Report of the Committee on Paleoecology, 1935-1936: Presented at the Annual Meeting of the Division of Geology and Geography, National Research Council, May 2, 1936 (1936)

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.

- 29 - (Appendix to 'Taleoecology of the Arthropoda") , "i • • HABITS OF TRILOBITES •' By William E. Schevill (Harvard University) Alimentary Apparatus and Diet In 1919 A. Born described the cephalic musculature of trilobites as his studies of Chasmops odini Eichwald led him to see it (6)*. Aside from his discussion of the muscles of the cephalic appendages, he maintained that the entire space between the hypostoma and the glabella was empty of ali- mentary tract, being occupied to its exclusion by a web of muscles operat- ing tiie hypostoma. This is his interpretation of the interior pustulation of the frontal lobe, and of the hypostoma. Schmidt's figures (32*, pi. II,'figs. 2, 3, 10, 16 and pi. t", figs. 1, 7, 9) of this and allied species show two rows of scars diverging from the center of the frontal lobe arid extending towards the anterior edge. He does hot illustrate their meeting after a sharp inward turn near this border, or the group of three or four dots, also made known by Born, a little to the rear. Born correlated this pattern with scars on the interior of the hypo- stoma which lie directly beneath, and suggested the existence-of a web of muscle's stretched between them, which would have restricted the alimentary tract to the space above and behind the mouth (6, fig. 4, p. 166, or 30, fig. 15, p. 87), which he places at the distal end of the hypostoma**. * Soo bibliography, pp. 41-43. ** 3o do Walcott (37, pi. IV, fig. 6), Beechor (4, pi. V, fig. 11), and Raymond (23, p. 69, and fig. 24, p. 81). Its situation in the Apodidae (the speci- mens examined were of Lepidurus productus (Kr6*yer)), and in cirriped nauplii (10, pl« V, fig. 7), is within the hypostoma, which in the Apodidae, for exemple, covers also the large mandibles* As Beecher says (4, p. 96) the appendages of Triarthrus ware doubtless displaced backwards witn the col- lapse of the ventral membrane against the.dorsal shell, — why not, there- fore, the metastoma as \7ell? (For evidence that the mouth could well have been in the lee of the hypostoma, v. Raymond (23, pp. 94-95, fig. 30). By alloTTing for this one might restore the mouth, still in front of the meta- stoma, but also in front of the distal tip of the hypostoma. Something of this sort is shown in Jaekel's restorations (18, fig. 28, p. 168; cf. also pi. VI) and in Richter's (30, fig. 16, p. 86)

- 30 - Richter (30), replying to Bornfs paper, begins by citing a number of genera, mostly Phacopidae, which show comparable markings: Chasmops, Pterygometopus, Phacops, Trimerocephalus, Dianops, Fhacopidella, Acaste, Asteropyge, DalmarJLtes, Dalmanitina, Cheirurus, Ceraurus, Cryptolithus, Solenopleura, and Scute Hum, Tlie last three are mentioned as not being strikingly comparable. He also restricts the discussion to the frontal lobe of the glabella, and the hypostoma, ruling out any other of Barrande's "impressions auxiliaires," but.recognizing the possibility that any other glabellar impressions of this nature may be connected with the stomach (as a matter of fact some of his examples - e.g. Ceraurus milleranus Miller and Gurley (24, pi. I, fig. 7) are of post-frontal markings")"! Great eni- phasis is laid on the "reversal of the impression." His figure 6, p» 80 (30) shows an internal cast of the head of Dianops anophthpiwiiR (Freeh) on which the glabellar furrows are raised ribs. The word "impression" is generally used by Richter to include both pustules and puncta, with their occasional linear extension! as ribs (keels) or grooves. In some forms there is but cne median impression, which may, as in Dalmanitina social is (Barraude), be a short groove instead of a circu- lar depression; in others, as in Dalmanites cristata Barrande and D_. hausmanni (Brongniart), there are three punctate impressions, outlining the angle as seen on, e.g., Chaamops. (Inese and others, up to five "fossettes," may be seen in Barrande's atlas (3, pi. 24-26)). Richter illustrates ex- amples not only of more than five impressions, but also of more than two rows, as on Acaste henni (Richter)and Crotalocephalus gibbus (Beyrich). The impressions need not be arranged in diverging rows, but may be scat- tered over the median sector of the frontal lobe, though in many cases, as in Acaste downingiae (Murchison), the impressions within the two main rows are not, as Richter admits, to be distinguished from the general sur- face granulation.^ So far all Richter's examples have been from the Proparia (he sug- gests some closer congenetic relationship "between the Cheiruridae and the Phacopidae on this account). However, he also cites Moberg's impressions _i_ and k_ (21, p. 296, fig. 1) on Nileus armadillo Dalman, as well as the lines on the glabella of Cryptolithus tessellatus Green, which Ruedemann (31, fig. 46, p. 147, and pi. 55, fig. 7) calls part of the "facial sutures" (and which Raymond (23) ignores in his discussion of Ruedemann's interpre- tation and elsewhere). Richter mentions also the nuchal tubercle of "some species of Scutellum, Nileus, etc." but considers the frontal impressions of the Phacopidae as sufficient for the present discussion, possibly in view of the great inclusiveness of the "etc.n He further notices that these aligned (or at least especially restricted) impressions are observable on animals with otherwise smooth tests and that as the surface of the glabella becomes roughened, the special impressions are lost. This might be con- strued to mean that he considered the general granulose sculpture of the shell to serve as attachment for a diffuse muscle web, whereas the dis- tinctly localized impressions indicated a more localized musculature, though he does not expressly say so. However, he avoids this imputation by follow- ing Born, who distinguishes the muscle scars by their independence of the "Durchporung der sonstigen Schale;" this distinction is supported by the

- 31 - nature of muscle insertions in the test of Recent macruran Decapoda. Richter's customary appeal to living forms leads him to a much. more grateful interpretation of these controversial markings, that of muscles rad- iating from the alimentary tract not only to the dorsal test, but also to the hypostoma. These, in Recent Crustacea, assist in persistalsis by opposing the action of annular (contracting muscles surrounding the canal (e.g., 10, p. 90, and pi. VT, fig. 6; this and other samples from Chun's "Atlantis" are of lepadid nauplii). Richter's fig. 9 (30, p. 83) is a sketch, after Chun (10, pi. V, fig. 7), showing that the only direct muscular (or ligamentous) connection between the dorsal test and the hypostoma attaches to the latter very near its proximal edge. Corroboration is found in Bernard's figure 13, page 56 (5), a diagram of the cephalic musculature of jjpus. Such an arrange- ment would be ample for the operation of the labrum, which after «n func- tioned chiefly, no doubt, as an operculum - - any grasping or holding would be performed by the paired appendages; and yet consider, for example, the power our biceps can apply to the forearm, where it exerts about the same a- mount of leverage. Richter does not, however, go into this, and instead calls attention to another drawing of Chun's (10, fig. 10, pi. 6) depicting on the dorsal surface of another lepadid nauplius a group of muscle scars roughly similar in plan to the frontal pattern of Chasmops, etc. Since these fasten the flexors of the tail and caudal spine such slight weight as they may have as 'testimony is in favor of Bornfs•interpretation. Bichter adequately explains the correspondence in ground plan of the frontal and hypostomal markings (as .on Chasmops) by pointing out'the similar dorsal and ventral silhouettes of the refle:ced alimentary canal. That is, from the ventral mouth the slender oesophagus widens forward Into the stomach, whereas dorsally the wide stomach narrows behind into .the intestine, this constriction being controlled, as he puts it,.by the mandibular muscles, as well as by the dorsal furrows, which may also have supported radial gastric muscles. Thus the similarity and correspondence of these impressions need not - could scarcely, under the circumstances - imply direct connection be- tween them, but merely a mirroring of their distribution to the walls of .the alimentary canal. Born noted considerable irregularity and occasional slight asynmetry in these scar patterns, even in a single species; Richter cites this as additional evidence that Hie muscles ran to a (relatively more vari- able) soft part rather than to a (more constant) member of the skeleton. It would appear, however, that one might interpret some, at least, of these muscles as hypostcmal; suroly one need not, like Richter in his diagram (30, fig. 16, p. 87), restrict them all to the radials. Indeed, one could, without overdoing, appropriate several .of the anterior of these scars to the levatores hypos tomas, for, as may be seen in Friederich Schmidt's figures (32, plates II and V), the "divergierende Punktreihen" are to a great extent on the steep fore part of the frontal lobe. Contrary to Richter's state-sent, there is in many cases considerable frontal space unoccupied by stomach, even in some of those nauplii to which he refers (vide 10, plo V-IX, XI-XIII). Bernard's sketch of' Apus (5, fig. 13, p. 56) shows hov? hypostoiaal muscles may run to the dorsal tost even though the stomach is far forward; they even seem to be somewhat deflected by the stomach. Moreover, would it not be reasonable to suggest that such prominent scars indicate the place of attachment of the

- 32 - more vigorous muscles rather than the more delicate peristaltic ones? How- ever, as Richter points out, Gerstaecker (15, p. 960) and Chun (10, p. 177) remark that not only may a muscle starting from one insertion spread into a more or less diffuse web and so reach to more than one attachment, but more than one muscle may be attached to one scar. That the glabeila had as its chief function the housing of the stomach and associated important organs, has long been the. natural belief of most students, Born being very nearly alone in his conclusions and quite so in his reasons therefor. Richter mentions BrBgger's observation (7, pp. 22 f.) that the glabeila and hypostoma constituted an admirable capsule for the reception of such vital organs, and further calls attention to the gibbous inferior sur- face of such doraally flat trilobites as Scutellum palifer (Barrande).* Although he agrees with most in repudiating the idea that swollen glabellae were gas bladders (esp. Deiphon and other bulbous Cheirurids), Hichter accepts the proposition that oil or lumps of fat may have been pres- ent (such being found often in Recent primitive Crustacea); since these are interpreted as being hydrostatically functional, the globules of oil or fat may have served to produce a buoyant effect. There is of course no basis for this speculation.. An interesting point is Richterfs absolute reversal of our generally accepted conception of the diet of the trilobites as inferred from their glabeliar volume. We customarily consider. the forms with large glabellae to have been given to mixed or even wholly vegetarian diet (e.g., Raymond, 23, pp. 81 and 103), while less expanded types have been considered to have been of carnivorous habits. Richter once more refers to Gerstaecker (14, pp.89 f.) to prove just the opposite; apparently Recent herbivorous Crustacea have a simple, elongate, undifferentiated alimentary canal (vegetarianism may, how- ever, lead to a widening of the tract if there is no possibility of length- ening it), whereas the carnivorous show a regional differentiation into a gizzai*d-like proventriculum and stomach proper, which are greatly expanded. Richter believes, moreover, that pronounced frontal muscle scars indicate a double stomach (apparently connecting the vigorous activity of the "gastric mill" with the exceptional development of the muscle insertions), and that this further confirms the conception that the Phacopidae, along with their cousins the Cheiruridae, were the most powerful and active.of trilobites. *By way of offerinc further evidence of the presence of the stomach in the glabeila Richter gives an original interpretation of one of Walcott's slices of Ceraurus pleurexanthemus Green (United States National Museum No. 68386, 38, pi. 104, fig. 5 - cf. Richter, 30,. figs. 13-14, p. 86) - an interpretation far more attractive than Walcott's own. Richter compares it with a'similar section through the head of a Recent lepadid nauplius (Claus, li, pi. Ill, fig. 11) which shows a strikingly similar T-section in the collapsed stomach, which might quite conceivably be preserved as a cast in a fossil.

- 33 - Locomotion ; The locomotion of thefc-ilobite may be considered under three main heads: benthonic, including any form of progression in or on a substratum; •nektonic; and planktonic. • Following Dollo (12) and von Staff aid Reck (35), themselves over- enthusiastic followers of Dollo, some rather glittering generalities on tril- obite locomotion have found their way into the literature. It was largely on this account that Richter wrote the first of a series of papers (27); in it he demonstrated some of the fallacies that.had gained widespread accept- ance, chiefly in relation to swinuiing; he touched upon sane of the benthonic aspects in the second (28). He complains that trilobites with pointed cephalic and pygidial shields have been grouped together as progressing by "Fortstacheln," i.e., by pushing themselves along -with -their spikes or points (forms with spinose pleura, etc., are also included). Thus Dalmanites limulurua (Green) has been described as progressing (cf. occurrences of Dalmanites hgusmanni (Brongniart) in the Devon- ian of Dworetz - vide 3, pi. 24, fig. 1; - and a more graceful individual (No. 1816) on exhibition in the Museum of Comparative Zoology) by fixing its caudal spine in the substratum and pushing with its body. This theory is allegedly (28, p. 24, and 35, p. 146) derived from Lockwood (20), but not rightly, for that author merely described the auxiliary use by the horse- shoe crab of its telson (which, incidentally, is movably articulated, as con- trasted with tho fixed spine of trilobites) in digging, where it aids by sup- plying a helpful purchase. Linmlus walks, or crawls, if you Till, normally by means of the legs. It is possible that trilobites with pointed or spinose pleura in the thorax, with the ends directed downwards, could progress some- what after the fashion of the snake, which may be said to walk upon the ends of its ribs, but it does seam over-elaborate, to conjure up such redes of progression when there are known appendages fully equal to the task of trans- portation. There is no dearth of other functions to assign to spines, which, as Richter says, must have been useful both as passive protection and hydro- dynamically. Richter rightly repudiates the interpretation of the greatly developed cephalic brims* on the Harpedidae and Trinucleidse as "mud-shoDS" or"aLime- skates." As he remarks, there could be no need for any such contrivance - he cites Leander, on its ten steeply inclined legs, walking easily on the softest ooze; the idea of a natatory animal.bogging in the sea floor does not seem very reasonable. Richter's chief point is stronger: s\rch an apparatus, used as a skid, would be more a hindrance than a help. In the first place the surface is decidedly non-skid, although the irregularities might readily pick up enough mud to make it slippery; moreover, it is only exceptionally of a form adapted for skiing. Consider especially Cryptolithus, whose brim slopes steeply downwards to the front as well as to the sides; it is true that the distal row of pores is on a more or less horizontal border, but this is narrow and is bounded by two abrupt carinae. This structure is comparable to the *This word seems better than the bodily transfer of Barrande's "limbe" which so easily makes for misunderstanding in 5inglish; cf. Bather.

- 34 - carapace of Linrulus, The situation in the case of Harpes is similar, though there are species in •which the xinder surface of the trim is approximately flat or convex downwards. The carinae — the Rand- and Z&npferleisten of Richter (26 and 29) — are much less obstructive than in Cryptolithus, and in some instances even turned up in front. Richter's profiles (28, p. 26, fig. 1, and 30, p. 208, fig. 3) are instructive, and suggest the parallel situation in Cryptolithus; figure _b is probably the nearest approach to a natural position, even for swimming, when the narrower posterior part of the body, furnished with smaller and weaker appendages, might well tend to droop, though a_ is likely if the animal had ploughed into the mud. Many of the specimens of this species (H_. macrocephalus Goldfuss) in the Museum of Com- parative Zoology are intermediate between a_ and b. As to real out-and-out burrowing, like moles or annelids, Richter is right in objecting that the trilobites were not particularly well adapted. Among Recent Crustacea the burrowers are distinguished by being cylindroid - at least less depressed than the rest of their kind, e.g., the Isopoda. He points out that real burrowers make tubular passages, whereas even the most nearly cylindrical of trilobites, and particularly those generally called fossorial, would make slit-like tunnels which vould readily collapse, so that even had any ever been made they would probably not have lasted long enough to be filled with mud or sand and so preserved. The Phacopid- Cheirurid type has the best cross section for burrowing, and Richter would not deny it the honors here any more than in other matters, as he has named it as the most powerful and versatile trilobite type. However, the append- ages of the Phacops described by Broili (8) are certainly not of the limuloid type of Cryptolithus. In the matter of the reflection of fossorial life in the external form, consider the burrowing crayfish of eastern. North America (Cambarus sp.)« This creature looks Just like any other non-burrowing crayfish; its tunnels extend for yards and culminate in the controversial "chimneys" (v., among others, 1, 2, 33, 34). Perhaps fossorial adaptation among the Crustacea is largely a matter of psychologyl Thus Richter is undoubtedly justified in saying that such trilobite burrowing as did occur was undoubtedly a sort of ploughing, like that of Limulus. (cf. Studer's remark on the similar habit of Serolis, quoted by Richardson, 25, p. viii.) Even at such shallow depths the Trinucleids with their long genal spines would have been in some difficulties when it came to turning (e.g., especially Cryptolithus goldfussi (Barrande)), for the at best only slightly flexible spines would have tended to keep the animal on its initial course. In any case really deep tunnelling is unlikely (cf. also 23, p. 103). As Raymond (loc. cit.) and others have pointed out, it is probable that many trilobites habitually covered themselves thinly with loose sediment by way of concealment, as do shrimps and many other Recent crustaceans; but how an animal so buried as were the Vogdesias at Elgin, Iowa (13, and 23, p. 102) could spring from ambush is not clear. In connection with this surficial ploughing one might notice some of the apparent digging tools mentioned and figured by Richter (28, pp. 41 f.' figs. 9-13); they include forward-directed points on the hypostoma,

- 35 - Euproetus cuvieri (Steininger)), epistoma (Homalonotus hippocampus Schwarz), and ploughshares and denticulate brimlets on the cephala of some Phacopidae. There has been some controversy over the swimming attitude of tril- obites - whether they swam with the back up or down. This undoubtedly varied in different types and at different times. Richter (27, p. 224) says about all that there is to say about this: such Recent Crustacea as exhibit this ability (e.g., the Apodidae, the isopod Dynamene, etc.) are generally equal- ly proficient in either position. It is probably largely a matter of what the animal is doing at the moment; if in search of food near the surface it may swim upside down (Raymond, 23, p. 99) The most likely assumption about the swimming methods of a trilobite is that they are not much unlike those in use by tire most trilobitoid of living animals. These are in general found among the Isopoda, and it ap- pears, both from Richtor (27) and other sources (I have not been able to observe any personally) that these customarily swim by means of the legs, with the body extended. Many, if not most, have broadened or setiferous swimming feet, or both. The trilobite exopodites approach the various forms of Recent crustacean pleopods most nearly, - much more so than any known form of trilobite endopodite. For this reason nearly all authors, except Raymond (23), interpret the exopoditcs as the trilobite's swimming feet. In support of Raymond's branchial explanation may be mentioned these same Isopoda, among which "the pleopoda are in some cases natatory, but their function is respir- atory for the most part" (25, pp. 2 f.). Of the six orders of isopods only one has the pleopods entirely natatory, whereas of the three in which they are entirely branchial one has them fitted for air-breathing. A member of this terrestrial group, Oniscus asellus Linne,' is, however, mentioned by Caiman (9, p. 201) as using the endopodites in breathing. Raymond has said, though unfortunately without supporting analogy in living forms, that trilobites with wide axial lobes probably swam forward by vertical undulation of the entire body (the fish's method rotated laterally through ninety degrees), though in his memoir of 1920 (23, p. 99) he speaks of this as an incident in backward swimming by means, primarily, of downward strokes of the pygidium. Once Richter suggests such wriggling, but as an incident to forward swimming primarily by means of the legs, in the case of Harpes (29, p. 211). This same author (27, pp. 221 f.) implies that Dollo (12) pictured Deiphon swimming in what Richter believed to be whale-fashion, i»e,, "by sculling or screwing with the pygidium" ("...durch Wricken Oder Schrauben des Schwanzschildes"). Dollo did emphasize the similarity in ap- pearance between the pygidium of Deiphon and the cetacean tail, but did not explicitly say that the trilobite suem in this manner. Neither, however, does the whale. There is in the Cetacea no such caudal musculature as that to which Richter refers (27, p. 222). In the four longitudinal reentrants of the posterior part of the vertebral column lie four muscles, the upper pair of which (cf» the trilobite's extensors) lifts the tail, and the lower (cf » the flexors) pulls it down. A very gentle and measured movement of the tail in this manner can propel the animal at a respectable speed*, as many ocean travellers must have observed in porpoises. Deiphon seems to have had room enough for a musculature adequate to flap his spare pygidium, but it is unlikely that it would be a highly efficient propeller. * Dr. G. M. Allen - verbal communication.

- 36 - Richter's objections to habitual backward swinciing by means of py- gidial slaps are for the 'most part well taken, particularly in view of such publications as v. Staff and Reek's (35). He doubts the suitability of the cephalic and pygidial shields for such swimming (in many trilobites, except those with long thorax and strong longitudinal muscles, the cephalon would of necessity perform a corresponding arc when the pygidium was abruptly flexed). It may be that these shields would not have been strong enough to stand the stress of such use; it is certain, though, that they would, as he says, be too stiff in recovery; and there is no way in which they could be feathered, like the telson of Recent macrurans. Thus, as their concavity is in many cases such (particularly would this be true in foms with a ra- chis so deeply arched as to admit of strong longitudinal muscles) that the inertia of the water they would catch and hold would brake their rearward impetus rather strongly. Richter recognizes the advantages of moderate spooning, as in oars. His contention that such forms as the Illaenidae and Phacops are scarcely suited, when enrolled, to any motion save that compelled by gravitation will readily be admitted, and the lack of any sort of a di- rective drag, such as is provided by the rigid cephalothorax and extended chelae of Homarus and its kind will readily be admitted. Richter also doubts that trilobites had adequate musculature for such methods of swimming, even those with a prominent rachis. The alimen- tary canal and the muscles of the appendages occupy too much room, and the sub-pleural spaces are too thin. Furthermore, he denies the existence on the pygidium of such muscle scars as have been described, crediting only those on the cephalon, hypostoma, and thorax. By no means isolated examples of pygidial muscle insertions are given by Hall and Clarke (17, pis. VIII A, fig. 15, XXV, fig. 8, and pp. 103 f.) and Walcott (37, pi. IV, fig. 5, and p.222). (It is true that this example of Walcott's is of extensor muscles - vide 23, p. 92). Richter also remarks on the flattened tests of those tril- obites with the largest pygidia, citing Scute Hum and the Asaphidae - saying that their musculature would be most inadequate. This does not by any means apply to all Asaphidae, and one might mention Scutellum campanifer (Beyrich) and S_. brongniarti (Barrande) as counter-examples. It has been suggested by Raymond that width even.without much height would insure sufficient space for longitudinal muscles. The volume of the rachis is doubtless more impor- tant than extension in any one dimension - either height or breadth - alone. Richter appears much too exacting about muscles; he credits thp muscles of the appendages with usurping an extreme amount of space in the axial lobe. Richter also lays great emphasis on the absence of a differentiated "Polsegment" in the thorax, i.e., the prominence of that segment which would form the cut-water in back-flipping. Few even of the Decapoda show such a feature. Some, as, for example, Palaemonetes, have a marked kink in the ab- domen, but the double imbrication referred to by Richter is a special, but not too uncommon, feature of the pleura and does not occur in the tergum of the segment in question. There is an apparent reverse imbrication in the rachis at the joint between the carapace and the first abdominal segment, but this is merely a deceptive appearance. Richter's point is that all Recent aquatic arthropods at all like

- 37 - the trilobites swim forward with the body extended, paddling with the legs. This sound rather inclusive> but is essentially true. His reasons, however, as shown above, are not all of the best. Among living Crustacea back-flipping appears to be known only in horaaroid animals. Von Staff and Reck mentioned esnong their criteria for a retrogressive- ly swimming trilobite eyes stalked or otherwise adapted for rearward vision. Other explanations may, however, be offered. For example, Raymond has sug- gested (23, p. 102) that Vogdesia vigilans was equipped for alertness in am- bush, and supported this with Finch's account of the animal's occurrence (13). A further development of this is to be noted in such forms as Asaphus kowaleskii Salter. The stalked eyes of sane of the spinose trilobites, as in the 0donto- pleuridae and Lichadidae, may represent a convergent line of adaptation. As these floated among their miscellaneous planktonic associates bits of seaweed and the like would become attached to them, probably serving well as camouflage. Thus the produced eyes would be used as probably by Asaphus kowaleskii, but might also be a necessity to enable the animal to see out of its dress of drift. Richter has made a somewhat similar suggestion in one of his papers. He is un- doubtedly right in declaring that stalked eyes are not a dysphotic adaptation - what good could they possibly do there? - but there is no necessity for call- ing them balancing organs (27, p» 233). An interesting feature of these eyes is their. forward raking, as if to make it easier to see directly ahead; even such noimal eyes as those of Asaphus expansus are directed very much down and to the front. Might not this be taken as evidence that the animals habitual- ly moved forward? Trying.to judge from the eyes alone, or from any other one or few features, is likely to lead astray. This is particularly well shown by Dollo's Paloontologie Ethologique (12). For example, Apus has dorsal eyes, and Richter says it likes to burrow, but is nevertheless no mean swimmer. Dollo does seem open to the accusation of having judged the arthropods by the fish. So also Richter when he fears that a bluff glabella might tend to coun- teract, by its resistance as a cutwater, the power indicated by the deep ra- chis; he himself has frequently called attention to the expert swimming of such seemingly.inept fonas as some of the Hydrachnidae. As to the weight of the test keeping the trilobites most of the time on the bottom, it seems no more likely than in the case of Recent Crustacea. I do not see that any generalization can be made about this, except that some doubtless were more completely benthonic than others, depending on a variety of factors, such as, indeed, weight of shell, but more immediately on their positive adaptations for swimming, such as musculature and nature of the ap- pendages, general shape, including such details as the disposition and nature of the eyes.* * I certainly see no point to the assumption (27, p« 231 - after Jaekel) that the Agnostidae, even if their shields do balance and are relatively very large, should have had to swim or float because the weight of their shell was too great for their "few" legs. Even if minimum values are taken, they would still have had from twenty to twenty-four endopodites, which should be sufficient for crawling (the Isopoda get along with fourteen)/ In any case, however, regardless of the number of legs', why should a shield consisting of fused segments weigh significantly more than the same number of free seg- ments?

- 38 - What Richter says about vising the dorsal shell as an index of na- tatorial ability is very much to the point: We know practically nothing about the relative development of the appendages on trilobites in general. Their size is surely more important ia this matter than variations in the design of the rest of the body, whose ill-adaptation they might mitigate or overbalance. Richter believes that the flmbriated exopodites were pleopods. On this account he assigns to Triarthrus and Cryptolithus, which he would not call nektonic from their shells, strong swimming ability. The chief point has been stated: the index-value of the dorsal shell is modified by the de- velopment of other parts of the body, chiefly the appendages and their nus- culature; but accoaodation for the viscera is of course also important. However, when he comes to naming his choice of the most active swim- mer he reaches practically the seme conclusion as Raymond (23, p. 103), who names Isotelus and Dalmanites, whereas Richter selects his "Pha.cops-type." Almost the only point of coincidence of the conceptions of criteria for swimming trilobites of Raymond and Richter is the agreement that large subequal shields indicate such habits. Whereas the former considers the py- gidium as a propeller, Richter thinks of the two shields as supporting planes, balancing one another. He also appeals to spines in the horizontal plane for the same purpose; thus the genal spines of the Trinucleids would carry the weight of the head far back and so balance it properly. The same applies to Harpes, which he considers almost exclusively nektonic, not seeing how it could do anything on the bottom except rest - because of the discrepancy in levels noticed above (and vide 28, fig. 1, p. 26, and 29, fig. 3, p. 208). Richter refers to Woltereck (39), but appears to have missed his point, for Woltereck declares that the spines are primarily directive - immediately mechanically and secondarily photostatically. Gurney (16, p. 462) mentions the great directive influence of the median spine of the Corystos zoaea, adding that spines (median) seem to be important balancing organs, which disappear as the "auditory" organs develop to efficiency. Richter mentions the raphiophorids as one of the types in which the head was supported in swimming by the long genal spines (vide above). The fron- tal spine, as of. Ampyx, would not do this in just the sane way, but would undoubtedly help keep the head up in forward swimming. Woltereek's work (39) on various kinds of Daphnia (esp. Hyalodaphnia) indicates that such a spine has great influence on both attitude and direction in swimming. Richter may be right in saying that no trilobite was entirely plank- tonic, i.e., could hang motionless in the water, without sinking, but seme of them probably made a fairly close approach to it. He speaks of Recent spinose plankton with exceptionally complicated appendages which may equal or surpass the ramifications of the test, and he would e:rpect similar devel- opments in the Odontopleuridae. Such appendages would, of course, be a nuisance to a benthonic animal, but would be of advantage off the bottom. By way of proof that trilobites were not genuine plankton he offers the testimony that their enrolled shells show that they sought refuge "when in trouble, i.e., frequently" by folding up and sinking to the bottom. Even if they were not entirely planktonic, the Odontopleuridae approached that con- dition very closely, and their locomotor apparatus may have been somevhat modified.

- 39 - With Richter's summary of trilobite locomotion we have very little quarrel - much less than with some of the evidence on which this sunroary is apparently based. He considers the trilobites as primarily benthonic, but also able to swim by paddling with the legs. Enrollment he looks upon as a reflex which was manifested in any emergency, no matter, naturally enough, what the animal was doing. But he makes no mention of burrowing, even of the ploughing kind, which must have been quite common. He compares trilo- bites to Serolis, as crawling and swimming vith equal ease; to this I inould add that they probably, in great measure, also burrowed to about as great degree as that •* isopod.* Enrollment Enrollment is almost the first thought when one is asked how a tril- obite could defend itself against attack.** This is because of numerous finds of the animals so disposed, as well as upon analogy with living isopods. In nearly all instances enrolled trilobites have been found to belong to fonas with the ends of the pleura facetted so that they slide easily upon one another and make a close-fitting series of imbricated joints ifcten the animal is flexed. Trilobites without such bevelling are not, as a rule, found en- rolled, though Ceraurus, for exanple, has been so collected. In this country it has been customary to speak of the facetted forms as capable of enrollment whereas this ability was implicitly or even explicitly denied to the rest. They have not been so restrictive abroad, however, and Richter (28) reflects this when he asserts that all trilobites could enroll; he cites Pompeckj*** as his authority as he continues to the effect that the friction of the slight- ly confined facetted pleura upon one another would keep the animal in enroll- ment even after the relaxation of the flexors (i.e., even in death, after the decay of the viscera), whereas the unfacetted fonns would unroll as soon as the flexors ceased straining. Pompeckj made a cardboard model of Para- doxides and found that it would enroll quite well, though not so perfectly as, for example Calymene or Asaphus. (M. Neumayr appears to have appealed to natural selection, and maintained that the rise of predaceous foes, such as fishes and cephalopods, induced an increase in the distribution of facetted pleura by putting a premium on ability to remain a long time enrolled). I have examined a number of Recent isopods (mostly Sphaeromidae) in search of facetted pleura and information on enrollment. The majority wore enrolled, and very finaly, so that attempts to unroll them or to enroll ex- tended individuals were abandoned for fear of injuring them (they were alco- holic specimens); intone of these did I observe anything like pleural facets. * See above, and 25, p, viii. ** Some have gone rather to extremes ia this. Jaekel, for example, supposed alarmed Agnostidae to enroll with such vigor that they drove -themselves deep into mud - out of sight! To Richter this does not, apparently, seem particu- larly far fetched, though he objects that they were so light that they prob- ably merely stuck in the surface of the mud (28, p. 32). Further, Richter scons to look upon enrollment as one of the trilobite's most important acts. He even (28, pp.29-30) refusea to consider enrollable trilobitos as burrowing, on the ground that they could not then enroll; he ap- parently did not believe that their need for enrollment was much diminished in such circumstances. *** Jahreschefte Ver, vaterl. Naturk.Wurtt.,XLVIII (1892), pp. 93-101.

- 40 - Thus, although facetted pleura may well have been of use in enroll- ment, they were evidently by no means indispensable. Although the accepted explanations appear to fit the facts very conveniently, they should undoubt- edly be modified. The general assumption has been that such passive defense as en- ro!3ment and spines afforded was about all ttiat trilobites were capable of; similarly they have been considered chiefly as scavengers, and portray- als of predaceous habits in trilobites have been repudiated (e.g., 23, p. 103). Richter, however, mentions Apus as successfully attacking active prey (28, p» 40, footnote; he mentions Branchipus, tadpoles, and annelids). Even this testimony is scarcely necessary to allow of a trilobite catching at least some living food, particularly if it be granted that 1ire creature could dart about suddenly.

- 41 - .. BIBLIOGRAPHY . . • I 1. Abbott, C. C. Are the "chimneys" of burrowing crayfish designed? Am* Naturalist, vol. XVTII (1884), pp. 1157-1158. 2. How the burrowing crayfish works; Inland Monthly, 1, 2, Columbia, Ohio, Feb. 1885, pp. 31-32. 3. Barrande, J. Syst. Sil., I, Planches, Prague (185S). 4. Beecher, C. E. Further obs. ventral struct. Triarthus; Am. Geol., vol. XV (1895),;pp. 91-100, pi. 45. I • • 5. Bernard, H.M. The Apodidae, London (1892). • t 6. Bornfc A. . Zur Organis. d. Tril. I: Das Kopfschild von Chasmops Odini. Senckenbergiana I (1919), pp. 159-171. 7. BrBgger, W.C. TJeber d. Ausbildung d. Hypostoms ... skandinav. Asaphiden. K. Svensk. Vet.-Akad. Handl. XI (1886), Stockholm, pp.1-78. * * •* - 8. Broili, J. Trilobit. mit Gliedmassen aus d. Uhterdevon d. Rheinprovinz. Sitzber. bayr. Akad, Wiss.(1928), pp. 71-81. 9. Caiman, W. T. Life of the Crustacea. London (1911). 10. Chun, C. Atlantis (Biol. Stud. ttb. pelag. Organismen). Bibliotheca Zoologica, XDC (1896), Stuttgart. 11. Claus, C. • Medianauge d. Crustacean; Arb. zool. Inst. Wien,' .. DC (1891). 12» Dollo^ L. Paleontologie e'thologique. Bull. Soc. beige de Geol., Paleontol., Hydrol., Mem. XXIII (1929), pp. 377-421. 13. Finch, G. E. Notes on position of individuals...etc. Proc. Iowa Acad. Sci. for 1903, vol. XI (1904), pp. 179-181, pi. 14. 14. .Gerstaecker, A. Klassen u» Ordnungen d. Arthropoden, [5J 1. Crust. (1. Hfilfte). Leipzig - Heidelberg, 1866-1879. 15. Gerstaecker, A. Klassen u. Ordnungen d. Arthropoden, von A. E. Ortmann - (2. HSlfte) Leipzig (1901). 16. Gurney, R. Metamorphosis of Corystes cassivelaunus; Q. J. Mic. Soc., vol. XLVI (1902), 3, pp. 461-478, pi. 17. Hall, J., and Clarke, J. M. Paleontology of New York, VIII, Albany (1888). 18. Jaekel, 0. . Ueber d. Organis. d. Tril..I. Ztschr. deutsch. geol. Ges., vol. LIU (1901), pp. 133-171, pi. 4-6.

- 42 - 19 - Jackson, R. T. Phylogeny of Pelcypoda. Mem. Boston Soc., N.H., vol. 17 (1890), 8, pp. 277-400, pi. 23-30. 20. Lockwood, S. • Horse Foot Crab-. Am. Nat. (1870), IV, pp. 257- 274, pi. 3. i 21. Moberg, J". C. Bidrag till •kannedomen am tril. byggnad. Geol. FOr i Stockholm, FOrh., vol. XXIV (1902), pp.295-302. 22. Packard, A. Limulus polyphemus. Anniv. Man. Boston Soc. N.H.(1880). 23. Raymond, P. E. Appendages, Anatomy, end Relationships of Trilobites; Mean. Conn. Acad. Arts and Sci., vol. VTI (1920). 24. Raymond, P. E. and Barton, D. C. Revis. Am. spp. Ceraurus. Bull. M.C.Z. vol. LI7 (1913), 20, pp. 525-543, plates. 25. Richardson, Harriet. Monoft. Isopods N. Am.; Bull. U.S.N.U., vol. LIV (1905). 26. Richter, Rudolf. Neue Beob. tlb. d. Bau d. Tril. - Harpes. Zool. Anz, vol. XLV (1914), Leipzig, pp. 146-152. 27. — Vom Bau u. Leben d. Tril. I (Schwirmen) - Senckenbergiana, vol. I (1919), 6, pp. 213-838. 28. ——- Vom Bau u.- Leben d. Tril. II (Aufenthalt a. d. Boden etc. - Senckenbergiana, vol. II (1920), 1, pp. 23-43. 29. Richter, Rudolf.' Beitrage z....devon.Tril. - Harpes; Abh. Senckcnberg. Naturf. Cos., vol. XXXVII (1921), pp.177-218, pi.16-17. 30. —-— - Vom Bau u. Leben d. Tril. Ill (Glatze u. Magen); Pol- aeontol. Hungarica, I, (1923), Budapest, pp. 77-89. 31. Ruodenann, Rudolf. Cephalic Suture Lines of Cryptolithus; Bull. N.Y. St. Mus., 189 (1916), pp. 144-148. 32. Schmidt, F.' Revision d. Ostbnlt. Sil. Tril., I. Mem. Acad.' Imp. Sci. St. Pot. (7) XXX-(1881), i., St. Petcrsbourg, pp. 1-238, pis. 33. Shufeldt, R. Chimneys of burrowing crayfish,' Observer, vol. VII (1896) pp. 85-89. * ' 34. ——, Some observations on crayfish; Shooting and Fishing, vol. XXV (1898), p. 227. 35. v. Staff, H., and Reck, H. Ueb. d. Lebensweise d. Tril.; Sitzber. Ges. naturf. Freunde, Berlin (1911), pp. 130-146.

- 43 - 36. Stelninger, J. Geognost. Beschreibung der Eifel. Trier (1853). 37. Walcott, C. D. The Trilobite, Hew and old evidence relating to its organization; Bull. M.C.Z., VTII (1881), pp. 191- 2S4, pis. 38. —— Notes on »»» Neolenus; Smiths. Misc. Coll. LJOTEI (1921), pp. 365-456, pi. 91-105. 39. Woltereck, R. TJeber... sog. "Schwebe - Forstatze" ... Zoologica, XXVI (1913), 67, pp. 476-550.