Flamant (1956) nearly ten years ago. She used a difficult but elegant technique which involved the photographing of the retina with an ophthalmoscopic arrangement while the subject viewed a vertical bar target. In more recent work, the sensitivity of the meas- urements has been greatly increased by the use of photomultiplier tubes allowing more detailed quantification. This report, therefore, will con- centrate on the later work. Work in this area has been stimulated by the recent burst of activity in optics in the application of sine-wave-response theory. An important event in this context was the appearance of Duffieux's book (1947) on the application of Fourier analysis to optical systems, but, as is often the case, this was preceded by important work of some of the luminaries of the 19th century, notably Michelson and Rayleigh. In using an ophthalmoscopic system to measure the light distribu- tion on the retina, one is faced with the problem that the retinal image cannot be examined directly. In animal eyes, as in DeMott's experi- ments (1959), it is possible to remove the back of the eye and measure the light distribution after a single passage through the optics of the eye. Figure 6 illustrates an apparatus which has been used to make the ophthal- moscopic type of measurements (Krauskopf, 1962). The subject views a bright vertical bar target, T, forming an image, T'. The light, diffusely reflected from the retina, passes out of the optics of the eye and is imaged by them back in the target plane, but part of the light is reflected by a beam splitter Ml so that an image, T ", of the retinal image, is formed in the plane of L4. This is the image that is to be measured, but in order to make provisions for controlling effective pupil size, a relay system L4 and L5 (which contributes negligible degradation) is included so that a last image T'" is formed. Another vertical slit, located in this plane, provides a window for a photomultiplier. When the photo- multiplier is moved horizontally, its output traces the light distribution in the image of the target slit, doubly degraded by the eye optics. A second set of images may be traced through the system, starting with the source filament and proceeding through Dl, the eye pupil plane, and D2. Since the eye is dilated by instillation of cyclogel, the effective pupillary aperature for the incoming beam is determined by the size of Dl and that for the outgoing beam by D2. Westheimer and Campbell (1962) used a similar but simpler and apparently more efficient system in which the pupil size was controlled by a conventional artificial pupil located near the eye. Yet another comparable arrangement has been used by Rohler (1962). Happily, this

157 Sin LIGHT PATH ELECTRICAL PATH LIGHT TRAP H^ Fig. 6. Photoelectric ophthalmoscope . is one of those situations in which the results of all investigators are in accord in all but the most minor details. Once the light distribution of the doubly degraded target has been measured, the problem is to deduce the quality of the image on the retina. One way to look at the problem is to consider the optics as con- sisting of two optical systems of identical character in tandem. On the hypothesis that rays travelling in opposite directions along the same path behave in the same way, this is a reasonable model. In this model, the image of the first system serves as the object of the second system. One can make this assumption since the light is diffused by the retina. A schematic representation of the light distributions produced on the retina and in the plane of the scanning slit is given in Fig. 7. The target, the uppermost graph, is assumed to be an infinitesimally narrow bright bar. On passing through the optical system once, the distribution becomes a roughly bell-shaped curve. This distribution is known as the "line spread function" of the system. If a point instead of a line target had been used, one would get a similar radial distribution known as the "point spread function. " On re-imaging the light diffusely reflected from the retina, the retinal image can be considered as composed of an infinite .number of infinitesimally narrow lines which vary in intensity in accordance with the line spread function. Each of these elements produces its own light distribution in the second image plane. All these distributions have the shape of the line spread function, but their heights vary according to

158 TARGET RETINAL IMAGE Fig. 7. Schematic representation of imaging of narrow bright line target in ophthalmoscope. their location in the first image. In all of this, it -is assumed that the light is incoherent, and, thus, the contributions of the elementary distri- butions can be summed to compute the shape of the resultant distribution of light in the second image plane. The problem in the ophthalmoscopic case is that the initial and final distributions are known and one seeks to deduce the intervening, retinal light distribution. This might be done in a variety of ways, but it turns out to be most convenient to apply the techniques of Fourier analysis. This procedure is not only simple but pays dividends in that it yields a sine-wave response curve, -characteristic of the optics, which allows one to deduce the light distribution of the retinal image for any target. The use of Fourier methods does not necessarily imply any special properties of the optics other than superposition as discussed in the preceding paragraph, but, in fact, the eye, like many other optical systems, may be considered as a low pass filter of spatial sinusoids. The use of filter theory in the case of electrical circuits is more familiar,

159 so it may be helpful to think in terms of electrical analogs. In the analog, the target may be replaced by a temporal voltage pulse, the eye in the ophthalmoscopic arrangement by two, equal, low pass filters in cascade. The retinal image is considered as analogous to the temporal voltage variation observed at the output of the first filter, and the second image as analogous to the voltage variation observed at the output of the second filter. 1 The input pulse can be decomposed into a series of sine waves of varying frequency and amplitude, and the same can be done with the two later wave forms. For the cascaded filter case, one finds that if the first filter reduces the amplitude of a particular component sinusoid, such that the output equals R times the input amplitude, then the output of two filters in cascade is equal to R2 times the input amplitude. In other words, the response function of the cascaded system is equal to the square of the response curves for the individual filters. Thus, if a Fourier anal- ysis of the output of the second filter is performed and it is divided by a Fourier-analysis of the input, the squared response curve of the compo- nent filters is obtained, and by taking the square root the component response curves are attained. The intermediate wave shape can be de- duced by determining how the input pulse is transformed by a filter of this sort. With this method the bonus of learning the response function is obtained. This function, which is mathematically equivalent to the line spread function, may be used to derive the shape of the outputs at either terminal for any input. In this form of analysis, fine lines and points may be considered as high-frequency disturbances, broad areas are rich in low frequency. In order to resolve small objects good high-frequency response is needed. An illustrative application of this approach to optics is to be found in automatic star trackers. The problem of detecting a star may be trans- lated to finding a high-frequency disturbance against a low-frequency background of noise (clouds and a light sky). By use of appropriate chopping reticles and electronic filters it has been possible to develop high-performance star trackers that "see" stars in the daylight sky. Figures 8-11 illustrate the application of this approach to the eye. Figure 8 shows the recorded output of the photomultiplier obtained with the apparatus described above. The abscissa is the horizontal scanning dimension, the height to the tops of each line gives the light level at that point. The target in this case is a bright vertical bar. By performing the analysis discussed previously, the response function shown in Fig. 9 is obtained. In this case, the abscissa is in terms of lines/minute (min) of visual angle and is logarithmic, while the ordinate is linear. The curves have been normalized by equating the B.C. levels of input and In the electrical analog voltages varying in time are measured; in the optical case light distributions varying in space are of interest. Thus, an electrical pulse is the analog of a narrow bright bar, a series of square waves the analog of a grid target.

160 i iiiiiiiiiiiiiiiiiiifirin :: Fig. 8. Sample record from photoelectric ophthalmoscope. I.0 -8 LU 2 o Q. LJ c o: .6 LU LJ .2 .0I 0.I 1.0 SPATIAL FREQUENCY-LINES/MINUTE Fig. 9. Response of eye to various spatial frequencies. Three experiments. Pupil diameter—5 mm.

161 output. The results of three experiments using a 5 millimeter (mm) pupil are shown. The scattered points in the right-hand side of the figure may be disregarded as artifacts, due principally to the fact that the high frequencies are not well represented in the target. Figure 10 shows response functions for various pupils from 3 to 8 mm in diameter. .01 0.1 ID SPATIAL FREQUENCY - LWES/MNUTE OF ARC Fig. 10 Spatial frequency response curves for various pupil diameters. Pupil diameters —3, 4, 5,6, 7, and 8 mm from bottom to top. Ordinates dis- placed . They were determined and plotted in the same way as the previous one. For clarity the data points are eliminated and the ordinates displaced. Pupil size varies in steps of 1 mm downward from 8 mm to 3 mm. These curves reveal the variation of performance of the eye with aperature, showing that imagery is best with moderate sized pupils (4-5 mm) and deteriorates considerably with larger pupils. Figure 11 shows the light distributions on the retina. Again, the 8 mm pupil is at the top, the 3 mm pupil at the bottom. The relative deterioration with large pupils can be seen here also.

162 20 15 10 5 0 5 10 IS VISUAL ANGLE - MINUTES 0F ARC 20 Fig. 11. Reconstructed retinal images of 1.6' bright vertical-line target. Pupil diameters 3, 4, 5, 6, 7, and 8 mm from bottom to top. Ordmates displaced. If it is assumed that the optics of the eye are radially symmetrical these results could be used to compute the image of any object. (If the assumption proves invalid, the calculations could still be made if the point spread function were known.) The process of determining the image distribution is known as convoluting the object function by the spread func- tion. The application of this procedure is illustrated in Fig. 12 taken

163 Fig. 12. Example of convolution. Target is knife edge. Bell-shaped curves indi- cate line spread function of optical system. Smooth curve is resultant light distribution in image obtained by summing of weighted spread func- tion for each linear element of target. from Perrin (1960). Although the mathematical calculations are carried out in terms of the summation of sinusoids, the process, as illustrated here, is precisely equivalent to summing the spread functions appro- priately weighted by the light distribution in the object, in this case a knife edge. Comparison of this figure with Fig. 7 may be useful. The performance of the eye may be compared with theoretical ex- pectation. A good model to use is the diffraction-limited lens, which is

164 assumed to be free of chromatic and spherical abberration and other defects. This is a condition which is often closely approached in good optics. Figure 13, taken from an Eastman Kodak publication (1962), shows the general response function for such a lens. The ordinate is the LIMIT Of ICSOiunON FOK A UN€ IMAGES. H) 20 30 40 SO 60 70 80 RELATIVE SPATIAL FREQUENCY (X of Limiting Value) 90 100 Fig. 13. Theoretical response functions for aberration-free, diffraction-limited lenses. Limit of resolution for par- ticular lens is determined by its aperature and wave length of light employed. normalized response. The abscissa is in terms of line frequency. It is also normalized in terms of the cut-off frequency. The curves are modern representations of diffraction theory and, with a little mathe- matical manipulation, can be related to cases which are generally more familiar, such as the Airy diffraction pattern. The cut-off frequency is determined by the wave length of the light used and the lens aperature. In angular terms, it is given by kd/X, where d is the diameter of the aperature, X is the wave length, and k is a constant to adjust for the units used. Some typical computed values for the eye are about 1 line/min with a 2 mm pupil, 2 lines/min with a 4 mm pupil, and 4 lines/min with an 8 mm pupil, when X = 550 millimicron. These limits are never closely approached experimentally. Of the probable causes for the failure to achieve maximal perform- ance the principal ones appear to be chromatic and spherical aberration. Experiments have been recently performed using monochromatic light. The results at each wave length are essentially identical to those obtained with white light. This does not mean that there is no chromatic aberra- tion. In fact, the existence of chromatic aberration was verified in the experiments by the variation of the power of the auxiliary spectacle lenses which had to be placed before the subjects' drugged eye to achieve best imagery. The results, however, agree with the data which show that acuity is little effected by the spectral composition of the light when luminance is controlled. In the case of acuity, the spectral sensitivity

165 of the eye determines that the light derived from the center of the visual spectrum predominates. In the ophthalmoscopic studies, the spectral variation in the light source, eye media, retinal reflectivity, and photo - multiplier response combine to produce similar selectivity. Thus, the white light measurements are not truly white, but, rather, give a good approximation of the effective performance of the eye in white light. If the eye exhibited simple spherical aberration it might be pos- sible to take corrective measures and achieve better than normal acuity. The older measurements of Ivanoff (1953) together with new detailed measurements by van den Brink (1962) and Smirnov (1961), demonstrate that the aberration is not simply spherical. While there is an overall radial variation, there is little radial symmetry. Nevertheless, experi- ments with zonal or annular pupils were attempted. The decision to do this was abetted by the suggestion in the literature that better resolution of certain classes of targets, specifically points and grids, might be achieved through the use of such pupils (O'Neill, 1956). This idea has a considerable history in the literature of optics. Apparently, the first to use it was the great English astronomer Herschel, who placed a central occluder in the plane of the objective of his telescope (Rayleigh, n.d.). The development of the modern sine-wave treatment of optics has allowed the detailed solution of such problems and a number of papers have ap- peared in the last ten years or so. In Fig. 4, taken from O'Neill, are plotted a family of normalized response curves for aberration-free lenses with varying degrees of central occlusion. It can be seen that with a perfect lens a relative improvement in high-frequency performance is predicted with central occlusion. It is also apparent that a lens exhibiting simple spherical aberration ought to be improved in performance by using Fig. 14. Theoretical response functions for aberration- free lenses with various degrees of central occlu- sion. Parameter is relative diameter of occluded region.

166 annular pupils. Thus, double benefits might well be expected. There is no need to dwell on this for the experiments failed to support any of the expectations. A variety of annular pupils were tried with uniform results: performance is always poorer than that achieved with circular pupils of the same outer diameter. Herschel's observations are the only data in which improvement with the eye is reported. There is little information available about the precise conditions involved. Perhaps the improve- ment he obtained involved the optics outside rather than inside the eye. Something of interest did show up in these experiments. In the course of performing the Fourier analysis of the output light distribution, one gets, as a matter of course, a measure of the total light in the image. It was observed that the amount of light in the images produced with annular pupils was less, in comparison with a circular pupil of equal outer diameter, than would be expected on considerations of area. That is, there seems to be an effect similar to the Stiles-Crawford effect to the utilization of light from different parts of the pupil (Stiles & Crawford, 1933). To investigate this further, the optics were rearranged so that the photomultiplier, supplied now with a small circular aperature, could be used to scan through an image of the pupil. Referring to Fig. 6, aperature Dl was opened up to fill the dilated pupil completely, the slit T was removed providing a circular target about 2-1/2 degrees in diam- eter, and the photomultiplier was moved to the plane of D2. Figure 15 shows the results of a horizontal pass through the pupil image. The horizontal axis is position, the vertical, light reflected from the eye pupil. It can be seen that the corneal reflex makes it impossible to measure throughout the whole of the diameter. This can be reduced, however. Figure 16 shows the effect of using crossed polaroids, one in the path before the eye, and the other in the return path. Comparison of these figures makes it clear that this procedure effectively removes the specular reflection from the cornea without altering the shape of the distribution of the light due to the pupil. In Fig. 16, one can identify the reflection due to the iris and discern the margin of the pupil. The ordi- nate is linear. Comparing the height at the margin of the pupil to the maximum, the ratio is about 1/3 to 1/2. It should be pointed out that this is considerably less than the ratio obtained by Stiles and Crawford. This record was obtained from the subject's right eye. The temporal margin is to the left. A similarly skewed distribution was obtained for the right eye of another subject. The similarity of these results with the psychophysical Stiles- Crawford effect is suggestive, and experiments comparing these results with psychophysical data on the same subjects are planned. In seeking an explanation of the finding, the idea of a light-trapping effect is attrac- tive. This might be reinforced by studies of subjects with "tilted" retinas. Experiments have been attempted with monochromatic light, but to date they are not sufficiently good to justify strong conclusions. It is reason- ably sure, however, that there are no large variations in the shapes of the distributions with wave length.

167 Fig. 15. Oscilloscope tracing of distribution in plane of pupil of light reflected by retina. No polaroids in beam. The doubts expressed earlier about the quantitative interpretation of the Alpern and Campbell experiments should now be clearer. The light reflected from the retina is directional; that which is scattered as stray light is relatively less than would appear by examining the light which gets out of the pupil. Furthermore, the obviously complicated situation makes correct quantitative interpretation of the photopigment seem difficult. It may prove useful to employ Fourier analysis to the processing of spatial patterns by the nervous system. An example of this approach

168 Fig. 16. Oscilloscope tracing of distribution in plane of pupil of light reflected from retina. Crossed polaroids in input and output beams. is the interpretation by Lowry & De Palma (1961) of their photometric studies of Mach bands. Applying this approach to the nervous system raises tricky problems of psychophysics and mathematics, particularly because of the non-linearities involved . Nevertheless, interesting things may be learned, particularly with .regard to the problems of summation and inhibition.

169 References Alpern, M., & Campbell, F. W. The spectral sensitivity of the con- sensual light reflex. J. Physiol., 1962, 164, 478-507. van den Brink, G. Measurements of the geometrical aberrations of the eye. Vis. Res., 1962, 2, 233-244. Campbell, F. W., & Rushton, W. A. H- Measurement of scotopic pig- ments in the living human eye. J. Physiol., 1955, 130, 131-147. Dartnall, H. J. A. The visual pigments. London: Methuen, 1957. DeMott, D. W. Direct measures of the retinal image. J . opt. Soc . Amer., 1959, 49, 571-579. Duffieux, P.M. L'Integrale de Fourier et ses application a 1'optique. Rennes: Societe Anonyme des Imprimerie Oberthur, 1947. Eastman Kodak Co. Modulation transfer data for Kodak films . Rochester (N.Y.): Kodak Pamph., 1962, No. P-49 . Flamant, Francoise. Etude de la repartition de luminere dans 1'image retinienne. Probs. contemp. Opt. (Istituto Nazionale di Ottica, Arcetri-Firenze)1956, 561-569. Ivanoff, A. Les aberrations de 1'oeil. Rev, d'optique theoretique et instrumentale (Paris), 1953. Krauskopf, J. Light distribution in human retinal images. J. opt. Soc. Amer., 1962, 52, 1046-1050. Lewis, D. M. Rat retinal photo-pigments by reflected and transmitted light. J. Physiol., 1956, 133, 55-56. Lowry, E. M., & DePalma, J. J. Sine-wave response of the visual system. I. the Mach phenomenon. J. opt. Soc. Amer., 1961, 51, 740-746. O'Neill, E. L. Transfer function for an annular aperature. J . opt. Soc. Amer., 1956, 46, 285-288. Perrin, F. H. Methods of appraising photographic systems. J . Soc. Motion Picture & Television Engrs, 1960, 69, 151-156; 239-249. Rayleigh, J. W. S. Scientific papers. Cambridge: University Press, n.d. Vol III, p~W. ~~

170 Rohler, R. Die Abbildungseigenschafter des Augenmedien. Vis. Res., 1962, 2, 391-429. Rushton, W. A. H. Human cone pigments. In Visual problems in colour. London: H.M.S.O- 1958, Pp. 71-105. (b) Rushton, W. A. H. Kinetics of cone pigments measured objectively in the living human fovea. Ann. N-Y. Acad. Sci., 1958, 74, 291- 304. (a) Smirnov, M.S. Measurement of wave aberration in the human eye. Biophysics, 1961, 6, 52-65. Speelman, R. G., & Krauskopf, J. Effects of chromatic adaptation on normal and dichromatic red-green brightness matches. J . opt. Soc . Amer., 1963, 53, 1103-1107. Stiles, W. S., & Crawford, B. H. The luminous efficiency of rays entering the eye pupil at different points. Proc. roy. Soc., 1933, B112, 428-450. Weale, R. A. Photosensitive reactions in the fovaea of normal and cone- monochromatic observers. Optica Acta, 1959, 6, 158-174. Westheimer, G., & Campbell, F. W. Light distribution in the image formed by the living human eye - J . opt. Soc . Amer. , 1962, 52, 1040-1045. '~" ' " Willmer, E. N. A physiological basis for human colour vision in the central fovea. Doc. Ophthalmol., 1955, 9, 235-313.

171 STABILIZED IMAGE TECHNIQUES Tom N. Cornsweet Department of Psychology University of California There is a large amount of evidence, both physiological and psycho- physical, to suggest that the human retina transmits information about changes.in retinal illumination very well, but transmits steady-state in- formation poorly, if at all. But so far there is almost no understanding of the processes by which steady-state information is lost, and this lack of understanding is due in part to the technical difficulties involved in producing a true steady state of illumination on the retinal elements. Fixate the dot numbered 1 in Fig. 1 as steadily as possible, occluding one eye. The blurred disk rapidly disappears but the sharp one remains Fig. 1. Demonstration of disappearance of steadily fixated object.

172 visible. If, after the blurred disk has disappeared, fixation is shifted to point 2 the disk will reappear and subsequently disappear again. The solid line in Fig. 2 is a plot of the retinal illuminance at the edge of the retinal image of the sharp disk in Fig. 1. The vertical c o jj d "5 rt rt Distance Fig. 2. Distributions of retinal illuminance at edges of retinal images of disks in Fig. 1 . Solid line repre- sents sharp disk, and dashed curve blurred disk. Rectangles on horizontal axis represent positions of receptor. rectangles on the horizontal axis represent a receptor. During "steady" fixation, involuntary eye movements cause the receptor to "shift" in relation to the image over a distance labeled 10' (10 minutes of arc). The actual displacement will depend on the exposure duration (Riggs, Armington, & Ratliff, 1954). Such "shifts" of the receptor cause the illuminance on it to change from level "a" to "b" and back repeatedly during fixation. The dashed curve in Fig. 2 represents the retinal illuminance for the image of the edge of the blurred disk. Note that the same involuntary eye movements produce much smaller changes in retinal illuminance on receptors near the edge of the blurred disk (from "c" to "d"), and the disk disappears. When fixation is shifted to the second fixation point, the receptors under the image of the blurred disk undergo relatively large changes in illumination (the new fixation point is represented in Fig. 2 by "X"'), and it reappears. Change in the illu- mination on receptors seems to be a necessary condition for seeing patterns. Visibility depends on the size of the shifts of the retinal image relative to the steepnesses of the gradients of retinal illuminance.

173 Retinal Blood Vessel Patterns Light arriving at the receptors in the human retina passes first through a layer of tissue containing blood vessels, and the vessels cast shadows upon the receptors. If the eye is dark-adapted, and then a white field is suddenly turned on, a trained observer can see this pattern of shadows for a brief time, but the pattern rapidly disappears and the field looks homogeneous. The shadows cannot move with respect to the retina, and there are thus no corresponding changes in retinal illuminance after the first change when the field is turned on. However, if the shadows are made to move with respect to the retina they appear again, but dis- appear as soon as they stop moving. Movement of the shadows maybe produced by changing the direction of the light incident upon the retina. For example, Campbell and Robson (1961) describe a device in which the spot on the face of an oscilloscope is imaged in the plane of the observer's pupil, so that he sees a uniform field in Maxwellian view. When the spot is moved, the blood vessel shadows move proportionately. The blood vessel shadows can be experimentally manipulated to some extent. They can be moved as described above. Their contrast can be changed by changing the wave length composition of the incident light. The shadows are darkest relative to their background when the light is monochromatic at 415 millimicrons (mji), where there is a very strong absorption band for blood (Glasser, 1950). Different mixtures of light at 415 rmj and light at another wave length poorly absorbed by blood yield different amounts of contrast. Similarly, it should be possible to provide a visual field in which the only thing that is changing is the stimu- lation of receptors behind blood vessels. This may be accomplished by showing the observer a field that is alternately lighted with 415 mjj and then with a mixture of two other wave lengths each of which is absorbed less strongly by blood, but so chosen that their mixture will match the 415 mjx light in regions not lying behind vessels. A retinal image that does not move with respect to the retina can also be provided by a device that might be called an auto-ophthalmoscope (Cornsweet, 1962). This optical device forms an image of the observer's peripheral retina on the fovea of the same eye with a magnification of xl and so arranged that it does not move with respect to the retina regard- less of movements of the eye. The observer looking into the device sees at first a sharp view of his own peripheral retina (with its blood vessels, optic disk, etc.), but the detail rapidly fades out and soon the field looks homogeneous, the detail reappearing only if the image is deliberately moved across the retina, for example, by jarring the apparatus. The salient aspect of the perception of each of the patterns discussed thus far is that detail rapidly fades out and disappears, the observer being left viewing a field that is apparently homogeneous with respect to the patterns that have been made motionless. Any superimposed patterns that do move across the retina, such as a fixation point or dust in the

174 optics, remain visible. The detail never reappears unless it is deliber- ately moved across the retina or its contrast is deliberately changed. In other words, only changes in illumination result in seeing. Other Stabilized Images Several procedures have been developed which allow the eye to move normally and provide a retinal image that moves along with the eye, so that the image remains relatively stationary with respect to the recep- tors. Changes in illuminance are thus reduced. These procedures may be divided into two types. In the first, a mirror is attached to the eye and the image to be viewed is reflected from the mirror in such a way that the retinal image moves with the eye (Riggs, Ratliff, Cornsweet, & Cornsweet, 1953; Ditchburn & Ginsborg, 1952; Yarbus, 1956; Clowes & Ditchburn, 1959). In the second type, the object to be viewed is itself attached to the eye so that its image moves with the eye (Yarbus, 1957; Pritchard, Heron, & Hebb, 1960). The attachment to the eye can be either by a tightly fitting contact lens or by a small cylindrical chamber attached by suction (Yarbus, 1957; Barlow, 1963). Retinal images rendered motionless with respect to the retina in this way are called stopped or stabilized images. Every report of images thus stabilized states that the stabilized aspects of the images disappear, and, in this way, these images are apparently identical with the blood vessel shadows and blurred images discussed above. However, when images are stabilized using attachments to the eye, it is also almost always reported that the images reappear from time to time. Whether this reappearance is the result of some artifactual movement of the retinal image, or is actually produced by an autonomous process in the visual nervous system, is a subject of some debate, and perhaps even of consequence (Hebb, 1963). It is obvious that the extent of stabilization of an image depends on the extent to which the attachments to the eye follow eye movements. Any slippage of a contact lens, or any change in the shape of the eyeball which results in motion of the retina with respect to the lens, will produce some displacement of the image with respect to the retina. It is equally obvious that any attachment to the eye must slip a little. The important question is whether or not the slippage that does occur is big enough to account for reappearance. Barlow (1963) has recently reported his measurements of the slip- page of two types of attachments to the eye, using a very sensitive meas- uring procedure. According to that report, what he calls a tightly fitting contact lens, fitted with a very strong lens and a stalk holding a target at a position where the lens and the optics of the eye image the target on the retina, slips about 31/2 minutes (min) of arc between the beginning and the end of a 5 to 10 second (sec) period of time. This is the kind of lens used by Hebb et al (1963). Barlow also measured slippage for the

175 cylindrical attachment held on by suction, such as first described by Yarbus (1956). According to Barlow, this device slipped only 40 sec of arc between the beginning and end of a 5 to 10 sec period. Barlow did not test the type of lens used by Riggs and many of his co-workers. Their lenses fit extremely tightly, and have only a very small mirror attached to them. Riggs, using a procedure similar to Barlow's, reports that his lenses slip about 15 to 30 sec of arc during 60 sec viewing periods and as a result of voluntary saccadic movements of a few degrees. * Consequences of Slippage The solid curve in Fig. 3a is the distribution of retinal illuminance produced when a human with emmetropic vision views a bar 10 min of arc wide through a 6 millimeter (mm) pupil. The dashed curve in Fig. 3a is the distribution for the same bar shifted sideways through 1 min of arc. The solid curve in Fig. 3b is a plot of the difference between the solid and dashed curves in 3a. In other words, it is a plot of the change in illuminance produced when a bar 10 min of arc wide is shifted 1 min of arc. The dashed line in 3b is a plot of the distribution of illuminance pro- duced by a bar 1 min of arc wide and having the same luminance as the bar in Fig. 3a. Shifting a bar 10 min of arc wide 1 min of arc sideways produces an increase in illuminance that is almost identical with the in- crease produced by presenting a bar 1 min of arc wide of the same lumi- nance. A bar 10 min of arc wide also produces a decrease in illuminance, so that the total change in illuminance is greater than that produced by the introduction of a bar 1 min of arc wide. Shifting a bar 10 min of arc wide 10 sec of arc sideways produces an increase in illuminance almost identical to the increase produced by its initial presentation. In general, a small displacement of a wide bar causes an increase in illuminance that is very nearly the same as that produced by presenting a bar whose luminance equals that of the wide bar and whose width equals the extent of displacement. This is true for dark bars on a light ground as well as light bars on a dark one, except that the word "decrease" must be substituted for "increase." The threshold for resolution of a dark line is about 1/2 sec of arc under optimal conditions (Hall & Mintz, 1939). A bright or a dark line 10 sec of arc wide is easily seen unless it has very low contrast. Ten sec of arc is just 3 per cent of the median size of involuntary saccadic eye movements, and only one-half of 1 per cent of the 30 min of arc saccadic movements that occur occasionally during fixation. Further, if the eyeball changed its shape (because of muscular pulls or pulse pressure, for example) in such a way that the fovea was displaced 750 m/i with respect to the optic axis, the target would shift 10 sec of arc. Therefore, it is quite likely that reappearance will occur during viewing ^Riggs, L. A., & Shick, A. Personal communication.

176 • o I ore shift 10* wide bar Retinal distance I e e • QJ O> u — l0'bar moved l' l' bar of luminance equal to shifted 10' bar Fig. 3. (a) Distribution of retinal illuminance for bar 10 min of arc wide seen through 6 mm pupil. Dashed curve is distribution when bar has been shifted 1 min of arc with respect to solid curve. Shaded area is change in illuminance produced by shift. (b) Solid curve represents change in illuminance shown as shaded in (a). Dotted curve is distri- bution of retinal illuminance produced by stationary bar 1 min of arc wide and having same luminance as 10 min wide bar. of an image stabilized with an attachment to the eye. (The appearance of the restored image may not be identical with that of the original image, e.g., if the field is a set of lines in haphazard directions, any slip would maximize the likelihood of the reappearance of lines perpendicular to the plane of slippage, the resulting perceptions being "simpler" than the original ones.) Figure 4 shows distributions of retinal illuminance for bars of different sizes. As the width of the bar decreases, the maximum slope of the illuminance distribution decreases, and this trend continues for bars smaller than those in the figure. As the slope of the distribution

177 I — O4 I. 364 (014 68 Retinal Distance From Center of Bar (Min of Arc) Fig. 4. Light distribution in retinal image of long bars for human eye, best focus, 6 mm pupil. decreases, the distance that the entire distribution must be shifted to achieve some fixed amount of change in illuminance increases. There- fore, it is to be expected that, using a stabilization system that is short of perfect, very fine lines might never reappear while wider ones (or ones with greater contrast) would reappear from time to time as a con- sequence of slippage, and such findings are reported in the literature (Riggs e_t aK, 1953; Ditchburn & Ginsborg, 1952). In general, the pro- portion of time during which a bar is visible during fixation would in- crease as the width of the bar increases, even if artifactual retinal image movement were the only cause of reappearance. To summarize, "spontaneous" reappearance of stabilized patterns has been reported only when the stabilization was produced by a system using an attachment to the eye. It is never reported for patterns of retinal illuminance which are known to be perfectly fixed to the retina. Slippage does occur in contact lenses and in devices attached to the eye by suction. While this slippage can be made very small, it is not neg- ligible in relation to the reappearance of many visual targets. A Device for Perfect Stabilization To study the processes that cause an image to disappear, and the other side of the coin—processes that render an image visible, it would be highly desirable to have a device which produced patterns of retinal illuminance which are perfectly stabilized and are manipulatable. The only perfectly stabilized patterns currently available, the retinal blood vessel shadows and related entoptic phenomena (Maxwell's spot, Haidinger's brushes, etc.), can be manipulated only within very restricted limits. But it is theoretically possible to produce perfect stabilization of any desired image. When an observer looks at a target, an image of his own retina is formed on the target. If it were possible to know, at each instant, the exact position of some landmark or set of landmarks in the image of the retina with respect to the plane of the target, and if the target itself were

178 then moved so as to be in a fixed position relative to the image of the retina, then the image of the target would be perfectly stabilized on the retina (Cornsweet, 1958). In other words, if a star were moved con- tinuously so that it was always located in the exact center of the image of the observer's fovea, the retinal image of the star would always be exactly at the center of the observer's fovea, and would thus be stabilized. So long as the relationship between the star and the image of the retina were fixed, the image would remain stabilized regardless of movements of the eye or the head, and regardless of distortions of the shape of the eyeball or changes in the position of the retina with respect to the optic axis. A first, and hardest, stage in building such a device is to construct a system which will give an electrical output which signals the instanta- neous position of the image of the retina, that is, a system to track the image of the retina. The output of such a tracker could then control the position of a visual target, for example a pattern displayed on the face of a cathode-ray tube. In order to track the retina, light must be sent into the eye, and the light reflected back out must be processed to extract the information it contains about the position of the retina. Since light is quantal in nature, it is necessarily true that perfect tracking cannot be achieved at all times. For example, there are times when no quanta arrive at the sensing sys- tem. Such a small proportion of the light incident on the retina is actually reflected out of a human eye that quantal considerations become important. That is, a very large amount of light must be put into the eye in order that enough quanta come back out to yield precise information about the loca- tion of the retina. For example, suppose that a point source were imaged on the edge of a blood vessel in the human retina. If the light were re- stricted to a wave length of 415 ± 2 1/2 mp (for optimal contrast between blood vessel and background), the source must deliver approximately 90 x 1012 quanta per sec per mm2 to the pupil of the eye in order that the light reflected back out of the eye contain enough information to locate the horizontal position of the blood vessel within 1 sec of arc every 10 milli- seconds (ms). That is roughly the intensity of a 500 watt high pressure mercury arc lamp. The intensity required is directly proportional to the precision in time, e.g., ten times as many quanta are required for a precision of 1 sec of arc every millisecond, and proportional to the square of the spatial precision, e.g., 10~2 times as many quanta are required for a precision of 10 sec of arc. The derivation of the function relating incident intensity to precision of tracking is included as an appendix to this report. In other words, if the light sensing and tracking machinery itself were perfect, that is, if the apparatus lost no information, tracking of this kind could be made just about precise enough to provide good image stabilization. This solution is marginal. With real apparatus, it is pos- sible that the intensities required would damage the retina. 2The writer is indebted to Mr. Michael Davidson, Department of Psychol- ogy, University of California, for the derivation.

179 The figures discussed above are calculated for an apparatus that images a single point source on the retina. If two sources were imaged and the sensing system acted upon both of them, the intensity of each could be reduced by one-half. Thus, it seems that the most promising approach to tracking of the human retina involves gathering light from a large region of the retina rather than from the image of a single point. If a section of the retina, say, for instance, the entire optic disk, were scanned at a very high rate, the intensity at each point could be reduced to easily achieved values, while excellent tracking could still be accom- plished . Appendix Michael Davidson Department of Psychology University of California Calculation of number of quanta required per unit time for eye -movement tracker. 1 . Incident light, nominal Let the line-spread function for the optical system be a Gaussian function with variance a2, and let a point source be such that Qs quanta per second (sec) are imaged onto the retina. Then the density distribution of incident light is qn(x,y) = [QS/21T02] • exp [-(x2 + y2)/2a2] (1) the units being those of Qs/<72, e.g., quanta • sec" • (minutes of arc) . Note that the peak density is given by 2. Incident light, stochastic Assumption: every photon which traverses a particular path from the source of the eye strikes the retina at precisely the same point. Then the probability distribution of the density of incident light at the point (x,y) on the retina is given by a Poisson distribution whose mean is the nominal intensity at the point, viz. P lq(x,y) = u} = [qn(x,y)]u • [u!]'1 • exp [ -qn(x, y)], (3) where qn(x,y) is the nominal intensity, given by equation (1). (Note: strictly speaking, one should use no. of quanta within a small time 6t, and a small area 6x6y, but this would lead to the same result.) Since the

180 emission from the source at a particular angle is independent of that at other angles, the distribution at each point is independent of that at each other point, under the assumption stated. 3. Reflected light, local As long as the number of quanta reflected during the time interval of interest remains large, the probability distribution of reflected light is also Poisson, with mean given by rn(x,y) = R(x,y) • qn(x,y). (4) R(x, y) being the reflectance of point (x,y). Now, assume that the retina is divided into two regions: R(x,y) = (5) R2, x>c. Then the density of reflected light, r(x,y), is given by P lr(x,y) = v] = [R1qn(x,y)J1' [i/! J"1 • exp [ -Rj qn(x,y)]. if (6) = [R2qn(x,y)]" [f!]"1 • exp [-R2qn(x,y)j, if x > c . 4. Reflected light, total Since the sum of two independent Poisson-distributed random variables is also Poisson-distributed, with mean equal to the sum of the separate means, the light reaching the detection device has a Poisson distribution with a mean given by the integral of the mean reflected density over the spot. Denoting the total number of quanta per sec by T, and the mean value of the distribution of T by Tn, one obtains Tn(c) =J J [QSRj ^ffa2] • exp [-(x2 +y2)/2a2] dxdy + j"j_1LQ8R2/2ffa2] ' exp[-(x2 +y2)/2a2]dxdy (7) = QS LR1*(c/a) +R2 (1 - *(c/a) )], where the error integral *(x) is given by the usual relation *(x) = J_^[2ir]"1/2 exp [-t2/2]dt. (8)

181 Then, since T is Poisson, P (T(c) = kj = [Tn(c)]k • [k! I"1 exp [-Tn(c>] . (9) 5 • Case of line close to center; normal approximation Letting C = 06, where |6| «1; then the Taylor approximation may be used ' *'(c/<r) |= = ±- + 6A/2ff (10) Further, if it is assumed that Qs will be quite large, the Poisson distri- bution may be approximated with the appropriate normal distribution: P tT(c) = k] «[2TrTn(c)]"1/2 exp [-(k-Tn(c) )2/2 'Tn(c)] . (11) Substituting (10) into (7) gives Tn(a6) = QS [(R1 +R2)/2 + (Rj - R2) • 6A/2ff] (12) as the common value of the mean and variance of the distribution (11), for the case where the edge is displaced by an amount 6 a from the incident distribution. One further approximation is made: since 6<K 1, the var- iance of the family of distributions under consideration may be approxi- mated well by Var = Qs [(Rj + R2)/2], whence, from (11), P {T(c) = k] = [ffQs (R1 +R2)]" exp [-(k-Tn(c) ) IQ^R^ + R2>]- (13) 6. The estimation problem Assume that the task is to follow the eye movements with a tracking machine so that the line is within a specified distance e of the center of the distribution a fraction (1-*) of the time. Let the machine be con- structed so that it works as follows . a. The time scale is broken into small intervals of T sec. b. During each period of T sec, the machine counts the quanta incident upon it. Let the number be qQ. c . The machine finds the value of 6 corresponding to qQ, by solving the equation:

182 qo = TQS [(R1 + R2)/2 + (Rj -R2) • 6A/2ir], i.e., 6 = (2ff)1/2(R1 -R2)"1 Lq0/TQs - (Rt +R2)/2]. (14) d. The machine moves the spot by an amount 6 a (to the right or to the left according as 6 is positive or negative) . Thus, this machine will move, at the end of a time interval, to the esti- mated average location of the line during that interval. This movement will be correct (1-«) of the time, to within error ±€, provided that the estimated value of 6 a is correct within a confidence interval of radius e, confidence level (1-01). This means that qo is the center of a confidence interval for the number of incoming quanta during the period T, the center and radius of which, from (14), are found as follows: • ( (Rj +R2)/2 + (Rj - Rg) [ (6a - e)/c] A/2? } ; q0±Aq = TQS • [ (Rj +R2)/2 +(Rj - R2) • 6A/2ff] ± TQS C (Rj - R2)/aV2ff. (15) But now this radius is also given by the distribution (13) to be: (l-°O/2 = J ^ Normal [o, TQS (Rj + R2)/2] dx, (16) i.e ., that value of x for which a normal distribution with variance TQS(R1 +R2)/2 has (1-«) of the area between -x and x. 7. Calculation rule a. Choose confidence level (1-«), allowable spatial error €, time sampling interval T. b. Find value of Z^ such that a normal distribution with mean 0 and variance 1 has area (1-01) between (-Z^) and Z.^. c. Find: (a) standard deviation a of the line-spread function, (b) reflectance R^ of lighter surface (this = probability that a quantum incident on the retina will enter the measuring instrument), (c) reflectance R2 of darker surface (similar).

183 d. Then the number of quanta incident in the entire spot per unit time is given by the solution of TQS(R1 -R2)c/or,/2¥=Z<x[TQs(R1 + R2)/2]1/2 which is ire2 z£ (Rj + R2) References Barlow, H. Slippage of contact lenses and other artifacts in relation to fading and regeneration of supposedly stable retinal images. Quart. J. exp. Psychol., 1963, XV, 36-52. Campbell, F. W., & Robson, J. G. A fresh approach to stabilized retinal images. J . Physiol., 1961, 158, 11P. Clowes, M. B., & Ditchburn, R. W. An improved apparatus for produc- ing a stabilized retinal image. Opt. Acta, 1959, 6, 252-265. Cornsweet, T. N- New techniques for the measurement of small eye movements. J. opt. Soc. Amer., 1958, 48, 808-811. Cornsweet, T. N- A stabilized image requiring no attachments to the eye. Amer. J. Psychol.. 1952, 75, 653-656. Ditchburn, R. W.. & Ginsborg, B. L. Vision with a stabilized retinal image. Nature, 1952, 170, 36-37. Glasser, O.(Ed.) Medical physics (Vol. n). Chicago: Yearbook Publishers, Inc., 1950. Pp. 1071-1072. Hebb, D. O. The semiautonomous process: its nature and nurture. Amer. Psychol., 1963, 18, 16-28. Hecht, S., & Mintz, E. U. The visibility of single lines at various illu- minations and the retinal basis of visual resolution. J. gen. Physiol., 1939, 22, 593-612. Pritchard, R. M., Heron, W., & Hebb, D. O. Visual perception approached by the method of stabilized images. Canad. J. Psychol., 1960, 14, 67-77.

184 Riggs, L. A., Armington, J. C., & Ratliff, F. Motions of the retinal image during fixation. J. opt. Soc. Amer., 1954., 44, 315-321. Riggs, L. A., Ratliff, F., Cornsweet, J. C., & Cornsweet, T. N. The disappearance of steadily fixated test objects. J. opt. Soc. Amer., 1953, 43, 495-501. " Yarbus, A. L. Perception of an immobile retinal image. Biofizika, 1956, 1, 435-437. Yarbus, A. L. A new method of studying the activity of various parts of the retina. Biofizika, 1957, 2, 165-167.

185 PRINCIPLES OF NEUROLOGICAL FEEDBACK CONTROL SYSTEMS FOR EYE MUSCLES Lawrence Stark Neurology Section Electronic Systems Laboratory and Biology Department Massachusetts Institute of Technology A review of recent work on neurological control systems indicates that certain system-design properties seem to be relatively widespread, although individual systems may employ only one or two examples of such mechanisms. Stability of system performance may be provided by either adequate gain margin, or adequate phase margin, or both. How- ever, in some cases, oscillations occur which may have a relationship to function. Non-linear properties such as scale compression or other types of saturation may permit the system simultaneously to maintain stability in one domain and oscillatory behavior in another. (Noise with various bandwidths and amplitude ranges may be present with or without a functional role .) Often, systems are composed of a symmetrical component whose interaction probably improves system performance at minimum cost. Even error signals are sometimes employed. Discontinuous control computations result in sample data properties with attendant character- istic features. These are often accompanied by prediction operators to enable closer following of repetitive signals. Interaction between differ- ent systems and time-varying characteristics, such as adaptation, extend the repertoire of design principles found in these biological feedback con- trol mechanisms. In order to study and analyze biological servomechanisms graphic displays such as Bode, Nyquist, and Phase plane diagrams are supple- mented by analytic descriptions such as linear transfer functions, non- linear describing functions, and higher order functional analysis. Simulation by analog models, special digital computer programs which simulate analog computers, and most recently, hybrid models have been employed. The hybrid models use analog elements for dynamics and time delays. Nonlinearities are obtained from the digital computer. An online digital computer is now being used to contain models and adjust parameters in real time during the course of an experiment as identifica- tion and matching becomes possible.

186 A number of different approaches have enabled dissection into the black box defined by the above input-output experimental analytic tech- nique. These approaches include the use of dissected invertebrate preparations; stereotactic, electrophysiological experiments on cats; and the use of pharmacological agents on normal subjects. Neurological patient material has been extremely valuable in providing subjects with altered control systems. Utilization of classical physiological literature has sometimes proven to be of great value for modelling, and conversely, the black-box analyses have often suggested crucial physiological experi- ments to be done. Chronic experiments in conditioned animals with im- planted electrodes are now being planned since it is deemed most impor- tant to identify the physical behavior of the physiological elements which together comprise any control system. The linear and nonlinear properties of the pupil servomechanism were reviewed in 1959 (Stark). Since then material on the pulse response of the pupil (Stark, Van der Tweel, & Redhead, 1962), rapid dark adapta- tion measured by means of a null pupil-response technique (Stark, 1962a), and environmental clamping of the pupil (Stack, 1962b) has been published . Current work on pupillary noise, nonlinear modeling with analog-digital hybrid computers, stereotactic electrode studies, higher order kernel approximation, and drug experiments has as yet appeared only in progress report form. The human accommodation system is a most interesting example of a complex biological control system. A nonlinear servoanalytic treatment of experimental steady state data (Stark, Takahashi, & Zames, in press) and a special examination of the evidence for the absence of an odd error signal mechanism (Stark & Takahashi, in press) have been reported recently. The eye target tracking system has been studied using electronic means to measure horizontal eye movements. It is important to eliminate the effect of the "prediction operator" by using unpredictable signals (Stark, Young, & Vossius, 1962; Stark & Young, in press—a). Then the discrete or sampled-data nature of the position (saccadic) and velocity (pursuit) dual control system operating together (Stark & Young, 1962; Stark & Young, in press —b) becomes apparent. A model formulated in engineering terms shows remarkable agreement with the real behavior of the system under variable feedback experimental operating conditions. Studying interactions between various loops of a single control sys- tem and between different control systems operating upon the same final effector is being continued. The classification of many previously unex- plained phenomena, and the prediction of novel phenomena as, for example, a high-gain oscillation, strongly justifies the effort required to apply the essentially mathematical concepts of servoanalysis to the motor control systems of the intra-ocular and extra-ocular muscles.

187 References Stark, L. Stability, oscillations and noise in the human pupil servo- mechanism. Proc. Inst. Radio Engrs, 1959, 47, 1925-39. Stark, L- Biological rhythms, noise and asymmetry in the pupil-retinal control system. Ann. N. Y. Acad. Sci., 1962, 98, 1096-1108. (a) Stark, L. Environmental clamping of biological systems. The pupil servomechanism. J. opt. Soc. Amer.. 1962, 52, 925-30. (b) Stark, L., & Takahashi, Y. Absence of an odd error signal mechanism in human accommodation. J. opt. Soc. Amer., in press. Stark, L., Takahashi, Y., & Zames, G. Nonlinear servoanalysis of human lens accommodation. J. opt. Soc. Amer.. in press. Stark, L., Van der Tweel, H., & Redhead, J. Pulse response of the pupil. ActaPhys. Pharm. Neerlandica, 1962, 11, 235-239. Stark, L., & Young, L- Dependence of accuracy of eye movements on prediction. J. opt. Soc. Amer., in press. (a) Stark, L.. & Young, L. A discrete model for eye tracking movements. Reglungitechnik. 1962 (Jan.). (Also: Bionics Conf., March, 1963, Dayton; Inst. Elec. & Electronic Engrs, Trans. Prof. Tech. Group on Military Electronics, in press.) Stark, L., & Young, L. R. Variable feedback experiments supporting a discrete model for eye tracking movements. Inst. Radio Engrs, (Trans.) Human factors —spec, manual control issue, in press, (b) Stark, L,., Young, L., & Vossius, G- Predictive eye movement control. Inst. Radio Engrs, (Trans.) Human factors in electronics, 1962, No. HFE-3, 52-57.

THE EFFECTS OF DRUGS ON VISION

191 INTRODUCTORY REMARKS BY THE CHAIRMAN Arthur Jampolsky Eye Research Institute Presbyterian Medical Center It is the purpose of this symposium to review the direction and results of some previous investigations, and to point out the probable directions for best future potential in solving some of the practical prob- lems, and acquiring more fundamental knowledge in this field. There is a large body of literature pertaining to drug effects on visual perception, behavior, and performance. There is a paucity of information on the specific sites and modes of action of drugs on the visual apparatus. It seems appropriate to examine the tools available to the psychophysicist, the electrophysiologist, and the biochemist to see what information may be derived from different approaches. The format has been organized to bring out the sites and modes of drug action on the visual apparatus; to assess the sensory input in electro- physiological terms; to discuss the drug effects on the extraocular muscles and on the accommodative-convergence mechanism; to relate these find- ings to psychophysical parameters, and to the more general topics of performance and behavior; and to point out the potentials, limitations, and blind spots requiring future examination. In the sixteen years since lysergic acid diethylamide (LSD) entered the scene, there have been thousands of descriptive reports vividly de- picting such visual experiences as the pristine beauty of the pearl-covered mountains described by a variety of laymen and scientists. An attempt will now be made to examine the pearls more critically, and to determine from the point of view of visual scientists how on earth they reached this unlikely perch.

192 OCULAR PHARMACODYNAMICS Albert M. Potts Department of Ophthalmology University of Chicago Although the desirability of obtaining drugs that influence the visual process favorably is perfectly evident, most of the actual effects here reported deal with deleterious influences of drugs. Investigation in this area is, however, by no means an exclusively rear-guard action, for the knowledge that is gained of the functioning of the visual system under the influence of deleterious drugs leads to a knowledge of normal function and how to modify it in both directions. The action of drugs on the eye is specifically conditioned by a num- ber of factors which are unique to the eye. One of these has to do with the way ocular anatomy and physiology determine the access of drugs to the intraocular structures where their actions are exerted. The ioniz- ability of a drug, the pH and tonicity of the solution carrying it determine how much will penetrate the multiple-layered cornea and reach the interior of the eye. To get beyond the anterior chamber, the drug must resist the sluggish but real current of aqueous humor which tends to wash it out once more. Furthermore, the location of the end-organ, whether in the ante- rior chamber as with the iris, or in the posterior chamber as with the ciliary body, will determine what effective concentration of locally applied drug will reach its destination. Systemically administered medications have different but equally stringent conditions laid upon them. They must either diffuse into the eye from the vessels of choroid, iris, and ciliary body, or they must be ac- tively secreted with the aqueous humor, or they must diffuse from the limbal vessels into the avascular cornea. Which of these occurs will largely be determined by the chemical nature of the substance in question, and one may find selective secretion into the eye of substances like ascor- bic acid, or selective exclusion from the eye as in the case of urea, or even extrusion from the eye as in the case of certain iodinated contrast media. The breathtakingly rapid advances in biochemistry have been re- flected in changing concepts of mechanisms of drug action. Now, when the actual three-dimensional structure of protein molecules, as in the case of myoglobin, can actually be diagrammed, the concept of molecular site of drug action can rapidly pass the stage of theorizing. The work of Wilson on the nature of the site for anticholinesterase activity, and the

193 work of Belleau on the nature of the adrenergic receptor site deal with the bonding of functional groups of the sort that occur in known protein struc- tures, and the day in which stoichiometric pharmacology reaches its ascendency may be very near, indeed. With pharmacology at the molecu- lar level in this way, the unique advantage that the iris offered as a test object behind a transparent window is no longer as all-absorbing as it once was. However, the new work in molecular pharmacology will con- tribute to the knowledge of how ocular autonomic structures operate, but a number of problems in the area still remain to be solved. This is par- ticularly true in the area of the control of intraocular pressure. Despite the possibility of dual control of rate of aqueous formation and rate of aqueous outflow, it is not completely understandable why a parasympa- thetic stimulator such as pilocarpine can lower the intraocular pressure, and a sympathetic stimulator such as epinephrine can accomplish the same effect. This is all the more curious when certain adrenergic inhibitors, given locally, can once again cause decrease of intraocular pressure. With the increasingly rapid rate of appearance of new drugs on the scene, the advent of some with unique actions on the eye is apparently increasing in frequency. Explaining these unique actions is a fascinating occupation. Witness one of the studies in progress in the author's own laboratory. Some years ago it was found that certain phenothiazine tranquilizers cause a pigmented choroidopathy in humans. Subsequent investigations revealed the curious phenomenon that phenothiazines and other polycyclic compounds store in the pigment of the uveal tract in higher concentration than anywhere else in the body. This storage, which also occurs with synthetic melanin, may well be attributable to the charge transfer reaction which is only now being studied intensively by biologists. It has been recently shown in the laboratories of the National Institutes of Health that the antimalarial drugs which exhibit a different type of ocular toxicity are also stored in the uveal tract. This is cited simply as an unsuspected and new aspect of the ocular toxicity of chemical compounds, and how knowledge of dynamics in this case may lead to prevention of damage from future similar compounds, and may also prove an effective way of directing desirable substances into the eye. Thus, preoccupation with the pharmacodynamics of harmful com- pounds is largely a measure of relative ignorance. As this ignorance is dispelled by increasing knowledge in the area, toxicity can be avoided and beneficial effects can be controlled.

194 THE SENSORY EFFECTS OF DRUGS: ELECTROPHYSIOLOGICAL INVESTIGATIONS OF THE MECHANISM OF THE ACTION OF DRUGS ON THE EYE Geoffrey B. Arden Department of Physiology University of California Medical Center* San Francisco Since only about six of the many thousands of references in each volume of ophthalmic literature deal with the electrophysiological investi- gation of the action of drugs, this review is in one sense easy to write. However, the fact that so little work is being done makes it impossible to cover the pharmacology of the retina in any comprehensive fashion. There- fore, this report discusses the methods available to the electrophysiolo- gist, and how they may be used to investigate the mechanism of drug action. Illustrations mainly from the author's own work are used since several of the most prominent workers in this field have recently given their own reviews of their experiments (Noell, 1958; Potts, 1962). The great advantage of electrophysiological techniques is that one can use them to determine the precise level in the retina that is affected by a particular drug. Perhaps the most interesting observations concern the steady potential of the eye, which is produced not in the retina proper, but in the pigment eipthelium which lies behind it. Since the rods and cones have no blood supply, their metabolic requirements must be met by the pigment epithelium cells. In particular, it has been shown that the pigment epithelium is concerned in the synthesis of visual purple. Vita- min A, liberated by the photolysis of rhodopsin, actually travels out of the rods into pigment epithelium cells. There it is modified before being transported back into the receptors (Dowling & Gibbons, 1960). These observations form the background of work undertaken (Arden St. Fojas, 1962a) to elucidate the mode of action of diaminophenoxyalkanes. These compounds have a schistosomiacidal action, and were at first thought to be non-toxic. Indeed, it was only when the research chemists concerned took the drugs themselves, that a retinotoxic activity was disconvered. Sub- sequently, it was found that cats and frogs too could develop a diamino- phenoxyalkane retinopathy, and histological studies in cats showed that the 1 Current address: Institute of Ophthalmology, Judd Street, London W.C.1, England.

195 outer part of the retina is affected (Ashton, 1957). With this in mind, a method for the assay of the retinotoxicity was developed, based on the decrease in the concentration of visual purple in the frog retina after administration of the drug (Goodwin, 1957). Of course, any chemical which kills retinal receptors will stop the accumulation of visual purple, but it seemed likely that the assay method was doing more than causing the death of rods: perhaps the diaminophenoxyalkane was affecting the visual purple cycle, possibly via the pigment epithelium. 1000 I WO 1*00 1*00 I2OO 1000 too 200 "I O o O e . M1B96BA IVISng/kg O e o o 9«* *A^ ^a' Of A 0000000<»>000 IO JO 3O 4O SO 6O TO 8O Fig. 1. Action of phenoxyalkane derivative on rabbit eye. Note increase of D. C. eye potential, and c-wave, while b-wave unaffected. (Arden & Fojas, 1962a) Figure 1 shows the effect of an intravenous dose of a modified phenoxyalkane in a rabbit. The steady potential of the eye increases very greatly, and at the same time, the c-wave amplitude also increases. This action is exactly the same as that caused by small doses of sodium azide, another drug which selectively damages the pigment epithelium (Noell, 1953). It will be noted, however, that the b-wave of the electroretinogram (ERG) is unaffected by the injection. This is not surprising because the rabbit does not suffer from a retinopathy after injection of phenoxyalkanes

196 (there is a marked species specificity), and the amount of the injection is smaller than would cause a retinopathy, even in a susceptible cat. It seems, therefore, that this class of drugs interferes with the pigment epithelium in smaller doses than are required to produce a retinopathy. This deduction has recently been confirmed. It has been shown that the P 32 uptake of pigment epithelium is abolished by diaminophenoxyalkane, but the uptake of retina is not affected (Glocklin & Potts, 1962). Of course, there was considerable interest in discovering what functional disturbance, if any, resulted from sub-toxic doses of diamino- phenoxyalkane. The author and his associates found it possible to give cats a small dose of a phenoxyalkane derivative which produced no ophthalmoscopic or long-term electrophysiological abnormality. How- ever, the drug had an effect on the retina, as was discovered when the ERG dark adaptation curve was measured. This is shown in Fig. 2. 1-3 05 4O Fig. 2. ERG dark adaptation curves before (•) and after (x) administration of phenoxy- alkane derivative. Drug did not affect waveform or sensitivity of dark-adapted ERG. (Arden & Fojas, 1962a) Each point represents the intensity of light required to elicit a b-wave of constant amplitude at various times after the end of moderate light adap- tation. After administration of the drug, the cat's eye dark-adapts more slowly than normal. The final dark-adapted threshold of the retina is unaffected by the drug. Also, the early part of dark adaptation is not really altered. This early part is predominantly "neural, " while the later part of dark adaptation may, in part, reflect the accumulation of rhodopsin.

197 Interpretation of this result was, therefore, that the drug had affected the rate of synthesis of visual purple, but the retinal neurones were unaffected, and it was speculated that the diaminophenoxyalkanes might be specific poisons for some part of the rhodopsin cycle. That suggestion will remain a speculation until the photochemists and biochemists attack the problem, but it is reasonable. It is interesting that many compounds which are relatively inactive in other sites in the body specifically affect the pigment epithelium. For example, the pheno- thiazines are specifically concentrated in the uveal tract, and although this has not been proven to be the mechanism of the retinotoxic effect, it seems likely. If one is looking for a difference between pigment epithelium metabolism and the metabolism of other structures in the body, the visual purple cycle, with its unique use of vitamin A, springs instantly to mind. Of all the ill effects produced by phenothiazine retinopathy, the impair- ment of dark adaptation is most striking (Burian & Fletcher, 1958). The clinical effects of the drugs mentioned to date are negligible. However, one occasionally retinotoxic compound, chloroquin, is in common use, and clinical accounts of chloroquin retinopathy are becoming more and more common. So far, the disease has not been reproduced in any animal, and experimental pathology is not available. In man, the symp- toms and signs of the retinopathy are variable. It is striking that the onset of the condition is quite sudden and occurs after many years of drug administration, during which there have been no obvious ill effects. Even though the drug is then withdrawn, the retinopathy may progress, so that though there may be some slight recovery from the acute phase, progres- sive deterioration of visual function follows. This is paralleled by a slow appearance of pigment in the fund us, It is possible to investigate pigment epithelium function in man. The steady potential of the eye cannot be measured directly, but, since it gives rise to the eye movement potential, a method based on the electro- oculogram can be devised to test the functional capability of the pigment epithelium (Arden, Barrada, & Kelsey, 1962). The technique is explained in Fig. 3. Owing to the geometry of the globe, the potential difference recorded between electrodes placed near the medial and lateral canthi varies with eye position, and a sudden standard eye movement can be recorded as an artifact-free voltage change. The magnitude of the voltage varies from person to person, and depends on the shape of the orbit and the position of the electrodes. However, if the eye movement potential alters in the same person from minute to minute, this reflects a change in the magnitude of the generator of the potential. It has been found that alteration of retinal illumination affects the steady potential in a character- istic way. It falls to a minimum in dark adaptation and rises to a peak in subsequent light adaptation. The percentage change in potential is rela- tively constant in normal eyes, and is decreased in disease. These find- ings have been elaborated into a clinical test, the electrooculogram (EOG). Since analogous experiments have not been performed in animals, there

198 Fig. 3. Curro* Flow round the ortft of tyr momme* polfnlKi! Change in tract Lighti —Ort— On Standing PoUntlal ma- /uuuv TmcownA 12 IS IB 21 24 Use of corneo-fundal potential as index of retinal function. Left —schematic cross section of eye with skin electrodes show- ing how rotation of eye alters p.d. between electrodes. If eye movements are standard, observed p.d. is constant fraction of corneo-fundal potential. Right—observed changes in corneo- fundal potential with change of retinal illumination. Sequence is that adopted in clinical testing, and more prolonged complex potential oscillations occur. (Arden, Friedman, & Kolb, 1962) is no absolute proof that the EOG potential is generated in the pigment epithelium. However, it is easy to show in man that the EOG is affected in retinal detachments and in choroiditis where the neuroretina is still functioning. These observations are analogous to the animal experiments which established the site of origin of the steady potential in animals, so the evidence is fairly compelling. Again, it is not understood why the potential changes are related to retinal illumination. It is easy to show that they are linked to the rhodopsin cycle, but this merely means that light has no direct action on the pigment epithelium and has no further localizing value (Arden & Kelsey, 1962a; 1962b). The potential is cer- tainly maintained by active metabolism —it is very sensitive to anoxia— and it is convenient to think of the EOG as testing the maximum working rate of some metabolic process within the pigment epithelium. In chloroquin retinopathy, the EOG is abnormal (Fig. 4). The figure summarizes some of the results obtained in a severely affected patient.

199 2 * 6 B IO 12 14 16 18 2O 22 34 Fig. 4. Some electrophysiological and clinical abnormalities in chloroquin retinopathy. Upper R—visual field. Upper L—dark adaptation curves, showing progressive deterioration. Lower L—ERG A & B—moderately light-adapted; C, fully dark-adapted; lower R-EOG. (Arden & Fojas, 1962b) There is great field loss, dark adaptation is incomplete, and the ERG is abnormal. The b-wave is tiny, and it is followed by an unusually large c-wave, which in this case is not a pupillary artifact. It recalls the large c-waves found in azide-poisoned rabbits. The EOG, lower right, is grossly abnormal and, though there are large "spontaneous" fluctuations in potential, the normal response to illumination is completely absent. In other cases of this condition that have been encountered by the author, even in mild ones where field loss was minimal and the ERG normal, there has always been a depression of the EOG. Even in patients taking chloroquin, in whom there are no signs of a retinopathy, and in whom the ERG and subjective tests are absolutely normal, the EOG may be de- pressed (Arden, Friedman, & Kolb, 1962). Since it has proved possible to use ERG and EOG to distinguish be- tween different sorts of drug action, it might be thought possible to use the various components of the ERG in the same way. This is especially true since the demonstration (Brown & Wiesel, 1961a, 1961b) that the

200 negative component of the mammalian ERG, PIII, is produced by the receptors, and the positive components are produced by cells in the inner nuclear layer. While this work will certainly be applied to under- standing the mechanism of drug action, it is not easy to do so, particu- larly in man, and there are right and wrong ways of going about this task. For example, in the investigation of the diaminophenoxyalkanes, Fojas and the author naturally gave larger doses than those mentioned above. When this is done, the ERG waveform changes considerably (Fig. 5). Immediately after the administration the ERG gets bigger; and then it declines in amplitude until the b-wave has vanished and only B 5 >L jaoo^v so msec Fig. 5. Effect on cat ERG of toxic dose of modified phenoxyalkane. Left- hand column: 1 . control record; 2. immediately subsequent to injection; 3. 2 min later; 4. 10 min; 5. 20 min; 6. 40 min. Right- hand column: effect of smaller dose: 1 and 2 as before; 3. 1 hr later; 4. 1 wk later; 5. 4 wks later. (Arden & Fojas, 1962a)

201 the a-wave is left. Such a sequence of events has often been observed, and equally often it has been assumed that the b-wave is selectively damaged by the drug. This assumption is quite unwarranted, as subse- quent results showed. Any and every toxic agent, from nembutal to anoxia, can produce this series of changes, so it is entirely non-specific. The trouble is that the ERG is the algebraic summation of two po- tentials of opposite sign (Granit & Riddel, 1934), and each in isolation is much bigger than the entire combined ERG. Therefore, a change like that shown in Fig. 5 could also be due to an increase in the size of the a-wave component. Another difficulty facing the interpretation of the ERG abnormalities is the fact that it is produced by the entire retina, mainly by light scattered in the fundus. A localized lesion will not affect the ERG and if in the presence of a large area of retinal damage the ERG is abnormal, it is not possible to decide, in any one case, what fraction of the recorded ERG is derived from the small area of normal retina, and how much represents the diminished response of the diseased tissue. Even if a microelectrode is thrust into the retina, as can be done in animals, it still records the ERG of the entire eye, unless special pre- cautions are taken. In clinical practice it is hoped that the use of weak, localized retinal illumination, and computer techniques will enable the response of small areas of retina to be distinguished. The microelectrode experiments referred to are not entirely convincing: but if the computers succeed, this will be obvious. The true focal ERG has a waveform which is in many respects dissimilar to that seen with conventional recording. Another reason for mistrust of ERG analysis is that alterations in the ERG amplitude may be due to non-visual factors, e.g., an alteration in the distribution of the recording resistance. Finally, in clinical practice, the technique is very difficult, so that although many workers have pub- lished very handsome individual records, the incidence of unrepeatable results is high. For these reasons, it is necessary to be extremely cir- cumspect in attempting to relate ERG and psychophysical data. One way in which this may be attempted is the analysis of dark adaptation, as shown in Fig. 2. However, even here caution is necessary. The dark adaptation curves in the figure extend over a small sensitivity range of only 3 log units. This implies that the light adaptation used was fairly weak, as was the case. It is possible to do better than this but then the ERG waveform changes considerably, and there seems to be no valid reason for relating the sensitivity of the eye to a constant voltage of ERG if, early in the experiment the response is a sharp spike, superimposed on a continuous negative drift, while later it is a gently rising positive wave. A good example of the potentialities and the limits of the ERG is provided by a recent investigation by Dr. Ikeda, in the clinic in London (Institute of Ophthalmology) on the effect of alcohol on vision and the ERG. When the fully dark-adapted subject takes alcohol by mouth, the flicker-fusion frequency decreases and the visual threshold decreases slightly, i.e., the visual system becomes more sensitive as though dark- adaptation had progressed further than is normally possible (Fig. 6).

202 50 45 Alcohol 40cc (Brandy lOOcc) i I * Log 2.4 Ill'11 OATE0FEXP. 13-7-59 • . • • „ o • • o ' • • o • o 11 III 1 0 15 30 9.25 9.45 45 60 75 90 105 120 10.40 11.40 A.M. A.M. A.M. A.M. Fig. 6. Effect of alcohol on flicker fusion rate of human subject. Ordi- nate —flashes/sec. Abscissa—time . Note time course of effect. (Ikeda, 1963) The effect on the intoxicated subject (in this case Dr. Ikeda herself) is small, and runs a peculiar time course. When lovely woman stoops to folly to the extent of 100 milliliters of brandy, the effect is not maximal after 15 minutes, nor is it over at about an hour. The ERG (Fig. 7) also increases and decreases in amplitude along the same time course, and the correspondence is so good that it is possible to state that alcohol has an effect on the receptors or on the inner nuclear layer. The change in amplitude in the ERG is rather striking, much more so than the actual change in sensitivity, but if one carries the analysis a little further, it can be seen that this is not really the case. In the fully dark-adapted eye, a large increase in b-wave amplitude results from a small change in light intensity. In addition (Fig. 8), the action of alcohol varies with the light TOO •oo 500 400 •Control . Alcohol O-TSj/KJ Alcohol 40 50 60 70 SO 90 IOO HO I2O ISO HO OO MIN IN DARK Fig. 7. Effect of alcohol on dark-adapted ERG. Note similarity in time course between ERG and psychophysical changes of Fig. 6. (Ikeda, 1963)

203 l-0 i B e 2-5 BEFORE ALCOHOL AFTER ALCOHOL SUPERIMPOSED Calibration 25O//V 0-25s«c Fig. 8. Waveforms of ERG's before and after alcohol. Note later part of ERG is af- fected. On right, two traces are shown superimposed, with that taken after alcohol shown dotted. (Dceda, 1963) intensity used to evoke the ERG- When the b-wave is small the alcohol has little effect, while intense lights cause a proportionately greater in- crease. Since all the stimuli needed to evoke ERG's were much more intense than those used in the psychophysical experiments, it is impos- sible to relate the b-wave to the subjective threshold, and only the corre- spondence of the time courses enables one to be sure this is the same phenomenon. Dr. Ikeda's ERGs are possibly the best ever recorded from a human subject. The contact lens electrode was individually fitted, and the sub- ject selected. As a consequence, the ERGs are large and artifact-free. They were recorded on a distortion-free system, and Dr. Ikeda has been able to superimpose ERGs taken before and after the administration of alcohol. It can be seen that it is the latter part of the b-wave that has been affected. In view of the fact that the spikes in the cat's optic nerve have a shorter latency than the conventionally recorded ERG, it seems most unlikely that the observed alteration in b-wave amplitude can have any bearing whatsoever on what the cat actually sees, or that it is pos- sible to extrapolate such results to human thresholds; yet this is commonly done by electrophysiologists. This is not to imply that electrophysiological

204 results cannot be used to infer anything about the mechanism of vision, but of course this must be done in terms relevant to electrophysiology. It is interesting to speculate a little on how one may interpret Dr. Ikeda's results. First of all, how do these experiments compare with others in which the eye has been treated with alcohol? The increase in amplitude of the b-wave was first demonstrated by Bernhard and Skoglund (1941), but they also observed that as the b-wave increased in size, the a-wave disappeared Their interpretation was that alcohol selectively reduced PIII, the negative component of the ERG. They were quite aware that the a-wave might merely be hidden in the enlarged b-wave, but other observations made them reject this idea. Now that it is known that the a-wave is the leading edge of a much larger receptor potential, it seems less likely that the interpretation of Bernhard and Skoglund was correct, and it seems pos- sible to explain their results in another way. For the moment, however, it should be pointed out that in Dr. Ikeda's experiments, the a-wave is unaffected by alcohol so that reinterpretation is necessary. There is a discrepancy here, which may be due to the different concentration of alcohol, or to the different stimulus conditions in the two series of experi- ments . This naturally brings to mind the question of under just what con- ditions can the b-wave be shown to be increased after alcohol administra- tion. In the human experiments, the dosage of alcohol is limited. A 120-pound woman can take 100 ml of brandy on an empty stomach, but not much more. Therefore, other stimulus parameters had to be altered. Chief of these was the intensity, and rate of repetition of the stimulus. When this was done it was found that the results after alcohol could be quite different. With a moderate repetition rate —say, 1 flash per second — the ERG was smaller, not bigger in amplitude. This is shown in Fig. 9. There is a curve relating ERG amplitude and stimulus-repetition rate, which to a first approximation is an exponential. What alcohol has done is to change the time constant of the exponential. This correlates with the fact that alcohol decreases the flicker fusion rate. With this finding, the effect of alcohol on the ERG is more under- standable, for it is exactly the same as the effect of anaesthetics, like nembutal, which also have the twin effects of making the single flash ERG bigger, but decreasing the amplitude of the responses to a series of flashes. This phenomenon has been analyzed in the cat (Arden, Granit, & Ponte, 1960); in man it appears quite similar. During retinal illumination, some process builds up in the retina with apparently the same time course as retinal excitation itself, and has the property of suppressing the response to a subsequent stimulus. For this reason, the process was called "suppression." It outlasts the stimulus, and decays exponentially after the end of the stimulus. The effect of alcohol and anaesthetic agents can be described by saying that they increase the amplitude of the b-wave, and increase the time constant of suppression.

205 3OO uV 5 IO 15 20 25 FREQUENCY - c/s Fig. 9. Amplitude of b-wave as function of flicker frequency. Administration of alcohol in- creases size of b-wave only if stimulus repetition rate is low. (Bceda, 1963) The next step is to find some electrical expression of this process, and it is possible that this has been achieved. In some recent experiments. Dr. Brown, Dr. Murakami, and the author have been investigating the effect of nembutal on the isolated receptor potential of mammalian retina. When the stimulus begins, there is an abrupt change in potential, cornea negative, and the potential is maintained throughout the subsequent illumi- nation. When the stimulus ends, the potential gradually returns to its previous resting level. The action of light seems to be merely to change the receptor potential to a new level, from which it recovers exponentially with a time course which approximates the ERG suppression; the action of nembutal is to increase greatly the size of the change in receptor po- tential produced by any light stimulus, as can be seen in the fast sweeps of Fig. 10. In addition, the recovery of the receptor potential is delayed. It is difficult to see this, for though it is easy to eliminate the b-wave (Brown & Watanabe, 1962), the receptor potential cannot be obtained without an interfering c-wave. However, by illuminating the retina, and temporarily interrupting the light —stimulating, as it were, by brief flashes of darkness—one can see the recovery process, and this is so slowed by nembutal that in some experiments no return of the receptor potential toward the base line level can be seen for as long as a second. So, although none of these electrical phenomena can be seen in the intact eye, there is a plausible explanation for the effect of nembutal on the normal ERG, and, extrapolating boldly, for the effect of alcohol on

206 Fig. 10. "Receptor potentials" of mammalian retinas. Records obtained with electrodes in vitreous and on sclera, after selective block- ing of retinal circulation. A. Waveform obtained from night monkey in which c-wave is very small. Two sweeps superimposed. At beginning of stimu- lus (downward pip) cornea-negative potential develops, and is maintained through 5.2 sec illumination. When stimulus ends, potential slowly returns to base line (cal. 500 jiV, 0.5 sec). B, C, D. Effect of injecting 5 mg nembutal i.v. on cat receptor potential. Only initial development of potential shown on these fast sweeps (cal. 500 fiV, 20ms). B—control. C—following in- jection. D—recovery. Note great transient increase in effect of light. E, F, G, H. Recovery of cat "receptor potential" in different stages of anaesthesia. E,G—lightly; F, H—deeply anaesthetised . Receptor potentials E & F are of equal amplitude, but recovery phase cannot be seen due to large positive C-wave. In G & H, retina constantly illuminated and signals show short periods (280 ms) of darkness. In this way, C-wave is eliminated and it can be seen that rate of recovery of receptor potential in G, lightly anaesthetised, is greater than in H. Calibrations, 500 jiV and 500 ms for E and G; 200 ms for F and H.

207 vision. It seems likely that both these drugs affect the receptors, so that a given quantity of light causes a greater depolarization of their terminals, but the repolarization in the dark is slower. A second stimu- lus, following close on the first, will now depolarize the receptors to the active level, but because recovery has been delayed, the change in ob- served potentials as normally seen (the a- and b-wave, and subsequent discharge of nerve impulses) will be smaller than usual. In psychophys- ical terms, one would predict that the dark-adapted threshold would de- crease, and that the flicker fusion frequency would also decrease. This is just what occurs. One is also in a position to explain Bernhard's in- terpretation of the effect of alcohol on the frog eye. Observing a diminish- ing a-wave and an enlarging b-wave, he suggested that alcohol depressed PIII—the negative component of the ERG. This interpretation was strongly supported by experiments in which the eye was reilluminated at the time of the "off response. " In the frog, this leads to the production of a large a-wave and the appearance of pre-excitatory inhibition. After treatment with alcohol, the a-wave is not seen, and this suggests that PIII is sup- pressed . Bernhard was unable to obtain either PIII or PII in a pure form and, therefore, his results can be equally well explained by supposing that PIII was in fact increased, but its recovery delayed. These observations, then, may not only explain an old and puzzling ERG phenomenon, but may aid in understanding the effects of drugs on vision. The most interesting point to notice is the progress made from the observation of an increased b-wave amplitude to an increased sensi- tivity and delayed recovery of a receptor potential. How far from the truth would have been merely an attempt to relate crude electrophysio- logical and psychophysical data. References Arden, G., Barrada, A., & Kelsey, J. New clinical test of retinal function based upon the standing potential of the eye. Brit. J . Ophthal., 1962, 46, 449-467. Arden, G., & Fojas, M. The mode of action of diaminophenoxyalkanes and related compounds on the retina. Vision Res., 1962, 2, 163-173. (a) Arden, G., & Fojas, M. Assessment of the value and the basis of electro- physiological abnormalities in pigmentary degenerations of the retina. AMA Arch. Ophthal., 1962, 68, 369-390. (b) Arden, G., Friedman, A., & Kolb, H. Anticipation of chloroquine retinopathy. Nature (London) 1962. Arden, G., Granit, R., & Ponte, F. Phase of suppression following each retinal b-wave in flicker. J. Neurophysiol., 1960, 23, 305- 314. -----.-

208 Arden, G.. & Kelsey, J. Changes produced by light in the standing potential of the human eye. J. Physiol.. 1962, 161, 189-205. (a) Arden, G., & Kelsey, J. Some observations on the relationship between the standing potential of the human eye and the bleaching and re- generation of visual purple. J. Physiol.. 1962, 161, 205-224. (b) Ashton, N. Degeneration of the retina due to 1.5 Di (p-aminophenoxy) pentane dihydrochloride. J. path. Bact., 1957, 74, 103-112. Bernhard, C., & Skoglund, C. Selective suppression with ethyl alcohol of inhibition in the optic nerve and of the negative component PIII of the electroretinogram. Acta Physiol. Scand., 1941, 2, 10-21. Brown, K., & Watanabe, K. Rod receptor potential from the retina of the night monkey. Nature (London). 1962, 196, 547 & 550. Brown, K., & Wiesel, T. Analysis of the intraretinal electroretinogram in the intact cat eye. J. Physiol.. 1961, 158, 229-256. (a) Brown, K., & Wiesel, T- Localisation of origins of electroretinogram components by intraretinal recording in the intact cat eye. J. Physiol., 1961, 158, 257-280. (b) Burian, H., & Fletcher, M. Visual functions in patients with retinal pigmentary degeneration following the use of NP-207. AMA Arch. Ophthal.. 1958, 60, 612-629. Bowling, J . E ., & Gibbons, I. The effect of vitamin A deficiency on the fine structure of the retina. In G. K. Smeltzer (Ed.), The structure of the eye. N.Y.: Academic Press, 1960. Pp. 85-100. Glocklin, V., & Potts, A. Metabolism of retinal'pigment of cell epithelium. 1. The in vitro incorporation of P-32 and the effect of diaminodi- phenoxyalkane. Invest. Ophthal., 1962, 1, 111-117. Goodwin, L. G., Richards, W. H. G.. & Udall, V. The toxicity of diaminodiphenoxyalkanes. Brit. J. Pharmacol., 1957, 12, 468- 474. Granit, R. & Riddel, L. The electrical responses of light- and dark- adapted frogs' eyes to rhythmic and continuous stimuli. J . Physiol., 1934, 81, 1-28. Ikeda, Hisako. Effects of ethyl alcohol on the evoked potential of the human eye. Vision Res.. 1963. 3. 155-169.

209 Noell, W. Studies on the electrophysiology and metabolism of vision. Brook AFB: USAF Sch. av. Med. Rep., 1953, No. 1. (Proj. No. 21-1202-0004) Noell, W. Differentiation, metabolic organization and viability of the visual cell. AMA Arch. Ophthal., 1958, 60, 702-738. Potts, A. Concentration of phenothiazines in the eye of experimental animals. Invest. Ophthal.. 1962, 1, 522-530.

210 STUDIES IN THE PHARMACOLOGY OF EXTRAOCULAR MUSCLES1 G. M. Breinin and J. H. Ferryman The literature has suggested that extraocular muscle is a primitive type intermediate between striated muscle of mammals and lower forms, and that it demonstrates an affinity with smooth muscle. These unusual features have been linked to the peculiar nerve supply and highly special- ized functions of extraocular muscle. The anatomical and histological distinction of extraocular muscle, its low innervation ratio, variety of nerve and nerve endings, and membrane-permeability characteristics have been adduced in explanation of certain aspects of ocular motility. Studies of the authors, however, tend to show that differences between eye muscles and limb muscles are quantitative rather than qualitative, and may be considered modifications related to the highly specialized nature of eye movements. This paper summarizes a series of experimental observations on the pharmacology of extraocular muscles in cat and dog, primarily addressed to the question of whether the postulated uniqueness of extra- ocular muscle actually exists and, if so, to what extent. The failure of epinephrine to elicit muscle contraction was seen. Epinephrine in mas- sive dosage injected directly into the carotid artery produced no meas- urable change in the rest tension of the inferior oblique, lateral, or medial rectus, while changes in the epinephrine-sensitive structures in the orbit (nictitating membrane and pupil) were readily observable. A gross comparison was made between the effect of acetylcholine in the extraocular muscle and that in the limb muscle . An injection of 80 micro- grams (^g) of acetylcholine into the carotid regularly produced maximal contraction of extraocular muscles. The drug was administered with a hypodermic syringe through a polyethelene cannula tied into the stump of the lingual artery. Eighty ^g of acetylcholine injected into the popliteal artery also produced a strong contraction of the anterior tibial muscle. With intracarotid injection, it required about 30 to 50 /ag of acetylcholine to produce a significant ocular muscle contraction, while anterior tibial contraction could be produced with as little as 2 to 5 /ig of acetylcholine. It is difficult to draw conclusions as to the relative sensitivity of ocular and limb muscles to acetylcholine since the responses are so largely affected by the technique employed. It is clear, however, that both types In M. B. Bender (Ed.). In the oculomotor system symposium, in press.

211 of muscle respond briskly to minute amounts of acetylcholine. Choline is said to produce a slow, tonic contraction of extraocular muscle and to be without effect on innervated skeletal muscle. Although large amounts are required to evoke a response, a definite contractile effect (with about 100 times concentration of acetylcholine) was seen in both extraocular muscle and anterior tibial muscle with close arterial injection. Nicotine was also tested. It was found that with close arterial injection as pre- viously explained, strong contraction could be elicited from the extra- ocular muscle with the intracarotid injection of 1 cubic centimeter of nicotine in a 0.1 per cent solution of saline, and with one-half of this amount in the normal anterior tibial muscle. Topical application of atropine in vivo was ineffective, while close arterial injection would pro- duce a short-term block in both sets of muscle. Intravenous or intra- carotid injection of succinylcholine produced a strong, sustained extra- ocular muscle contraction. Close arterial injection of the anterior tibial muscle, however, produced a comparable result. Differences in the route and technique of drug administration can produce marked differences in result. An intracarotid injection is in reality a close arterial injection of extraocular muscle. A comparison of effect in other muscles can be made only when they also receive a close arterial injection, especially when a drug is quickly inactivated in the blood as in the case of acetylcholine. The amount of drug, the rate of injection, and the volume of diluent are interrelated factors in making a comparison of drug effect in different muscles. The effector site is the motor endplate. The peripheral skeletal muscle has proportionally fewer effector sites than the extraocular muscle, and these are spread out because of larger mass. The flow velocity and percentage satura- tion of blood is higher in extraocular muscle and lower in peripheral skeletal muscle. Taking many of these factors into consideration, much of the so-called uniqueness based on intrinsic physical or chemical factors falls away, and many of the reactions attributed to the peculiar physiology of eye muscles become understandable in terms of ordinary skeletal muscle behavior. The distinctions that exist are quantitative rather than qualitative. References Adler, F. H. Physiology of the eye. (3rd ed.) St. Louis, Mo: C. V. Mosby Co., 1960. Breinin, G. M. Electrophysiology of the extraocular muscles. Toronto: Univer. of Toronto Press, 1962. Brown, G. L., & Harvey, A. M. Neuromuscular transmission in the extrinsic muscles of the eye. J . Physiol.. 1941, 99, 379-399.

212 Cooper, S., & Fillenz, J. Afferent discharge in response to stretch from the extraocular muscles of cat and monkey and the innerva- tion of these muscles. J. Physiol., 1955, 127, 400-413. Duke-Elder, W.S. New observations on the physiology of the extraocular muscles. Trans. Ophth. Soc., (U.K.) 1930, 50, 181. Lincoff, H. A., Ellis, C. H., DeVoe, A. G., DeBeer, E. J., Impastato, D. J., Berg, S-, Orkin, L., & Magda, H. The effect of succinyl- choline on intraocular pressure . Amer. J . Ophth., 1955, 40, 501-510. McCouch, G. P., & Adler, F. H. Extraocular reflexes. Amer. J. Physiol., 1932, 100, 78-88. Rowley, P. T., Wells, J. B., & Irwin, R. L. Tension response of mammalian muscle to intra-arterio acetylcholine. Amer. J . Physiol., 1960, 198, 507.

DISCUSSION on The Effects of Drugs on Vision

215 DRUGS AND EYE MOVEMENT RESPONSES IN MAN Gerald Westheimer School of Optometry University of California Barbiturate nystagmus, a well-known clinical phenomenon, was shown by Rashbass (1959) to be tied in with the observation that barbitu- rates selectively abolish the smooth pursuit eye movements while leaving saccadic movements unaffected. Barbiturates also interfere with dis- junctive eye movements: vergence movements are appreciably reduced in speed and amplitude. Under the influence of barbiturates, the near point of convergence recedes, the distance phoria changes in the direction of esophoria, and the near phoria changes to exophoria. Accommodation remains unaffected. These changes are found with moderate therapeutic doses and have a time course that suggests that they are good indicators of the level of intoxication. A study of the accommodation-convergence synkinesis reveals that it is quite drastically reduced by barbiturates. On the other hand, amphetamines enhance it while leaving all other oculo-motor responses substantially unchanged. References Rashbass, C. Barbiturate nystagmus and the mechanism of visual fixation. Nature, 1959, 185, 897-898. Westheimer, G. Amphetamines, barbiturates and the accommodation- convergence synkinesis. AMA Arch. Ophthalmol., in press. Westheimer, G., & Rashbass, C. Barbiturates and eye vergence. Nature, 1961, 191, 833-834.

216 THE EFFECTS OF DRUGS ON VISION Leon S. Otis Stanford Research Institute The excellent papers presented by Drs. Potts, Arden, and Breinin illustrate how drugs may be used as tools to explore the physiology and structural characteristics of the visual apparatus. The emphasis, how- ever, has been on the peripheral organ —the eye. To a behavioral psychologist, the eye is part of a larger system —the individual. When drugs are used, it is the possible alteration of the information-gathering function of the eye and the behavioral consequences that may attend such alteration that are of primary interest. That is, the psychologist would like to know to what degree drugs change how well the eye can see, what it sees, and the adequacy of the individual's performance within his visual world. These comments in no way detract from the importance of the excellent work reported by the previous authors. They are meant, how- ever, to point up the vastness of the chasm that exists between knowledge of how the eye works, and knowledge of how the individual processes and utilizes visual information in performance. The problem is difficult enough in its own right; it is especially complicated when drugs are introduced. A literature relating drugs to changes in visually dependent behav- ior is practically non-existent; a brief review of some recent papers illustrates the paucity of good experimental data. The articles are dis- cussed in terms of five categories that reflect, in descending order, the degree to which the experimental design included attempts to correlate performance with visual functions, or with underlying neural or chemical mechanisms. The first and by far the most voluminous category (also the least satisfying from the point of view of controlled observation) consists of clinical reports of side effects when drugs are used in medical practice. Illustrative of such reports is the following excerpt from a recent paper (Murray, 1961). A patient, described by the author as "a man who had a previously unblemished record of over 20 years' driving connected with his work . . ."(p. 168) was involved in a series of relatively serious automobile

217 collisions during a 90-day period following his release from the hospital. During this period he was on a main- tenance dosage of 75 mg. of Librium daily. As a con- sequence of the accidents, ophthalmological evaluation was performed and the patient was diagnosed to be suf- fering "from a mild anisometropia unequal accomoda- tion, esophoria for distance, and exophoria for near vision and poor depth perception." Librium was dis- continued for seven days and he was again tested for ocular function, using the Maddox-Rod muscle balance test, and the Howard-Dolman depth perception test. With- out the drug, the patient's responses were completely normal. The medication had affected his depth percep- tion as well as his eye muscle balance . (This occurred in 2 other patients who showed similar reversible changes. One of them was on a 30 mg. daily dosage and the other on a 50 mg. daily maintenance dosage. Ten other patients who complained of visual difficulty and found to require new lens prescriptions.) Mr. L.K.'s daily dosage of Librium was reduced to 25 mg. daily and on retesting, his vision was within normal limits. Since then he has had no further automobile accidents or driving offenses, (p. 169). Clearly, this report leaves much to be desired. One wonders to what degree the deterioration in driving skill was due to postdrug alter- ation of visual functioning, and to what degree it was due to other causes -- for example, changes in motor functioning, emotional reactivity, judgement processes, and the like. Be that as it may, the clinical literature has implicated a number of drugs as causing altered visual functioning, with attendant behavioral impairment. Among those most frequently mentioned are the pheno- thiazines, certain of the monoamineoxidase (MAO) inhibitors, and the psychotomimetic drugs. Chlorpromazine and related phenothiazines are frequently reported to precipitate oculogyric crises and blurred vision (Affleck, Booth, Forrest, &Mackay, 1962; Apt, 1960; Kozinn &Weiner, 1960). Structural changes have also been reported. One phenothiazine (NP-207), that happily did not reach the market, has caused impaired dark adaptation, loss of acuity, and retinal pigmentary degeneration (Apt, 1960); damage to retinal elements has also been reported after thioridazine (Weekley, Potts-, Reboton, &May, 1960). Pheniprazine, an MAO inhibitor recently removed from the market, has been reported to induce toxic amblyopia, decreased ability to discriminate colors (especially reds and greens), blurred vision, and impaired depth per- ception (Jones, 1961; Lear, Browne, &Greeves, 1962; Palmer, 1963). Perphenazine has also been reported to produce transient blindness (Apt, 1960; Johnson, 1960), and disturbance of color vision may occur after tridione (Cox, 1961). Mescaline, lysergic acid diethylamide (LSD), and

218 psilocybin have been implicated in impaired color discrimination, blurred vision, and, of course, visual hallucinations (Hollister &Hartman, 1962); LSD also caused overestimation of the apparent horizon (Wapner &Krus, 1959) and increased variability of size consistancy judgements (Weckowicz, 1959). Acute myopia has been reported for prochlorperazine (Yasuna, 1962); and altered depth perception, poorer accomodation, and blurred vision have been reported after chloriazepoxide therapy (Murray, 1961). Amitriptyline causes disturbances in accomodation (Lambert, Charriot, VuDinh, &Versme'e, 1962). The widely used compound, imipramine, may result in poorer accomodation, blurred vision, and hallucinatory disturbances (Fleming &Groden, 1962; Friedman, DeMowbray, & Hamilton, 1961; H8hn, Gross, Gross, &Lasagna, 1961; Pollack, 1962). As these drugs have a potent effect on the central nervous system, it is, perhaps, not surprising that the visual system is so vulnerable, including, as it does, about 12 per cent of the cerebral cortex. Little can be said for the remaining categories. Too few research papers exist to justify more than a passing recognition of pioneering efforts in a neglected field. The second category includes papers that report the effects of drugs on visual (or perceptual) functions, per se, using relatively simple indi- cator responses as performance measures. Postdrug impairment of dark adaptation (Apt, 1960), size constancy, depth perception, apparent motion, spiral after-effect and after-images (Costello, 1960a; 1960b; 1960c), brightness perception (Weiner &Ross, 1962), visual discrimina- tion (Fuster, 1959), and visual thresholds (Blough, 1957; Carlson, 1958; Krill,. Wieland, &Ostfield, 1959) have been reported in both man and animal. The third category involves studies that measure drug effects on relatively complex visuo-motor performance. The classic studies of Payne &Hauty (1955; 1958) on the facilitating effects of amphetamine and other stimulants on visual monitoring, and those of Mackworth (1950), Kornetsky (Kornetsky, Mirsky, Kessler, &Dorff, 1959), and others on vigilance, practically exhaust this category. Papers comprising category 4 attempt to relate postdrug changes in visual response to higher order theories of behavior or brain func- tioning. Costello (1960a; 1960b; 1960c) has reported changes in apparent movement, after-images, and spiral after-effects following clinical doses, 400-600 milligrams (mg), of meprobamate which have no effect on visual threshold (Melikian, 1961). He relates these changes to Eysenck's general theory of brain functioning which states that depressant drugs decrease brain excitability potential and increase inhibitory potential; stimulus drugs are presumed to have the opposite effect (Eysenck, 1957). Unfortunately, meprobamate was the only drug investigated. It would

219 have been desirable to include other depressant-type drugs, e.g., chlorpromazine and pentobarbital, in the design as well as a stimulant for comparative purposes. Hollister and Hartman (1962), using a Latin square design, reported that LSD-25, psilocybin, and mescaline reduced color discrimination of normal subjects and increased the frequency of reports of color exper- iences after non-adequate stimuli, i.e., tones and/or an anchromatic visual flicker stimulus. These findings tend to support their hypothesis that the heightened and unusual perception of colors following psychotomi- metics may be due to excitation of the central "color response" by non- adequate stimuli. The mechanism, however, remains to be specified. Studies by Krill et al_(1959), Fuster (1959), and by Carlson (1959) are illustrative of papers that attempt to relate changes in visual per- formance to underlying mechanisms (Category 5). Krill_et^ al_studied the electroretinogram (ERG) and dark adaptation after low and high doses of two hallucinogens, LSD-25 and JB-318.1 Con- trol drugs were nonhallucinogenic analogues, Methysergide (UML) and JB-808.1 Fifteen subjects hallucinated after 75 micrograms (pg) LSD and 7.5-15 mg JB-318, and significant changes in ERG, increase in scotopic b-wave amplitude, or in dark adaptation, elevation of the entire rod thresh- old and a delay in rod-cone break, were observed. The analogues had no discernible effect on vision. The authors suggest that the visual disturb- ances associated with the hallucinogens may have been due to the hypoxia or to toxic retinal effects, possibly induced by LSD and JB-318. Fuster trained five monkeys in the WGTA to discriminate visually a cone from a 12-sided pyramid. LSD at 2-8 pg decreased the accuracy of discrimination and increased the latency of response—findings that are at variance with Blough's report that LSD improved visual dis- crimination in the pigeon (1957). Fuster (1959) correlates these findings with known excitatory effects of LSD on axosomatic fibers, i.e., sensory fibers or fibers that carry information to the brain, and inhibitory effects on axodendritic fibers, i.e., fibers which emanate from the diffuse thaiamic system and reticular formation and which control "activation" of the brain. Carlson (1958) reported that photopic threshold was raised more than scotopic threshold after non-hallucinatory doses of LSD—findings that are consistent with the occurrence of inhibitory effects on cortical processes as reported by Purpura (1956); elevation in threshold was shown to be unrelated to possible inattention and inability to concentrate. This then, is the present state of the art; a fair number of clinical reports implicate at least some drugs in the deterioration in visual processes and performance, but few "hard" experimental studies have been done. The challenge should and, hopefully, will be met. 1 JB compounds are experimental drugs of Lakeside Company.

220 References Affleck, J. W., Booth, J. C. D., Forrest, A. D., &Mackay, K.J. The effect of thioproperazine administered by a continuous method on long-term schizophrenic patients. J. ment. Sci., 1962, 108, 862-864. Apt, L. Complications of phenothiazine tranquilizers: ocular side effects. Survey Ophthal.. 1960, 5, 550-5. Blough, D.S. Some effects of drugs on visual discrimination in the pigeon. Ann. N. Y. Acad. Sci., 1957, 66, 733. Carlson, V. R. Effect of lysergic acid diethylamide (LSD-25) on the absolute visual threshold. J. comp. physiol. Psychol., 1958, 51, 528-531 . Costello, C. G. The effects of meprobamate on perception. I. apparent movement. J. ment. Sci., 1960, 106, 322-326. (a) Costello, C. G. The effects of meprobamate on perception. II. the visual after-image. J. ment. Sci., 1960, 106, 326-330. (b) Costello, C. G. The effects of meprobamate on perception. III. the spiral after-effect. J. ment. Sci., 1960, 106, 331-336. (c) Cox, J. The effect of tridione on colour vision. Brit. J . Physiol., 1961, Opt. 18, 79-84. ~~~ Eysenck, H. J. Drugs and personality: I. theory and methodology. J. ment. Sci., 1957, 103, 119-131. Fleminger, J. J., & Groden, B. M. Clinical features of depression and the response to imipramine ("tofranil"). J. ment. Sci., 1962, 108, 101-104. Friedman, C., DeMowbray, M.S., & Hamilton, V. Imipramine (tofranil) in depressive states: a controlled trial with in-patients. J. ment. Sci.. 1961, 107, 948-953. Fuster, J. M. Lysergic acid and its effects on visual discrimination in monkeys. J. nerv. ment. Dis., 1959, 129, 252-256. Hauty, G. T., & Payne, R. B. Effects of analeptic and depressant drugs upon psychological behavior. Amer. J. publ. Hlth., 1958, 48, 571-577.

221 Hohn, R., Gross, G. M., Gross, M., & Lasagna, L. A double-blind comparison of placebo and imipramine in the treatment of depressed patients in a state hospital. J . psychiat. Res ., 1961, 1, 76-91. Hollister, L. E., & Hartman, A. M. Mescaline, lysergic acid diethylamide and psilocybin: comparison of clinical syndromes, effects on color perception and biochemical measures. Compr. Psychiat.. 1962, 3, 235-241. Johnson, W. Toxic amblyopia from perphenazine (fentazin), J. ment. Sci., 1960, 106, 352-354. Jones, O. W. Toxic amblyopia caused by pheniprazine hydrochloride (JB-516, catron). Arch. Ophth., 1961, 66, 29-36. Kornetsky, C., Mir sky, A. F., Kessler, E. K., &Dorff, J.E. The effects of dextro-amphetamine on behavioral deficits produced by sleep loss in humans. J. Pharmacol., 1959, 127, 46-50. Kozinn, P. J., & Wiener, H. Oculogyric crisis after a small dose of perphenazine. J. AMA, 1960, 174, 166-167. Krill, A. E., Wieland, A. M., & Ostfeld, A. M. The effect of two hallucinogenic agents on human retinal function. AMA Arch. gen. Psychiat., 1959, 1, 417-419. Lambert, P. -A., Charriot, G., VuDinh, G., & Versmee, A. Apropos d'un nouvel antidepresseur: le Ro4-1575. J. de Medecine de Lyon, 1962, 1008, 635-640. Lear, T. E., Browne, M. W., & Greeves, J. A. A controlled trial of cavodil (pheniprazine) in depression. J. ment. Sci., 1962, 108, 856-858. ~~ Mackworth, N- H. Researches on the measurement of human performance. London: (HMSO) Medical Res. Council Spec. Rep., 1950, Series No. 268. Melikian, L. The effect of meprobamate on the performance of normal subjects on selected psychological tasks. J. gen. Psychol., 1961, 65, 33-38. Murray, N. Covert effects of chlordiazepoxide therapy. Amer. J. Psychiat.. 1961, 118, 168-170. Palmer, C. A. L. Toxic amblyopia due to pheniprazine. Brit. med. J., 1963, No. 5322, 38.

222 Payne, R. B., & Hauty, G. T. Factors affecting the endurance of psychomotor skill. J. exp. Psychol., 1955, 50, 382-389. Pollack, B. Imipramine-promazine therapy for depression. Amer . J . Psychiat.. 1962, 118, 842-845. Purpura, D. P. Electrophysiological analysis of psychotogenic drug action. II. General nature of lysergic acid diethylamide (LSD) action on central synapses. AMA Arch. neurol. Psychiat., 1956, 75, 132-143. '""~~ Wapner, S., & Krus, D. M. Behaviorial effects of lysergic acid diethylamide (LSD-25). AMA Arch. gen. Psychiat., 1959, 1, 417-419. Weckowicz, T. E. The effect of lysergic acid diethylamide (LSD) on size constancy. Canad. psychiat. Assoc . J., 1959, 4, 255-259- Weekley, R. D.. Potts, A. M., Reboton, J., & May, R. H. Pigmentary retinopathy in patients receiving high doses of a new phenothizaine. AMA Arch. Ophth.. 1960, 64, 65-76. Weiner, H., & Ross, S- Effects of d-amphetamine sulfate on time and brightness perception in human subjects. Psychopharmacologia, 1962, 3, 44-50. Yasuna, E. Acute myopia associated with prochlorperazine (compazine) therapy. Amer. J. Ophth., 1962, 54, 793-796.

223 SOME POTENTIAL OF RESEARCH ON DRUGS AND VISION Richard Trumbull Office of Naval Research The author's interest in the effects of drugs on vision goes back to some experiences of the early 1950's while testing anti-seasickness drugs. A number of the more promising, based on laboratory studies, were taken to sea in a "double blind" evaluation in the operational situation. They were standard compounds, available to the public. Troop trans- ports with 10,000 subjects available provided an ideal situation. Although anyone would expect individual differences to appear in response to the drugs and their effectiveness, it was surprising to find that the range reported ran from decrement to actual improvement in vision. One man apologized for not filling out the questionnaire card because he couldn't read the print. Indeed, his pupils were so dilated that "sack-time" was his only occupation. Others commented on ability to see farther and more clearly. Similar comments were made about improved auditory function. This situation necessitated checking the literature to see just what history had to offer (Trumbull & Maag, 1958). There are the rather dramatic instances where mescaline has produced artistic products judged by experts to be far superior to those normally produced. Cer- tainly, anecdotal claims of this type are well known, as are those related to contrast enhancement. The improved sense of timing gained has been one reason for drummers, and others seeking exotic off-beats or new rhythms becoming addicts. Then there are more reliable data where the laboratory and control are found. It is surprising how many times re- searchers have reported improvements in sensory modalities resulting from drugs. The author's interest was purely on the positive side, where performance was improved, maintained, or restored from deficit by ingestion of some biochemical. Here are a few examples of what were found. 1. "(0.6 - 1 •) mgm. of carcholin was administered to human sub- jects with a resultant lowering of dark adaptation threshold and improved dark adaptability corresponding to the dosage. " (Fang, Hwang, & Hall, 1953) 2. A study of caffeine influence on color naming—". . .stimulation at all doses but better at lower." (Hollingworth, 1912)

224 3. "It can be concluded that the decline of important elements of pilot skill can be delayed for at least four hours by pharmacological techniques." (Payne & Hauty, 1953) 4. Mescaline and lysergic acid produced ". . .greater sensitivity to sounds and most striking effects of visual stimuli. Colors had added 'purity, ' 'vividness' and 'brilliance'." (Berlin et al., 1950) There are many additional items (Trumbull & Maag, 1958) where there is direct reference to the sensory modalities. If one considers the multitude of performance studies where separation of sensory, central, and motor aspects is impossible, one is impressed with the potential of this area of research. By that is meant the large number of studies in which ability in mathematics, coordination, and complex reac- tion time might have improved as much from a simple visual improve- ment as it would from a central or motor one. Color naming, where time is not an element, is easier to claim. In studies of stress or fatigue, there is a major requirement for more knowledge as to just what influence these conditions have on sensory systems —the base from which it is desired to produce improvement. This experience caused the authors to try to induce the Navy and the Department of Defense to concern themselves with "positive psycho- pharmacology, " seeking some fraction of the money now spent to degrade or destroy man for purposes of realizing the full potential of his systems. The Stanford Research Institute became a keystone in this effort by doing a critical review of various scientific disciplines which had relevance to the problem. Leon Otis' presence in this effort has been a vital factor in seeking the more promising techniques and researchers, and develop- ing a sensitivity to the problems and holes in present knowledge. His paper presented in this section has highlighted much of what has been learned. There is no reason, after this review, (Plotnikoff, Birzes, Mitoma, Otis, Weiss, & Laties, 1960) to be any less enthusiastic about the potential that can be made available by this means. The three papers and two discussions preceding this presentation serve to underscore the fact of this potential while warning of the demand for improved methodology, adequate control, and more basic work on the biochemistry operating at every point in the human system involved. Research on drugs and the muscle system of the eye has also been presented. Obviously, anything which influences ocular shape, motion, transmission, or other functions can be used to further the knowledge and use of vision. Thus, the history and potential of drugs that produce these effects merit similar consideration as those to which reference has been made. It is hoped that pharmacologists will turn some attention to sensory modalities. It is urged that those interested in sensory and perceptual capabilities of man can broaden their research approach, or seek pharmacological co-investigators to answer some of the challenge and seek realization of some of the potential of this area.

225 References Berlin, L., e_t al. Studies in human cerebral functions: the effects of mescaline and lysergic acid on cerebral processes pertinent to creative activity. J. nerv. &ment. Dis.. 1955, 122, 487-491. Fang, H. S., Hwang, T. F., & Hall, A. L. Effects of carcholin on dark adaptation and visual purple regeneration. Pensacola: USN Scn. av. Med. Rep., 1953. (Proj. No. NM 001 059.30.02) Hollingworth, H. L. The influence of caffeine on mental and motor efficiency. Arch. Psych. (N.Y.), 1912, 3, 22. Payne, R. B., & Hauty, G- T. The pharmacological control of work output during prolonged tasks. Randolph AFB: USAF Sch. av. Med. Rep., 1953, No. 2. (Proj. No. 21-1601-0004) Plotnikoff, N., Birzes, Lucy, Mitoma, C., Otis, L., Weiss, B., & Laties, V- Drug enhancement of performance. Menlo Park: SRI, 1960. - — Trumbull, R., & Maag, C. An annotated bibliography and critical review of drugs and performance. Washington: ONR Rep., 1958, No. ACR-29.

226 CONCLUDING REMARKS BY THE CHAIRMAN Arthur Jampolsky Eye Research Institute Presbyterian Medical Center Drs. Potts and Arden have pointed out how the biochemist and the electrophysiologist seek to determine the sites and modes of action of drugs on the visual apparatus. Toxic effects, and the consequent visual decrements, are valuable ways to get at the sensory side of the picture. This information is basic to the problem of maintenance and enhancement of visual performance. The different electrophysiological tools tap off different parts of the visual input system. Dr . Breinin has emphasized the similarity in drug effects on the extraocular muscles and other skeletal muscles, and indicated the mode of action of drugs on the accommodative-convergence relationship. Dr. Westheimer has related some of the examples of the visual psychophysical investigations in this field, an area in which there is a relative paucity of work. Dr. Otis has stated that despite the large literature on the behav- ioral aspects of the problem, there is a lack of acceptable experimental evidence to elucidate specifically the performance changes. The factors of inattention, excitation by non-adequate stimuli, and the multiple sites of drug action, make it a difficult task to unscramble the variables. Dr. Trumbull has presented a background picture of the past history, the present status, and the future potential of research seeking to answer some of the specific problems in practical terms, especially the maintenance and enhancement of man's systems. Some important monographs in this field have been brought to the attention of this group: they relate and coordinate different disciplinary approaches that point out some of the basic needs and practical problems. From the above discussions it appears that the psychophysicist and the behavioral scientists will have difficulty in evaluating the response effects, judgmental changes, and performance results. Controls are difficult to establish, since drugs may decalibrate the subject for visual psychophysical determinations. It is often difficult to determine whether one is affecting the visual apparatus, per se, or the subject generally in

227 his response to the visual performance task. Double blind studies do not avoid the effect of a drug or placebo administration. The biochemist and electrophysiologist have the unique opportunity of assessing nervous-tissue function in terms of the eye as a peripheral brain. Computer techniques materially aid the detection of faint signals that are all but lost in the roar of biological noise. Multiple stimuli add another analytical dimension. Electrophysiological techniques do not always avoid the limitation of drug-altered neurophysiological recording, which may be monitored by polarographic electrodes to assess blood flow and metabolic effects. More must be known about the specific drug alterations of neuronal conduction time, neuronal recovery time, synapse blocking, shuttling, myoneural junction, storage-resynthesis, etc. There is a unique opportunity to assess the retina as a plug of central nervous system tissue, with blood vessels and human responses attached. There is a need to know much more specifically how drugs affect the visual apparatus, per se, especially with respect to the sensory input and motor output visual systems, which lend themselves to easier direct analysis than does "that messy, raveled knot of pinkish jelly. "

THE NATIONAL ACADEMY OF SCIENCES—NATIONAL RESEARCH COUNCIL is il private, nonprofit organization of scientists, dedicated to the furtherance of science and to its use for the general welfare. The Academy itself was estab- lished in 1863 under a Congressional charter signed by President Lincoln. Empowered to provide for all activities appropriate to academies of science, it was also required by its charter to act as an advisor to the Federal Govern- ment in scientific matters. This provision accounts for the close ties that have always existed between the Academy and the Government, although the Academy is not a governmental agency. The National Research Council was established by the Academy in 1916, at the request of President Wilson, to enable scientists generally to associate their efforts with those of the limited membership of the Academy in service to the nation, to society, and to science at home and abroad. Members of the National Research Council receive their appointments from the President of the Academy. They include representatives nominated by the major scientific and technical societies, representatives of the Federal Government, and a number of members-at-Iarge. In addition, several thousand scientists and engineers take part in the activities of the Research Council through member- ship on its various boards and committees. Receiving funds from both public and private sources, by contributions, grant, or contract, the Academy and its Research Council thus work to stimulate research and its applications, to survey the broad possibilities of science, to promote effective utilization of the scientific and technical re- sources of the country, to serve the Government, and to further the general interests of science.