^<^U ^^' IMAGE EVALUATION TEST TARGET (MT-3) 1.0 I.I 1.25 UIM WIS ■^ 1^ mil 2.2 Mi K4 1.8 1.4 1.6 6" Photographic Sciences Corporation 23 WEST MAIN STREET WEBSTER, N.Y. 14580 (716) 872-4503 •NJ \ ;V \\ V O" CIHM/ICMH Microfiche Series. CIHM/ICMH Collection de microfiches. Canadian Institute for Historical IVIicroreproductions / Institut Canadian de microreproductions historiques .,.i^„m!Mmmmmmtmmmm^-^i^ Technical and Bibliographic Notes/Notes techniques et bibliographiques The Institute has attempted to obtain the best original copy available for filming. Features of this copy which may be bibliographically unique, which may alter any of the images in the reproduction, or which may significantly change the usual method of filming, are checked below. L'Institut a microfilm^ le meilleur exemplaire qu'il lui a 6X6 possible de se procurer. Les details de cet exemplaire qui sont peut-dtre uniques du point de vue bibliographique, qui peuvent modifier une image reproduite, ou qui peuvent exiger une modification dans la mdthode normale de filmage sont indiquds ci-dessous. I J? Coloured covers/ Couverture de couleur [TTT Covers damaged/ k\} Couverture endommag6e D D D D D D D Covers restored and/or laminated/ Couverture restaurde et/ou pellicul6e Cover title missing/ Le titre de couverture manque I I Coloured maps/ D Cartes g^ographiques en couleur Coloured ink (i.e. other than blue or black)/ Encre de couleur (i.e. autre que bleue ou noire) Coloured plates and/or illustrations/ Planches et/ou illustrations en couleur Bound with other material/ Reli6 avec d'autres documents Tight binding may cause shadows or distortion along interior margin/ La reliure serr^e peut causer de I'ombre ou de la distortion le long de la marge intdrieure Blank leaves added during restoration may appear within the text. Whenever possible, these have been omitted from filming/ II se peut que certaines pages blanches ajout^es lors d'une restauration apparaissent dans le texte, mais, lorsque cela 6tait possible, ces pages n'ont pas 6t6 film^es. Additional comments:/ Commentaires suppl^mentaires; □ Coloured pages/ Pages de couleur □ Pages damaged/ Pages endommagdes □ Pages restored and/or laminated/ Pages restaurdes et/ou pellicul6es I — I Pages discoloured, stained or foxed/ D D D D Pages ddcolordes, tachetdes ou piqudes Pages detached/ Pages d6tach6es Showthroughy Transparence I I Pages detached/ I I Showthrough/ Quality of print varies/ Qualitd indgale de I'impression Includes supplementary material/ Comprend du materiel suppl^mentaire Only edition available/ Seule Edition disponible Pages wholly or partially obscured by errata slips, tissues, etc., have been refilmed to ensure the best possible image/ Les pages totalement ou partiellement obscurcies par un feuillet d'errata, une pelure, etc., ont 6t6 film^es d nouveau de faqon 6 obtenir la meilleure image possible. This item is filmed at the reduction ratio checked below/ Ce document est film6 au taux de reduction indiqu6 ci-dessous. 10X 14X 18X 22X 12X 16X 20X 1X1 24X 26X 30X 28X 32X I i B Stalls s du lodifier r une Image The copy filmed here has been reproduced thanks to the generosity of: Library of Congress Photoduplication Service The images appearing here are the best quality possible considering the condition and legibility of the original copy and in keeping with the filming contract specifications. L'exemplaire film6 fut reproduit grdce A la g^nArositd de: Library of Congress Photoduplication Service Les images suivantes ont 6t6 reproduites avec le plus grand soin, compte tenu de la condition et de la netteti de Texemplaire filmd, et en conformity avec les conditions du contrat de filmage. )S Original copies in printed paper covers are filmed beginning with the front cover and ending on the last page with a printed or illustrated Impres- sion, or the back cover when appropriate. All other original copies are filmed beginning on the first page with a printed or illustrated impres- sion, and ending on the last page with a printed or illustrated impression. The last recorded frame on each microfiche shall contain the symbol — ^- (meaning "CON- TINUED"), or the symbol V (meaning "END"), whichever applies. Maps, plates, charts, etc., may be filmed at different reduction ratios. Those too large to be entirely included in one exposure are filmed beginning in the upper left hand corner, left to right and top to bottom, as many frames as required. The following diagrams illustrate the method: Les exemplaires originaux dont la couverture en papier est imprlm^e sont film^s en commengant par le premier plat et r^ terminant soit par la dernidra page qui comporte une empreinte d'Imprasslon ou d'illustration, soit par le second plat, selon le cas. Tous les autres exemplaires originaux sont filmds en commenpant par la pramidre page qui comporte une empreinte d'impression ou d'illustration et en terminant par la dernlAre page qui comporte une teile empreinte. Un des symboles suivants apparaitra sur la derniftre image de cheque microfiche, selon le cas: le symbole — *> signifie "A SUIVRE", le symbole V signifie "FIN". Les cartes, planches, tableaux, etc.. peuvent dtre filmds d des taux de reduction diff^rents. Lorsque le document est trop grand pour dtre reproduit en un seul clich6. il est film6 d partir de Tangle supArieur gauche, de gauche d droite. et de haut en bas, en prenant le nombre d'imagas nicessaire. Les diagrammes suivants iilustrent la m^thode. errata to t pelure, on d n 32X i i 1 2 3 1 2 3 4 5 6 lU I On the Use of the Interferometer in the Study of Electric Waves *■ ■ SUBMIISTtD TO THK FaCULT«8 OF THE GRADUATE SCHOOLS OF Arts, Literature, and Soence, in Candi- dacy FOR THE Degree of DOCTOR OF PHILOSOPHY {Dipartmmtof Phyna") BY GORDON FERRIE HULL (l^rinted item die Phtoigal Rbvikw, VoL V., Na 28, OctolKr, «897 ) r'^, CHICAGO, MAY i, 1»97 •=-§■%•- r* '*■ r^ ~- - > il.^^a^V~ .^.i*. -H- iA.c ■ . . . . iv-iSf ■~r^- THE PHYSICAL REVIEW. '■A JOURNAL OF EXPERIMENTAL AND THEORETICAL PHYSICS. CONDUCTBO BY BDWAKD L. mCHOU, ■busst mibritt, and fbbdkkick bbdbll. Beginning with Volume V. (July-December, 1897), two vol- umes of The Physical Review will be published annually, these volumes beginning in July and January, respectively, and con- taining at least five numbers each. The price of subscription is two dollars and fifty cents a volume (five dollars a year), or fifty cents a number. Subscriptions should be sent to the publishers. The Macmillan Company, 66 Fifth Avenue, New York; Messrs. Macmillan & Co., Ltd., London; or to Messrs. Mayer and Mueller, Beiiin, Volumes I.-IV. of The Physical Review are annnal volumes, each containing six bi-monthly nunibers, beginning with the July- August number, 1893. These may be obtained from the publish- ers at the former subscription price, three dollars per volume. Conrespondence relating to contributions should be addressed to the editors, at Ttkaca, New York. Manuscript intended for publication in The Physical Review must be communicated by the author; when publication in other journals is contemplated, notice to that efiect should be given. The authors of original articles published in the Review will receive one hundred separate copies in covers, for which no charge will be made ; additional copies, when ordered in advance, may be obtained at cost PUBLISHED FOR CORNELL XJNIVERSITY. THE MACMILLAN COMPANY, NEW YORK. Cbc "lanivcrsitv) of Cbicaoo lofspKI) I'.Y .imiN 1>. Kill K I'll M.I K On the Use of the Interferometer in the Stiidv of Electric Waves A DISSERTATION SniMirTKi) 1(1 riiK FAeri/riF.s ok thk (Ikadia ik Smooi- oK Arts, LriKKATiRK, and Sliknli:, in Candi- dacy I'OK riiK I)k(;i^-/ ^c , ox Till': L'Sl". ol' Tl STUDY 01 ■: i\ri:Ri-i:R()Mi:ri':R in r i:li:ctric \v.\vi:s. liv (i. I', ill II. OlTMNi:. 1. Introduction. — Use of tlic intcrfLTonictcr in the study of lij^ht. 2. Historical statrnicnt. — .Xnalo^y between iiijht and electric radia- tion not complete. Work of .Saiasin and de la Rive, . . . for Hertzian apparatus pro\in_L^r that tlie \va\e-lenj4th nieasuretl depends on the recei\er. Is the ratiiation from a I lert/ian vibrator simi)le or complex ? I'",.\i)eriments anil theory dealiiiLC with this point. lieariiiL,^ of this theor\- in our experiments. 3. .Apparatus. — Interferometer; conditions ^a)verninj; choice of \i- brator and receiver ; sensitiveness ; absence of diffracted aiul scattered radiation ; preliminary observations. 4. Measurement of /.. — DeL;rce of accuracy. 5. Ivstimation of " />' " from the interference curve. — Proof that this does not ^^ive the loj^arithmic decrement of the vibrator, as has been suppo.sed. Comparison of experiment with theory. 6. Influence of receiver upon the interference curve. — Inference as to the period and dampin<,r coefficient of our receiver. 7. Dependence of ?. upon vibrator. — No agreement between \alues of / found by other observ. ■ for vibrators similar to our own. 8. Determination of /i. — Degre. . accuracy. Conclu.sion. I. Use of tiik Interferometer in the Studv of Licht. The interferometer, as a special form of refractometer has been called by Professor Michelson, has been found to be a very- power- ful instrument in the study of light. By its mean.s the nature and wave-length of the radiation from a source and the index of refrac- tion of a transparent medium have been found with the greatest accurac)'. It is true that a grating or prism spectroscope must be used for the study of complex light, but for comparatively simple /. //Y //, [VmI, V radi.iliini tlu' inltrtlniiiKtrr pusscsscs afar ^naUr pouir of analysis. It was witli the oliji'ct of constructing an intcrfcronictiT for electric waves and of nsinj,' it in tlnir study liiat tliis ivse mtIi was nndrr- takiii. 2. 1 Iciv it may he well to point out that electrical and lij^dit oscil- lations differ in an important respect. The hreadth of a spictral line of a liomoj;enei)Us ^as has been accounteil for upon various assumptions amon;^ which is the one that the oscillations of the molecules of the ^as ^raduidly die down owinj^^ to its communi- cating^ ener^)' to the surrountlin^ medimn, or to other causes. l'",xperiments, however, have failed to \erify this assumption.' In the case of electrical radiation, cm the other hand, theor\' indicates and eNperimeiit proves that the oscillations are alwaj's damped. Another difference is in the apparatus used for detecting thi' radi.i- tion in the tui) cases. I'"or li^dit, tlu' eye is the usual tletector. 'rhoui;h it is sensitive for radiation lyin^ within a very small range, it has no periiul of its own and does not possess the power of influ- eiicin^f the measured wave-length. That electrical receivers, as a rule, have this power has been pro\ed by the experiments of Sara- sin antl lie la Rive," Klemencic antl Czarmak,' and (|uite recently b\- those of Wiedeburg.' In fact, for the different forms of I lertzian vibrators and receivers used by these ph\'sicists, the wave-length measureil depemled almost entirely upon the receiver. The explanatioi; of this fact led to the important cjuestion whether the. radiation from a Hertzian vibrator is simple or complex. Ui)on this point opinions have differed. I'rom the phenomenon of multi- ple-resonance as discovered by Sarasin and de la Rive- and Klem- encic and Czarmak' many physicists were led to the conclusion that the radiation is comi)lex. The experiment of Zehnder'' show- ing that the rays of electric force are analyzed by a grating into a spectrum pointed in the same direction, while Garbasso and Asch- kinass'' found that rays, in passing through a prism made up of 'A. A. Mii'liclson, .\stiophysital Imirii.'il, ]>. 251, Xnvcmlicr, iSyS. 2 Sarasin aiul dc la Rivo, t'nmpt. KlikI., 1 1 5, ]>. 439, 1892. !< Klciuciicic and C/ariuak, Wiod. .\iin., XOl. $'>, 1S9J. < WicdclniiK, NVied. .\iin., \ol. 59, p. 496, 1S96. ''•/clindor, Wicd. Ann., VdI. 5,5, p. 172, 1894. oGaiiiasso and Asclikina.ss, Wicd. .\nii., Vol. 53, p. 534, 1S94. 4l r.VA ('/• /AV/A7/ A'(',l//7/ A'. ^;lass pl.itcs up'ii uhicli \\\\v pastcil strips of tiiituil, an' disporsfc! ami coiicliidicl lliat rays of ckctric fi>\w iiia>' \n- lonsidi inl imt necessarily as iiiDiiotlirdniatic hut, with the sanu' jiistifuMtidii .is in the case of li^lit, as c Jiiipositc, ( )ii tlie other h.md HJimUiu's' iisJM^r I.i'eher's arrail^'enieiit •' attemptid topicve that tlie radiation is (hic to a simple clamjied oscillation of tlic form . /, "• sin {iiZ + a'). Tile theory of his nuthoil, u hicli iliffered from the latir work of KIcmeiicic ami Czarmak chictly in tiiis that he used waves aloiin wires, uhili' they useil waves in air is as follows : .Assuming' that the oscillations are of the form ./,""' sin (t/, he finds the effeet upon the receiver of two infniite tr.iins of waves, one of which is direct from the vibrator anil the other rellected from the ends of the wires. This effect (usin^ notation and limits of inte|^n-ation to suit our experiment) is j^iven by ./= C' J<--"'.sin-,?/,//4- ['^^ [-h-'siv ,//+./,-"("■'"' .sin a i/+ ~'\Y .1- I "■'• .IV = 1 1 -ft' - cos 2a I. In lijerknes' experiment .r was the distance of the receiver from the end of the wires ; in ourexperiments it is the distance of a mirror from the " zero position " ; in both cases 2.ris the total difference in path of the two trains of waves. Thus the curve wiiosc ordinates are i)ro- porlional to the intensity of the electrical radiation for a jjoinl v\ hose abscissa is .r , is leaving out a constant, a tlamped cosine curve, It was found that the experimental ai,n-eed fairly well with tlie theo- retical curve. It was therefore concluded that the oscillations fol- lowed the law assumed. This conclusion is open to two possible criticisms which are su^^- ge.sted by the questions : first, may not the wires with their terminal plates or the receiver e.xercise a selective action on the radiation, and second may not the interference curve be obtained assuming; an- other iaw for the oscillations. Indeed ]5jerknes found that the inteiference curve was infliienced by the chanj^e of the distances between the plates of the vibrator and the terminal plates of the ' lijoiknes, NVicd. .Ann., Vol. 44, ji, 51J, iS(,i. ^I.echcr, Wicd. Ann., Vol. 41, \i. S50, lSi,o. G. /•; ///•// IV.. I.. V. wires — tli.it ,is tliis ilistancc iiicivaNtil tin.' u,i\rs uiii.' imt su i,i|)- i(ll\- (l.miprd hut were laintt-T — a result wliich aj^rirs uitli tlu' thc(iry tliat the mnie teil)le llie lailiatimi the less rapid is its de- cay. I'ri)l),il)l)- the uiris were too Inn^' to atlect the result hut that is a point uhieh in ^eiier.d slmuld he considered. The results ohtaiiied h)' .Sarasin ;iiul de l.» Ri\e ami the other ph)sicists named .ihow, indicitiii^; th.it the radi.ition is complex, m.iy he .iceounted for h_\ .issiimin^f th.it the oscill.itions are simple ■md d.imped .iiid th.il the receiver h.is ,i piriod of its oun and is conip.iratively und.imped. On the other h.md, lU'^lectiuj^ the criti- cisms hroiiL;ht forward, it would he diflicult to account for the rej^'ii- l.irit)' of the interference curve ohtainetl !))• Kjerknes on the assump- tion that the radiation is complex. On the whole the (piestion has not Ik en satisfactorily settled. We are justificil in expectini^^ that the prohkni may he approximatel)' soUed hy means of the inter- ferometer. The ([iiestion which concerns us is this — e'ven if electric radiation is, in general, suflicientl)- simple to be analyzed hy the interferom- eter, how shall we inter[)ret the interference curve — is its form influ- enced by the receiver? The latter point is a matter for experiment to decide in the case of the receiver u.sed. Xoti'. — The theory which applies here is rather unsatisfactorj'. Assuming; that the receiver, when no outside forces act, executes dampetl pendulum motions and th.it the oscillations of the vibrator arc, of the same nature wc obtain the equation : 'J.+:-i 'J + (/'^)f =^i'-'"sin {at+c) where ^f represents the potential difference between the parts of the recei\ er. I,et us impose the conditions 9'= , =0 when t=o. 1 he solution may be expressed in the form — f=C sin I ^ (+(■') where .1- (rt-_/r)--f-(rt--//-)(«--^i-) ■- ^ Nn. 4,] is/-: 0/ A\7/.A7/.A('.I//. //.A'. I 4- cos I - ,,4./, d"-^')'" ^-[i-/>)i:n^{'i-/'y-('i-^ii^u\(.i-/'y) + sill I Tlif in'tion nil the R'c\i\ir diir to tin- t\vult, it sclius iiL-L-cllcss to expect assistance from this liii(>r\'. The work so far descrilxcl was pirformed with the I lert/ian \ibra- tor and receivers. I am not aware tliat an)- eorrespoiulin^ work has been done witli s|)heres unless it is that of l.anj^' and of IJose." I .;i\v^ obtaineii interference effects by a metliod an.doj^nms to that of (Jiiinke in sound. But it seems probai)le tliat retlectioiis from tiie tubes to which tlie raiiiation was led would obscure the effect. Hose states that the radiation from a sphere under the conditions existinij in his experiment {jives a ////<■ s|)ectruni from which the inference is to be drawn that the \ibrations are not on!)- simple but wry sli^ditly damped, lie also states that the wave-lenjjths measured were iiule- pendent of the periodicity of the receiver. If this were true it must ha\e been on account of the dead beat character of the receiver — not on account of the method used. The experiments to be described in this paper will show (1) that the oscillation due to electrical disturbances on spherical conductors are for the most part, rather liiyMy damped sine functions, (2) that the interference curve is influenced by the receiver — sometimes to such an extent as to change the apparent wave-leiiyth. 3. Al'I'.AKATL'S. The interference device shown in Fig. i was patterned after Michelson's simple but effective interferometer. Radiation coming from the parabolic mirror /^ falls at an incidence of 45° on a sr/>ii- ratiitg surface S which reflects part to M and transmits the rest ' Victor von Lang. I'hil. Mng., Kob. 1S96. «Iiose IVoc. Roy. Soc, Oct. 16, 1896. 'I G. /■: HULL. [Vol.. V. which {,foc.s to J/'. Of the radiation reflected from .1/ aiul J/' the parts transmitted and rellected by .S" fall toj^^ether upon A' (See VVg. I ). Difference of path is obtained by the motion of one or both mirrors. To obtain as lart;e an effect as pos:-ible in A' the separat- ing surface should transmit anil rellect one-half of the radiation I'all- \\v^ on it. For this purpose there were pasted on a piece of card- board, half a metre st[uare, strips of tin foil 1x40 cms. At first the adjacent edges of two neighboring strips were one centimeter apart. Later it was found necessary to remove every alternate strip, mak- ing the adjacent eiiges 3 cms. apart.' The reHection and transmis- Fig. 1. iThis agrees with the conclusion of Lord Layleigh deduced from theoiy, viz.— a narrow aperture parallel to the electric ' Uirations transmits very much less than is re- flected by a conductor elongated in the same direction Phil. Mag., April, 1897, p. 272. N". 4.] r.S'A ('/• l.\ II.KI I.K'KME I'EK. siini powers of the Ljratiiit; wrrc (ktiTiniiKi,! 1)\- well -known iiuaiis and it was found that when the stiijjs were paralkl to tlie ek'ctric oscillations, this <;ratiii|4 retlected and transmitted about half of the radiation fillin;^ on it. The woollen arms t>t 40 cms. and an aperture of 30x40 cm. As it was intendeti to use wa\es not lon^^er thm 10 cm. tlie ilimensions of J/ and J/* and ■'! the ajjcrturc of /'were lari;e enouj^di to obviate any serious diflrac- tion effects. The focal len^^th of /' was chosen about one-fourth of the wave-length in order that the direct and reflected waves mi.;ht be in the .same phase. But as the radiation was too stront^ (unless the galvanometer was short circuited) this precaution was unnecessar)-. The arrangement for the vibrator usually consisted of two spheres connected by fine wires to tlie secondary terminals of the induction ct-il and sparking to each other in oil. ICach sphere was screwed on the end of a hard rubber rod which passed through a rubber cork and this was inserted in a glass tube about 10 cm. long and 2.5 to 3 cm. in diameter. This tube was partiallj" filled with oil — usually vaseline oil, sometimes paraffine oil. A small hole blown in the side of the tube allowed one, by rubbing paper between the spheres, to partially clean them without removing the vibrator from Its position. Righi's arrangement' was also used. Four hard rub- ber rods each carr>'ing a s[)here, were attached to a vertical rod in the mirror /'. All distances were adjustable. The tliird sphere was fastened by means of melted wax in a small glass cylinder of si:eh a size that it held sufficient oil for the spark gap and yet allowed the spheres 2 and 3 to approach until their surfaces touched. The arl- \antage of this arrangement was that all the spark gaps could be fairly well cleaned by paper without removing the spheres. The rod .//» could be rotated so that the spheres could be brought out towards the aperture of /'or moved back to the focal line. ' Kidii, Mem. del K. Acivl. del Se. del Insl. di l!nl(if;iiii, T. IV., 1894. 8 c. /. ni i.L. [V,,i.. V. The rcccivLM- chosen for this experiment must fulfill certain condi- tions. It must be sufficiently sensitive to respcjnd to radiation reaching: it after reflection and diffraction, sufficiently constant to ^mvc (luantit.ilive results, ami as we desire to determine the nature of the radiation it should respond ecjually well to waves of all periods sent out b)- the vibrator. Now it has been shown that 1 lert/ian receivers do not fulfill this last condition. It was thought that a receiver which responds to electric wa\'es of periods extending' over a wide rant^e would measure accurately the radiation falling on it. Having had some experience in the use of the colicrcr (as named b)' Lodge, though the name seems misleading) we thought that if it could be made constant it would answer our purpose. Consequently we used a liranley or Lodge receiver consisting of small nails or pieces of wire about I cm. long, in a glass tube i 5 or 20 cm. long and filled with a lubricating oil. This was placed in the focal line of a semi-circular zinc retlector of .iperature 10x20 cm. A sheet of tin foil over part of the aperture served to shut out, if neces.sary, part of the radiation. Lead covered wires comiected the receiver with the gaUanometer and battery which were enclosed in a tin box. Thus, with the cx- cei)tion of the aperture of the mirror, the whole circuit was enclosed in metal. In order to make the deflections small enough to be read, the galvanometer, a simple D'Arsonval, was short circuited by a resist- ance varying from one-fourth to one-half of an ohm. It is hardly necessary to state that in finding the proper grating the ordinary phenomena of polarized radiation were observed. For example, if the receiver, grating and \ibrator were placed in a line, with the grating immediately in front of the receiver and its strips iiorizontal (perpendicular to the vibrator), almost all the radiation passed through. If the strips were vertical only one-half pas.sed through. A block of wood with its fibres vertical transmitted much less than it did when its fibres were horizontal. It was found that the receiver responded nearly as well when the electric oscillations were vertical as when the>- were horizontal. In this respect, therefore, it differs from the Hertzian receiver. In order that the diffracted and scattered radiation be as small as pos- sible, the axis of the mirror, and therefore, also the vibrator, were N... 4.] r.vA or ixri-Kir.ROMEi'ER. vertical, for in this pnsition it was nut luccssan- to have metallic ends in the mirror. In our experiments the total scattereil radiation was only three per cent, of the effect when there was zero difference of path in the two beams. Considerini,' the faci that the recei\er was .sensitive to radiation reachiiii; it in all directions, whether from the air or from the wires leadin;^^ to the ^^al\anonu■ter, and that it responded .sli^ditly to the loni; waves of the induction coil, which was not enclosed in metal, this result was felt to be satisfactor\-. A small induction coil with an ordinary automatic interruptor, operated by one storage cell, completed the apparatus. A number of interruptors were tried, but they were not found to be more con- stant than the one belont,nnL,' to the C(.il, and were more complicated. Usually the key was closeil for about one .second. Closini,' it for a longer time did not increase the deflection, but onl\- served to destroy the sparkin;^ surfaces. \ 4. ]\Ik.\SL-I, c, the series n b c b a was taken. If we are concerned with the wave-length and not with the interference cur\-e it is necessary to find only the maxima and minima positions. But the accuracj- with which one of these positions can be determined depends on the sharpness of the curve at that point. Consecpiently the first step in the determination of / {or even of ft) is to roughly plot the interfer- ence curve, obser\-c the sharpness of the maxima and minima and choose the sharper of these from which is to be estimated the desired quantity. In order to see to what degree of accuracy / may be found, the positions of four successive minima were observed. The following results were obtained, the numbers referring to the position of the mirror on the scale 10 (;. /•• J/L/.L. [Vol.. - — Second 1 behind. FirBt ! First behind. before. Second before. Mean for A a. V.syt I . 6. 45 11. 15.55 20.2 4.58 nn. Kx,.t. -» 6.55 10.95 15.5 20.1 4.53 .111. Lxit, J- • ■, "■= 1 10.97 15.53 i 20.17 1 4.55 nil. 1 and tliL- mean of these j^nves /./2 = 4.-,6 cm. to witliin a fi-action of one pel- cent. The vibrator con.sisted of two .spheres 1.93 cm. in diameter arraiiijed as in l*"ig. I . 5. K.STIMATION OK " (V ." II we accept the theory gi\'en on pat^e 5 as correct we can from the curve find the logaritlimic decrement of the vibrator. Let t^^, a.,, 5=0.74. But it is certain that this theor\-, though it may apply when re- ceivers are used which are dead beat compared with the vibrators, docs not apply in the case of Klemencic and Czarmak's experiments nor in our own ; for we have found that the interference curve is in- fluenced by the receiver — a fact which this theory neglects. How- ever, it is interesting to compare the values of 2 log ' ■ (w hich we will call o' and which has some connection with the damping of the \ibrator or receiver or both), found by Klemencic and Czarmak and others, for Hertzian vibrators and receivers, with the correspond- ing value for the spheres and coherer. Their values varied from 0.3 to 0.5 while ours varied from 0.5 to i. This indicates either a more complex radiation or a higher damping coefficient in the case of our apparatus than in that of the earlier experiment. But these Nm, 4.] C ■.VA ( >/■ A\'/"/;A7 A A'( M/A / AA'. I I values (0.5 to ■'ii'c coiisidcrabl)- lower tliaii that sui;L;cstr(l by theory. I'or the aniphtmle oftlie oscilIali ; ; ,. u— r> -• — .J- I I I I I . I I I I ! I I I 1 : ! ^ I I I I I I I I I M I I I I ' I I I ' I '! I I I I I I I I I I I I I I ' I I I I I I I I I I . ' ' I ' Fig. 2. V^ N... 4.] r.s/-; ('/■' /.\7y;A7'7;A'('.i//;yA7v'. I ; view of tlie fact that receivers 11 ami 111 f'rthis vibrator -ave a \vave-leni;lh («f 4. J to 4.4 cm., we arc led to believe that the com- ponent of the shorter \vave-leiii;th is due to the \ ibrator and that the other component, which is present probably to a sliL;ht extent in the radiator, owes its prominence to a resonance action of the receiver. The interpretation that receiver 1 has a period correspond- in-r to the ion"- wave receives confn-mation in the fact that the cnrvc for vibrator 1 has such well defined maxima and minima. Hut, \vhate\er interpretation be placed on the interference cur\e, it is evident that it cannot always be taken to accurately represent the radiation. On the other hand the wave-len|^th for either vibrator is maintainetl by all the receivers (leavini; out curcvs 6, 7, 8, for which the wave-lengths are very indehnite). This fact shows that our receiver has far less influence on the interference curve than has the Hertzian receiver. In other wiuxls, the damping coefficient of the former is far less greater than that of the latter. Our expecta- tion, therefore, that the coherer used is highly dami)eil has been fulfilled, but it appears from our work that its damping coefficient is of the same order of magnitude as that of the vibrator and con.se- quently it influences the interference curve. In view of this action of the receiver we are not able to accurately analyze the radiation, but using a number of receivers we are able to arrive at a fair estimate of its nature. We are thus led to state that the chief component of the radiation from spheres is due to a damped oscillation. 7. DeFENDEN'CE of / II'ON VlI!K.\T()K. Besides the vibrators i and 2 we used the following arrange- ments : Vibrator No. 3, consi.sting of one sphere of No. i, to which sparked a small platinum bead. The interference curve is given in No. i„. The wave-length is the same as that of No. i, but the maxima and minima are less pronounced. Vibrator 4 dif- fered from No. 3 only in having one of the spheres of vibrator 2 in place of one from No. i. The interference curve is that of 2,,. Here again the wave-length is the same as when the two spheres arc used. We also used two steel spheres 0.79 cm. in diameter together and singly. Their curves, which are similar to those al- H c. /•; //i/.L. [Vmi.. V !■ ^^■ read)- found, arc not ■;iv(.ii. The wavc-lcn^^tli was a'xmt 4 cm. Rij;Iii's form of vibrator, with the two sj)licrc.s of No. I sparking' in oil and tho.sc of No. 2 .sparkinj^f in air, \va.s nseil. The waw-len^^th wa.s a^^ain ai)])ro\imatel\' efpial to that of No. i. It is here seen tliat tlie wave-len^'th is nearl\' proportional to the diameter of the sphere (the law is e.xpressed more nearl)- by /. = ///(^r+(i) where /// and o are constants and /•= railius of the .sphere), and that it is indeiiendent of the arran^fenienl b>' means of which the tlischar{;e is broujjjht about. 1 was not able to observe anj- change of wave-lenj;th due to a change of sparkling distance. It is iiUerestins.' to compare the values of / obtained by other ob- servers with those which we have found. It will be seen that no aL;reement exists between the results of any two observers. Dia. of Spheres. mm. 39.7 24.4 10.6 lfr^rwr^'ATor\ ^"-ver. M.V;?^^°f Observer. Reference. nun, 88 I Kighi'.s 82 80 80. 200 37. S 106 8. 26 7.8 18.4 19.3 91 9.3 43 7.9 40 I.LMl(,'th of ivliiulcr. 8x1 mm. SO 1.3x0.5 " 6 Nail I I Coherer. Interference l.anj;. " througli , " " two tubes. " Kiglii's Sliver on Interference Rigiii. " Glass. .Spark " " " j Ohservcd. " ' " 18.4 Lodge's one Siiirnl spring Grating. \ lk).se. Sphere, j Coherer. j Righi's and Coherers. Interferom- Lodge's. , eter. Righi's Thermo- element. Cole. Interference Lebedew AVied. Ann., 3, I 1896. l.'eclair.age Llee- trique. Xo. 3, p. 360. 1S95. llVoe. Koy. .Soc, j I.X, 361, 1S96. jSee pages 13-16. Physical Review. 1S96. Wicd. Ann. No. 9, 1895. Surely but one interpretation can be placed upon the values of/ as found by Lang. It is that th.e wave-length measured is due chiefly to the receiver. 8. Determination of /x. In order to determine accurately the index of refraction of a medium transparent to electric waves by the interferometer it is No. 4] IS/: VI- IXri.HllKoMl.lhK. 15 necessary that uv have an iiUrrtrriin\ (.iirvr with urll dirmcd maxima and miiiiina. The iiilliitnce ot' tlie irci'ivir 011 tlii>.tiir\c is, of coursi-, a iiialUr of no i-oikl';ii except that it ^houlil assist in ^'ivin;^r this ileilnitioii. Neither docs th.e ([iiestion of tlie iiaturi' ol the r.idiation concern us unless (wliat is exceeth'n^dy iniprohahU) the medium in (|uestion exercisi's siKetiw; absorption li)r the-^e Iim^^' wa\cs. To prewnt the radiation uliieli is reflected from tlie surf ices of the plate from reachiii<; the receiver, the j)l.ite should he inclined .it an aULjle to the ra\'s. If the index of refraction is not much greater than unity, tiiis ])rec.iution is uiuucessary. If, however, the plate is kept normal to the rays, its thickness should he varied gradually tiir()UL;h at least a wave-lenL;th and the mean of all the results taken.' In oriler to find the ile;j;ree of accurac)' with which the value of// may be determined by this method, obserxations were ma{t)Ydt\- r [o(/') + /'(/)]V/ where l\t) is found by taking account of the radiation rellected successively from tlie surfaces of the plate. This itself is a complicated expression. i6 (,-. /• mil.. [V,.|, V. iKiiu III wi arc -^ali' in ^.i) iiil; lli.it tliat i K iiKiil i-> diir to an loiilLi- ticii ut' a iliTinitf iRTi
    riY 01 (■llir\(,(i, May I, I.Si)7. w \ ^r JB.5 lyoi 1,1- at tlu- icrc not ant oiy om II-:- The Physical Review. A 'Journal of Experimental and Theoretical Physics, CONDUCTRD BY EDWARD L. NICHOLS, ERNEST MERRITT and FREDERICK BEDELL. XXVI. Auffust, 1897. TABU OF COHTBMTS. FnmttopiM*. Ob tlM CoHTtnlon of BlMtrlo Bawgy in Dlctoetriet. in. Richard THRiLrAix. A Dttwmliutkm of tho SpoeUe SMttUaoo uA Tomporaton CotlBclaBt of Oil la Thin rUna, and the Application of thaao Katolta to the Meaaurement of the Thlckneaa of the OU Film In Joomal Bona. F. L. O. Waoswohth. Certain Pnpertlea of Beat Warea of Great Waye-Length. I. H. Rubens and E. F. NiCHOU. Minor Contribution : On the Relation Between the Onnotic Preaure and the Vapor Premre of Solntioni. A. A. Ntytt and C. G. Aiiett Vote : Alfrad ManhaU Mayer Hew Booka: T. W. WrffAt: Elements of Mechanics, y. S. Amti.- Theory of Physics. IV. Ackrcyd: The Old Light and the New. E. P. Thompson and tV. A. AntAony.- Roentgen Rays and Phenomena of the Anode and Cathode. 'AT. J. Morton: The XRay or the Photography of the Inrisible. Giorgtt Vitomx: Les Rayons X et U Photographle de I'lnrisible. XXVII. September, 1897. TABU OF COHTBBTS. Are Spectra. Arthur L. Foley. Certain Propertlea of HMt WaToo of Oroat WaTo-Ieagth. n. H. Rubens and E. F. Nichols. The Snrfaoe Tenaioa of Water and of Certain Dilate Aqnaona Solatloaa, I>eter> mined by the Method of Klpplea. I. N. Ernest Dorsby. Minor CoatrlbatlOB: A New Method of Determining the Specific Heats of Liquids. Robert L. LUek. New Bottka: IV. A. Stont: Experimental Physics. A. E. DoUtar: First Principles of Natural Philosophy. C. P. Sttinmttt : Alternating Current Phenomena. A. C. Crthore and G. O, Sfuitr: A Polarizing Photochronograph. E. Mack: Contri- bution to the Analy^ of the Sensations (transUted by C M. Williams). Adams : Scientific PRpeis of John Couch Adams. W.G, PUBLISHED FOR CORNELL UNIVERSITY. THE MACMILLAN COMPANY - - - NEW YORK. ^ I m ^ afej^(MW Wg^g UA^'»>^ JJ«i^a-Ma,i Mg^. „,.- --tj