Click for next page ( 152


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 151
CHAPTER X DETERMINATION OF " g " BY MEANS OF THE FREE STINGING PENDULUM C. H. SWICK U. S. Coast and Geodetic Hi very INT110DUCTION The at.tractio~.~. oil the earth upon a unit mass on its surface varies from place to place. This -['act was accidentally discovered in 1671 by the French astronomer, Jean Richer, who found that his astronomical clocl<, which had been care:t'ully regulated in Paris, lost 2~ minutes a day on the Island of' Cayenne, South America, and required a shortening, of its pendulum o-l' more than 2 millimeters to malice it keep correct time at the latter place. Although this variation caused a great deal of comment and speculation amon`, the scientists of that time, the first hint of the correct explanation was given in Newton's Principia, published sixteen years later. In that great work it was stated that the decrease in gravity between Paris and Cayenne Night be due to the centrifugal force of tile earth's rotation and to a slight bulging, of the earth at the equator. At the time o:l' Richer's discovery the pendulum clock was a compara- ti.vely new device. It had been inventecl only about fifteen years earlier by Christians Huy~,ens, tl~e great Dutch astronomer and mathematician. 1-luygens in his astronomical observations had been confronted with the need for more accurate timepieces than were then available He Recense interested in the possibility of Usi.ll~, a pendulum as. a clock re~,ula.tor and -finally invented the pendulr~ elm. It is quite remarkable that in nearly three centuries since T-l:u~'s i.r,.vention no more accurate device for regulating ~ timepiece leas ever been discoverecl. than the pendulum. It is equally remarkable that during the two centuries since the beginning of gravity determinations the pendulum has been the most accurate device known for measuring the intensity of gravity. While worl~in`~ on his invention Huy~ens. made an exhaustive mathe- matical study of the pendulum and discovered many of the underlying, principles which made possible its. use later :t'or gravity measurements. He discovered, for example, that the period of a free swinging pendulum varies as the square of its. length within certain limitations. However,. :J'-or a compound pendulum (a simple pendulum is possible in theory only) he found that. the length must be tolled as the distance between the center off:' suspension and the ce~:~.ter oft' oscillation. 151

OCR for page 151
1 ,32 FIGURE OF THE EARTH The center of oscillation he found to be that point in a compound pendulum where a, particle moves of its own accord at the rate forced upon the other parts. of the pendulum. He found, in other words, that the particle at the center of oscillation moves at exactly its own free rate and so tends neither to retard nor accelerate the rest of' the pendulum. Those parts of the pendulum which are near the center of suspension tend to move faster than those farther away and so force the latter to move more rapidly than they otherwise would, and vice versa. The distance of this particle at the center of oscillation from the center of' suspension is exactly equal to the length of a,theoret;.cal simple pe~:llllun~ ha,v-in', tl~e same period of oscillation. ABSOLUTE DETERMINATIONS OF GRAVITY The first attempts to measure the intensity of' gravity were all made with some i'orm of compound pendulum which approximated as closely I, ., as possible a theoretical s:~mpie penctu~um. Henry Inter was the first to depart I'rom this principle. He applied to gravity measurements Huygen's theorem that the centers of suspension and oscillation of a. con~- pouncl pendulum are reciprocal if at unequal distances from the center of `,ravitv. He made a pendulum which had two Loire edges and two sliding weights on the stem. By swinging, this pendulum first in the usual man- ner with the bob clown and then on the other l~nil:'e Aloe with the boll up he was able to adjust the weights until the period of the pendulum oaf orate l~nif'e edge was practically the same as its period on the other l~ni-fe ecl~,e. The distance between the two l~ni-le eclges, after c.eriain corrections had been appliecl, was then equivalent to the length of' a. theoretical simple pendulum having, the since period, end the intensity of' :,ravitv could be computed groin this le~:~,t,l-~ and the observed period of:' the penclulum. Improved types of the reversible pendulum were later developed as, -for example, the Repsolcl pe~dulu~n Lehigh by means oil-' a hollow shell at the opposite end from the bob was made syn~netrical at the two ends so that the air resistance woulcl be the singe wl:~eu the penclulu~n was swells on either knife edge. Although the reversible pendulum is the most precise means Anon for determining, the absolute i.ntensit.v of ~,ravitv, it has certain inherent limitations. that impair its accuracy. For example, the distance between the l~nil'e ecl~t,es is very Difficult to measure with sufficient exactness. By ma.ki~l~ this measurement with the }pendulum hanging, on a. l~n;i'e edge the stretching deice to its own ve;~ht is included in the remeasured length. To talkie account o-l' the stretching, clue to tile centrifu~,a.1 force of its vibra- tory motion, however, is much more clif~cult and uncertain. )Tan,v other small inaccuracies enter into the absolute determination of' gravity with the reversible pendulum.

OCR for page 151
DETER,lII.VATION OF " 9 " B Y 1WEA,YS OF PENDVLU,llI 1.33 REL NTIVE DETERMINATIONS OF GRAVITY In contrast to absolute determinations it is comparatively easy to mea- sure the difference ire gravity between. two places with very great preci- sion. It is only necessary to swing an invariable pendulum under the same conditions at both places. and to determine its. change of period. I-i' one of the places at which the pendulum is swung is a base station or some other station at which the intensity of gravity has been previously determined, the value of gravity at the new station can be readily com- putecl by the formula Pal, A,, .~8 = p2 in which g is the intensity of gravity and P is the period of the pendulum the subscripts s and b referring to the new and base stations, respectively. Since it is difficult, or practically impossible, to make the conditions under which a pendulum is swung, exactly the same at two different sta- tions, certain corrections must be applied to the observations in order to talkie care o-! the differences. What is usually done is. to reduce the obser- vations at both the base station and the new station to certain standard conditions., corrections being applied for temperature, pressure, a.~npli- tude of vibration, and flexure. All modern precise gravity detern~;nations are nude by the relative method, and practi.callv all o:L' them are based on the absolute value of gravity at Potsdam, Germ..~n>. The absolute value at Potsdam is' the result of a lone, series of' very care.-l'ul n~ea.surements extending over manic years and is probably as accurate as it is possible to obtain. The preci- sio~ of' this determination, however, is. not so important as it Piglet seem. The important thing :is to have a.llthe ~-ra.vity measurements in the world leased on the same system, tl:~at is, referred to the ogle base sta.- tics-. '1'1~is -is especially essential. in fi-u-~:e-oi'-the-ea.rtl-~ investiga.tio~-~s where the variation of' ,r..~vitv ~.i.tl~ hate of latitude is tile ~->rincipal rector to lie considerecl in clerivin~, the shape of' the eartl~. Nearly every country engaged in measuring the intensity of ~t,ravitv has its own base station whose value has been cleterm;~ed with more or less care by the re]a.tive method :t'ro~n Potsclan~. or groin the previously established base station of tonne other coun.tr\r. All relative gravity ~neas- ~rements in ~ country are usually referred. to its own base station. The periocls of the pendulums are detern~;ne~l there T~el'ore anal after each Held season. not only to obtain the l',, for use in the l~receLlin~, I'orn~ula but to make sure that. the pe~clulu~ns super no cha.~',e o-t le~-~th d-uri~:, their use ifs the -held.

OCR for page 151
~4 FIGURE OF THE EARTH The first pendulums used for relative determinations were patterned more or less a.f'ter the pendulums used for absolute determinations. They were a meter or more in length, heavy and cumbersome to handle, and were much inferior to the modern apparatus in accuracy. MODERN GRAVITY APPARATUS There are two general types of modern gravity apparatus for use on land. The older of these two types is 1 OCR for page 151
0\ O~ ~ ~y t~ # 0F ~155 flab ~p~~US Tntedcrometer Two chronometers Thermometers; manometers; etc. [IG. 1.~0DiulUulS of the Coast and Geodetic Survey. A ~ the dummy pendulum gab ~ thermometer ~tt~cbed to as ~em. ~ is one of the principal pendulums us it appears Ruben Ida OD the knife edge. C small ~uxiL~ry pendulum carrying ~ levy vial Ahab is used in 1CVeliD/tbC knife edge. D add ~ are the other priLcip~1 pendulums. Tbc pendulums are of Ample Construction. (See Digure 1.) Each con- ~sts of ~ thin stem With ~ heavy tot at one end and an inverted stTr~l~ at the otbe~. ~ piece of agate carefully ground to sat optic~1 pl~ne.on its lower side is secured in the stirrup in surb ~ position that file plane. is at right Andes to the vertic~1 axis of the peninlum. A1I pats of Me pen- dUln~ ate rigidly ~tt~cbed to egg other t~ means of closely fitted joints

OCR for page 151
FI GLiRE OF THE EAR TH aIld rickets to prevent ally rha.r~',e ;~ the length of' the pendulum under ordinary usable, except the change due to temperature variations. The placing oft' a Platte instead. of' the usual l~nit'e echo in the head of the pendu~un~ ma: seem at first thought to be a.n unnecessary reversal of position of these i.mporta.~t parts of' the apparatus. As originally con- structed~ each pendulum bad a. L:nit'e edge in the head and swung on. a plane attacher!. to a shelf' in the receiver. After the apparatus had been usecl. a short twinge in the Delco it was feared that the knife edges might eventually wear appreciably and cause a. lengthening of the pendulums and so the planes and kDi.-i'e eclges were reversed in position. A:lter more than thirty years encl. considerable use the l~r!,i1'e edges now show no indi.- c~a,tion of wear anal there is some doubt as to whether this change was ~,ecessarv or clesirable. NVi.th the present arrange~ner~t. it is quite essen- tial to center the pendulum carefully ol1 the lignite edge, but if the Loire ecl~,e were ifs the pearl of'' the pendulum the position of the peIlclulum on the support woulcl slot be so important. The crummy pendulum its similar to the other pendulums in all parts but the head which lies a. different form-to prevent the dummy from talking, up sympathetic vibrations with the swinging pendulum. The purpose of the dummy pendulum is to carry a thermometer in contact with its stem and thus malice possible a close approximation to the tem- perature of the s,wingi.n~ penclulum. The latter, of course, could not be made to carry a thermometer without causing a. change of virtual length which would destroy its ac~cura.c~r. As the dummy is made of the saline material and l~a.s. pract;.cally- the same shape as the swinging pendulum there is every reason tc believe that it acquires almost exactly the same temperature as tl.~e latter cluring the observations, especially as the two are only a i'ew centimeters apart in the receiver and are usually placed ir position several hours. before observations are begun. The knife edge on which the pendulums are swung is a most impor- tant part of the gravity apparatus L;,l~e the plane in the head of the pendulum, it is made of' agate. It is rigidly attached to a, brass plate which is secured by a. screw to a, shelf in the receiver. The knife edge is surround to give as close an approximation as possible to an ideal kni:f'e edge, a geometrical line. The most nearly perfect knife edge that can be made will be found when tested to have what has been called the miss- ing triangle, that is, the two planes forming the edge will not quite meet. The base of this missing triangle on the knife edges of the Coast and Geodetic Survey'~,ravity apparatus has been found by optical means to be less than one micron in length. As the pendulums swing through a very small arc, the two planes which Fleet to form the l~n;i'e edge are ground at an angle of 13(~ with each

OCR for page 151
DETERlIINATION OF ",J'' BY MEANS OF PENDULUM 106 other. This Dies an ecl'~,e strolls a,.~.cl rigid a.~cl not Basil: i~-~jurecl. In order to prevent any tendency for the pendulum to rock ice a. d;.recti.o~ parallel to the l~n.il'e edge, the ~niclclle third of the edge is cut away. The flash apparatus consists of' a box oil. a stanfl with an obse.rv;.~, telescope mounted above it. The box contains an electro-mag~et. which is connected to a, shutter ire the side of the box towarcl, the per~dulum receiver. It also contains a small electric light which is di.rectl~ bestial the shutter. NYhen the electro-magnet is placed in circuit with ~ break- circuit chronometer it will open the shutter :E'or an instant at each break of the chronometer and thus pern~.it a beam of li.gr,ht to emerge al; one-second intervals. This beam of' light is clirected through a, windows in the :t'ront of the receiver onto; two small mirrors, one attached to the l~nife-edge plate a,ncl the other to the swinging pendulum. From these mirrors the beam is reflected in two parts baffle into the observing tele- scope. The coincidence observations, used in obtaining the period of' the pendulum in relation to the chronometer, are made by noting the times when the two reHect.ed parts of the beam are in the same horizontal ]i-~e in the telescope, as explained in more detail later. The interferometer (see :Figure 2) needs no particular description. It is simply a, special form of the Michelson interferometer adapted-to the purpose of measuring, the flexure or horizontal movement of the kni:t'e edge caused by the horizontal component of the stresses induced by the swinging, pendulum. It hats a strong source of light of Knowles wave-len`~th a.ncl two accurately ground plane-glass reflectors, one on the interferometer itself' and the other attached by means of a, light aluminum arm to the l~nife-edge plate in the receiver. When the two reflecting surfaces have been placed close together and a.djusl;ed to a- slight angle to give inter- ference fringes, any movement of the knife edge will cause a, siclewise shifting, of the fringes. By determining the amount of this, shifting in terms of the width of one fringe it is possible to obtain the corresponding movement of the knife edge within ~ few thousandths of a micron, and -T'rom this the correction to be applied to the period of the penclulun~. GRAVITY ])]3TERMINATIONS Ire all modern cletermir~atio~ls of' ,ra.vity on la.ncl dI1 accuracy of' one or two parts ill a nonillion is desired and usually attained. As the average intensity or acceleration of `~,ra.vit~: is about 980 cm/see" this accuracy corresponds to an error of about 0.001 or 0.002 cm/see" in the total in- tensity. To attain this accuracy it is necessary to determine the a,vera~e~ Period o:l' the pendulum within less than a millionth of a. second of time. It is :t'ai.rly cl.itficult to malice any physical measurement, within one part ice a million lint an interval of' time is one of:' the most difficult of all to

OCR for page 151
~ ~- ~e as 100 , . FIC~ 0F r~ ~6 T~ ~S~:ES~:E~ESE:E.E:ESE:ESESES~:E:ESE:ESE:ESESESESE:E.E.ESESE.E:E:ESE:E.ESE.E.E.E:ESE FIG. 2.-[eDdulum receiver bud interferometer, Coast and Geodetic Survey upp~utus. Tbe extremely small unmount by ~bicb 1be receiver is swayed by the osculations of the pendldum is measured limb the interferometer in terms of Me ~aved~D~ of barium light.

OCR for page 151
DETERMINATION OF "I" BY MEANS OF PENDULUM 159 measure with such a high degree of' precision. Fortunately, a free swing- ing pendulum will maintain a constant period of oscillation, except as it is affected by variations of temperature and pressure and by a. decrease in its amplitude of oscillation. If corrections are applied for these variations it is possible to obtain the period of the pendulum by dividing a. fa,irl~, long interval of time by the number of' its vibrations during that interval and thus. secure considerable precision. The gravity pendulum swings in a. partial vacuum, under a pressure of less than a tenth of ad atmosphere, and is supported on a very accu- ratelv ground knife edge. It is, therefore, subject to very little friction of any kind and, if started with a, total arc of' vibration of 1~, it will swing for 12 hours and still have a. total arc of nearly 2S', which is large enough to permit of' accurate ~oinc.idence observations. In 12 hours, the pendulum oscillates about 86,000 times and this males possible a very exact determination of its period. The period of the pendulum is obtained in terms of some timepiece. Due to its portability, a. chronometer is usually employed for this pur- pose. There. is at present no timepiece in the world of the portable kind that will maintain a constant rate within one part in a million, the a.ecu- rac.y required for gravity world. It is. necessary, therefore, to determine the rate of the chronometer or other timepiece used in making the gravity observations and to apply a correction for this rate. A fairly sim- ple and entirely satisfactory way of obtaining the chronometer rate is to make chronographic comparisons between the second beats of the chro- nometer and the time signals, of the U. S. Naval Observatory as sent out by radio from Annapolis, Arlington, or Bellevue, using an automatic re- cording device for this'' purpose. It is necessary even then to apply a cor- rection for the variation in the error of' the sending clock at the Naval Observatory from day to day, although this variation usually amounts to only a few hundredths of a. second. The relative precision required in making the various observations :for a determination o:t' gravity is well worth considering. An error o-t' 0.000 0~00 So second in the period o:l' the pendulum corresponds, to 0.00] c~n/sec" in the derived value of the intensity of gravity and may be caused by any one of the -~:'ollowing errors: 1) A total relative error of 10 seconds in the coincidence readings at the beginning and end of' a 12-hour swing. 2) An error of' 0.04 second in the observed daily rate of' the chronometer. 3) A variation of 0.()2 second in the time required for the armature to open the shutter ifs the flash box between the beginning and ending of 1.2-1~our swing. 11

OCR for page 151
160 FIGURE OF THE EAR TH 4) An error of 0.06 C. in the mean temperature of' the pendulum if it is made of bronze or 0.9 C. if it is made of i.nvar. b) An error of 3 mm. of mercury in the observed mean pressure within the receiver. 6) Frrors of 10' and b' in the observed arc of ~-ibra.tion at the begin- ning and end, respectively, of a 12-hour swing. 7) An error of 0.014 wave-length of sodium or helium light in the observed flexure of the pendulum support. Item No. 4 in the preceding list of errors shows very conclusively one advantage of invar over bronze pendulums. Bronze pendulums, must be swung in a constant temperature room as it is impossible to obtain the mean temperature of the pendulum with sufficient accuracy over an inter- va.l of several hours if there is very much fluctuation of the surrounding temperature. The invar pendulums, on the other hand, can, ordinarily, be swung in a. tent as it is only necessary to obtain their mean tempera- tures within a. few tenths of a degree Centigrade. Although i.nvar metal is rather peculiar in its behavior and seems to be subject to molecular and other uncertain changes, the pendulums of this metal, which have been used for all recent gravity determinations of the Coast and Geodetic Survey, have given no particular trouble in this respect.. Inva,r pendulums are magnetic, however, and must be tested for magnetism at each station and kept as nearly demagnetized as possible. In regard to item ~ above, it may seem to the reader that the specified accuracy would be impossible to obtain. It is accomplished by varying the conditions and taking the mean of a large number of independent obser- vations. The size of' the flexure correction depends to a, large extent, of course, on the kind of support used for the receiver. In order to avoid the delay incident to building a pier the receiver is, often mounted directly on a concrete floor or on stone brooks. or bricks which rest on a, concrete floor. If the floor is thin the flexure is apt to be large in such cases. COINCIDENCE OBSERVATIONS The gravity pendulum is a free-swinging pendulum. It has no, directly connected mechanism to drive it or count its oscillations, as that would seriously affect its accuracy. The exact number of its oscillations over an interval of several hours must be obtained, however, for use in comput- ing its period. It would be almost a hopeless task for an observer to watch the pendulum for several hours and count the half-second oscillations. It would be even more difficult i'or him to determine by direct observation the exact chronometer times within a hundredth of a second at which the pendulum passes through its neutral position, or some other definite point in its oscillation, at the beginning and end of the interval.

OCR for page 151
DETERMINATION OF "I" BY MEANS OF PENDULUM 161 The method of observation by coincidences is a. very simple way of overcoming these difficulties. A coincidence observation. consists in noting the chronometer time when the pendulum and the chronometer are beat- ing exactly together. They are considered as beating together if the pen- dulum passes through its neutral position, corresponding to its. position of' rest, at exactly the same instant as the shutter in the flash box is opened by the break-circuit title ok the chronometer. At that time the two images of the illuminated slit in the flash box, one reflected from the stationary mirror in the receiver and the other from the moving mirror on the pendulum, appear in the telescope in the same horizontal line. The pendulum is. constructed of such a length that it has. a period slightly greater than one-half of a sidereal second. Therefore, when the shutter is opened by the break-circuit ticl: of the chronometer next fol- lowi~g a coincidence it hats not quite reached its neutral position arid the two indulges of the slit no longer appear opposite each other in the tele- sc.ope. With each succeecli~:~g sec.o~d the movable image gets farther and i:'arther away front the stationary one until it passes out of the held of view. Al'ter a short tilde it reappears in the telescope butt moving in the opposite direction. l~ina.lly the two! images appear again in a horizontal line and give another coincidence. In the interval between the two coincidences the pendulum has made ens 1~s oscillation than twice the number o seconds ticked off by the chronometer. The number of' its oscillations is tl~erel'ore easily obtained. The interval between two coincidences is known as. the coincidence in- terval. If the chronometer has a steady rate this interval is fairly con- stant at a station I'or any one of' the pendulums. Its average value can be determined by observing 3 or 4 coincidences when the pendulum is first started to oscillate, and 3 or 4 more after 10 or 12 hours. The num- l~er of coincidence intervals, or the number of times the pendulum has lost a beat on the chronometer, can then be obtained by dividing the total time between the first and last coincidences by the' average coincidence interval. With those data the number of oscillations of the pendulum during the 10 or 12 hours can be readily found and.. finally, flee peri.ocl of one oscillation ;: terns ol' the chronometer second. ~ .^v .~ OTHER OBSERVATIONS At the time the coincidence observations are made, the temperature and pressure within the receiver and the length of the are through which the pendulum is swinging must also be read. These observations are made through windows in the receiver. To facilitate the arc readings, a small observing telescope is secured to the receiver which can be moved by a race`; alla pillion to a position such that a cross wire in the telescope co;~- .

OCR for page 151
~2 FIGURE OF THE EARTH cides with the index line on the bob of the pendulum at first one extreme off its motion and then at the other. Just below the pendulum bob there is a slightly curved scale, graduated to millimeters, attached to' the bottom of the receiver. The amplitude of the swing is obtained by reading on this scale the positions of the index line of the pendulum at the two extremes of -its vibration. The flexure observations at a station are usually made after the pen- dulum observations have been completed. The receiver is left exactly as it was during the other observations except that the air is let into it and a plug is removed near the top to permit the insertion of the aluminum arm which is attached to the knife edge plate and which supports the sepa- rate reflecting plate of the interferometer. The observations consist. in reading the width of one fringe and the movement of the fringes on a decimal scale located at the focal point of the telescope. After five obser- vations of each quantity have been made the observing conditions are varied by cha.ngin~, slightly the angle between the reflector plates and thus varying the width of the fringes or bit changing the amplitude of vibration of' the pendulum. Altogether, four different sets of conditions are used. The movement. of the l~ni.fe edge plate when the receiver is supported on an ordinary pier and the pendulum. is swinging through an arc' of five millimeters amounts to about one-fifteenth of a fringe width of sodium or helium li.~.t. This corresponds to about' 0.02 micron, or less than on ore-millionth of an inch. If' the pier is too small, or if the soil around it its loose, the fle,xure may be two or three times as much as this. Its stile depends also on how far the foot screws project below the receiver and on how tight the lock nuts are set. It has been found by careful tests that on a good pier the greater part of the flexure is. in the receiver and its foot screws. Probably this is due to ~ large extent to the elastic compres- sion of the metal at the lower ends of the foot screws. VON- STERNECK PENDULUM APPARATES The VOll Sternecl: apparatus is quite different in many of its details from Blat of' the Coast and Geodetic Survey, although similar in several essen- tial respects, such as in the length of its pendulums. The pendulum sup- port consists of a well-braced frame secured to a heavy circular base plate which rests on foot screws. All levers for manipula,tin, the pendulums and for making any of the acljustments, required when cha,ngin`~, from one pendulum to another are located in the base plate to permit the use of a bell-jar type of' receiver. There are four principal pendulums and ea.cl:~ has a liaise edge secured in its head. Each hats its own place iffy the receiver and each. its own sup- portin', plane. All four pendulums are placed on tl~.is France at the same time. This is clone before the receiver is placed ;~ position. A system

OCR for page 151
OFTERMINA LION OF 9 B Y MEA NS OF PEND ULU.M 1 (; .3 of' mirrors and prisms controlled -i'rom the outside by means of levers working through pac.kin~, boxes in the base makes it possible to change the observations from one pendulum to another without opening, the re- ceiver or changing the position of the observing, telescope. During the observations on any one of the pendulums the other three are prevented from talking up sympathetic vibrations by means of clutches controlled frown the outside of the receiver which OCR for page 151
164 FI CURE OF THE EAR TH GRAVITY DETERMINATIONS AT SE! WITH PENDULUM APPARATUS Geophysical investigations have heels greatly impeded in the past by the lack of accurate gravity determinations over the oceans. Many at- tempts were made, by a number of scientific men o-t' different countries using several different kinds of specially designed apparatus to meet this lack but the results were of little geophysical value. IJntil recently it seemed impossible to obtain an accuracy at all comparable with that of land determinations. The problem was finally solved very successfully by Dr. Meinesz who adapted the von Sterr~eck apparatus I'or use on a, sub- merged submarine. The method used by Dr. Meines% depends for its success upon com- bining the simultaneous oscillations of two pendulums in the same re- ceiver swinging in nearly opposite phases in the same plane. When the two pendulums are thus combined, there are obtained the oscillations of a hypothetical pendulum which is free from the effect of any small l~ori- zontal accelerations of the ship in the direction in which the pendulums swing. The apparatus is used in a submerged submarine as the motion of a surface ship, even when the sea. is almost calm, is. ordinarily too great to be corrected for in this manner. In order to be able to combine the oscillations of each pair of pendu- lums swinging in the same plane Dr.: Meinesz designed a, photographic apparatus which ~,a.ve a continuous record on a moving film of the oscilla,- tions of' the pendulums. By an analytical process. the simultaneous oscil- la.tions of each pair were combined to give the oscillations of the cor- res~ondin~ hypothetical pendulum. Both pairs of' pendulums of the - Or ~ = --a 1~ ~ von Sterneck apparatus were thus recorded and combined and in this way the effects of both the endwise and s.idewise motions. of the ship were eliminated from the results. After making a series ol about thirty determinations between Holland and Java through the Mediterranean Sea, Dr. Meanest found that the computations involved in combining, the records of' the pairs of pendu- lums by the analytical method were discouragingly difficult and laborious. As soon as he returned to Holland, therefore, he set about modifying the apparatus in such a way as to shorten the computations. In this he was very successful. lIe virtually designed a new apparatus, although a fen of the details were copied from the Volt Sterneck apparatus. The new Meinesz apparatus has three principal pendulums, all of which swing in the sane plane. The two outside pendulums are started oscillating at the same tinge by means of a device which automatically starts them out of' phase with earls other. The middle pendulum does not oscillate except as it is started by the motion of' the ship. A complicated but ingenious systems of prisms and mirrors combines the oscillations of

OCR for page 151
DETERMINATION OF " g " BY MEANS OF PENDULUM 165 the middle pendulum with those of each of the other two directly on the photographic film. In this way the record is made to show the oscillations of two hypothetical pendulums which, as previously explained, are un- affected by the horizontal acceleration of' the ship in the direction in which they swing, and a great amount of' exacting computations is, thus eliminated. In order to keep the apparatus as nearly as possible in a level position, it is supported in a cradle hung in gimbals. The pen.du- lums swing at right angles to the axis of the ship. Besides the three principal pendulums there are two short, heavily damped pendulums, one of which records the tilt of the apparatus, at right angles to the plane of' oscillation and thus takes account of the endwise acceleration of the ship, and the other serves as err object of reference in recording the a.mpli- tude of oscillation of the middle one of the principal pendulums. These data are needed in making the computations. Dr. Meinesz first tested out his new apparatus by making a few deter- minations in 1920 between Holland and Port Said. Then in 1926 he used it for an extended series of about one hundred and fifty determinations between Holland and Java by way of the Panama Canal. In the fall of 1,'328 he brought the apparatus to the United States and in cooperation with the United States Navy and the Carnegie Institution of Wa.shi.ng- ton he made about fifty determinations, in the Gulf of Mexico and in. the waters surrounding the West Indies. This thorough trial has proved beyond any doubt that the problem of determining OCR for page 151
~ 6~; FIGURE OF THE EAI?TII error which may be introduced from this cause may be inferred from the fact that an east or west velocity of one knot at the equator causes a. change of about ~ part in 130,000 of the total intensity of gravity. SUGGESTIONS FOR ADDITIONAL READING It is suggested. that those interested in making a more detailed study of the different kinds of pendulum apparatus and methods used in determining gravity, consult the following books and articles: Fran~,ois, Ch. Note sur l'appareil gravimetrique de l'Observatoire Royal de Belgique. Anna,les de l'Observatoire Royal de Belgique. Brussels. p. 1-36 (1920). Galbis y Rodriguez, Jose. Determinacion relative de lo intensidad de lit fueler de ,gravedad. Memorias del Instituto Geografico y Estad~st,ico. Madrid. NIII: 7~D (1905). Haasemann, L. Bestimmung der Intens,itat der Schwerkraft auf 35 Stationen ..... Veroffentlich. Kg. Preen. Geodat. Ins,t., No. 71. Berlin. p. 1-32 (1916). Meinesz, F. A. Vening. Observations de pendule sur la Mer Pub. commis sion geodesique neerlandaise. Delft. p. 1-16 (1923). -- . Projet d'un nouvel appareil pendulaire. Bulletin geodesique No. '5. Paris. p. 1-10 (1925). . The determination of gravity at sea in a submarine. Geog. J. London, 65: 501-521 (1925). Mendenhall, T. C. Determinations of gravity with the new half-second pendulums of the Coast and Geodetic Survey Appendix 15, U. S. Coast and Geodetic Survey Report for 1891. Washington. p. 503-531 (1892). Sterneck, Robert von. Der neue Pendelapparat des kaiser!. konigl. militar-geo- ~r,,raphisch. Institutes. Mitteilungen k. k. militar-geogr. Institutes. Vienna 7:83-116(1887). Swick, C. H. Modern methods for measuring the intensity of gravity. U. S. Coast and Geodetic Survey Special Publication No. 69. Washington. p. 1-96 (1921).