The Determination of the Electric Dipole Moment of Certain Organic Compounds Citation Daily, Charles Robert (1929) The Determination of the Electric Dipole Moment of Certain Organic Compounds. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/n134-1109. https://resolver.caltech.edu/CaltechTHESIS:08202024-204418826 Abstract No abstract. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Physics) Degree Grantor: California Institute of Technology Division: Physics, Mathematics and Astronomy Major Option: Physics Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1 January 1929 Record Number: CaltechTHESIS:08202024-204418826 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:08202024-204418826 DOI: 10.7907/n134-1109 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16656 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 21 Aug 2024 18:07 Last Modified: 21 Aug 2024 18:14 Thesis Files PDF - Final Version See Usage Policy. 11MB Repository Staff Only: item control page

THE DETERMINAT ION OF THE ELECTRIC DIPOLE MOMENT OF CERTAIN ORGANIC COMPOUNDS Thes i s by Charles R ,. Daily In Partial Fulfillment of the Requirements f or the Degree of Doctor or Philosophy California Institute of Technology Pasadena, California 1929 Tabl e of Contents l. Introduct ion . 2. The ory of Ele ctric Dipole Moments . 3. Appara t.us Deve l opment . A. s t abi l ized Hartl ey Oscillator. D. Meis sner Di e l e ctric osc i l l ator . 4 • Experi mental . Procedure e Ao Ptl:ri f ioa tton of Chenioa l s . B. Determi nat i on of Di ele ctr ic Constanta . C. Determinat i on of Mol e Fra otions , Dens i ty, Index of Refra ct i on . 5. Experiment a l Results and Discussion . 6• summar y • 7. Bibliography . 3o 1. Introduction; The determination of t he electric dipole-moment of molecules has assumed considerable importance in r e cent years as a means of r ecognizing and distinguishing various chemical structures . Re cent experimental ork 1 2 in this field by w1111ams ' ' 3 14 • 5 • smyth6 , Errera 7 , ' • 8 . Hojendahl, and others on the basis of the Debye theory of binary liquid mixtures has shown that the dipole moment of polar molecules may be considered as a vector quantity. The magnitude of this moment may t hen be used in interpreting the molecular structure. The purpose of this paper will be to consider certain special chemical types containing CH3 and Br groups 1n various molecular combinations . The development or constant radio frequency oscillators constituted the first part of the experimental work undertaken and data is presented on two types which have been tested. The construction am testing of the circuits used for the diele ctri c constant determinations is des cribed in detail. 2. Theory or Electric Dipole-Moments. The complete theory of the dipole moments of gases 9110 11 and dilute solutions was first presented by Debye • and his treatment of the subject forms the basis for most of the reoant experimental work in this fie.ld. The Debya theory ha.s since been eheeked theoretica l ly be R • de L. Iu-o:nig 12 and Mens L'"lg & Pau11 of t he new quantum mechanios . 13 on the ba.s is C. Manne.b aek14 has also da:r•ived the Dabye equation using Schrlbtiingera. 9s wave meeha11ies e Thus it may oo assumed that it is wall verified by the more recent advances in physical theory . According t o this theory, 1£ a diele ctric is placed il1 an axterna 1 f 1e ld of fores F, we may write for the molar polarization of a molecule which is free to assume a perfectly random orientation 1n either a gas or in a : dilute solution: (1 ) where 19 = diele otrio constant , U = molecular weight , d = density , N =no . of grams 1n a gram molecule= 6 e062x1023 , 0 = molecular pola:r1zab!lity, k = molecular gas constant= • 1.372 x 10 -16 , f = electrie dipole moment of a single molecule, and T = absolute temperature . 0 A'9 is the polarization produced by the deformation, or shi ft of charges, in the structure of the molecule induced by the external field of force e The molecule is assumed to have an elastic structure and since the displa ced electrons have a very small mass , this component of the polariz~tion will be maintained even for very high freq uenoies • 0 B" is the polarization :resulting from the induced moment per molecule produced by the acting rield . It thus represi:m'i:,s the effect cf the orientation of the :molecules in th® .field and is the only term ef the polari~ation that is affected by temperatl.l:r " changes. If now we have two components in s0lut1on, by the ~volume mixture law the m()la.r-- polarization for the solut ion may be written: e .. 1 (2 ) P12 = ~ + 2 • where r 1 and f 2 , M1 and M2 , P1 and P2 refer t o the mole fractions , molecule.r weights ,. and molar polarizations of components l and 2 in the sol ution. 1'he validity of the volume-mixture law for the Claua1us•Mosott1 l aw which is involved i n the derivation of th1s formula has bean checked by Hubbs.ro 15 , Hojendahl 8 , Ph111p 16 , ami others , and :may be assumed to be oo:rre ot. If P1 refers to a non.,.pole.r solvent suah as benzene , then P2 which represents the molar polarization for the polar solute molecules :may be obtained graphically be extrapolating the P12 curve t o infinite d ilut ion as shown by the dotted .f' line 1n Figure f; which is incluo.ed in the exre rimental section of this ps.par . This extre.polation will g ive the molar polarization due to a single pole::r molecule completely surrounded by the nonupolar solvent. The just1fica•t i on of this extrapolat i on and the interpretation of this form of curve is explained by Debye 11 • As shown by Eq. l, the polarization for the pure sol ute i s composed of two parts which may be written: p - ( pt + P" ' ) + 2 2 2 - ~ ( 3) 2 P~ corresponds t o BJ T i n eq . l and represents the orientation polarizat ion produced by the external f i eld and it is this quant i ty that we wish to determine. 2 P corresponds to A in eq. 1 and represents the polar - izat ion due to deformation and is called the mole cular refra ction or the molecular el ectron-polarization sine& only the displa cement of electrons i n the molecular s tructure have a part i n determining it. It may be appr oximated by the Lorenz-Lorentz relation, the proof for which has been demonstrated by Lorentz 17 , Lange 18 , and Debye 11 • P2 = N~ - 1 • M ( 4) ti~ + 2 d where N0 rep:r0sents the i nd.ex o:f re.fraction extrapolated to zero f:r•equeney s.nd. is valid providing no i nfra-red al:usol"pt 1on occurs . Since little is known concerning this 2. absorption , the val ue of P used. may· be in error by a snall amount tiut serves as a fair~ approxi w~tion i n the calculation of the d.ipole moment . P2' represenis tbs pola~ization produced by the displace- ment of polarized or i onized atgms an:i groups ·of atoms or rad icals comprising the molecul~. This quant ity t low freq ... uencies i s generally negligible a.nd has been riegleoted 1n this treatment. P2' becomes zsro i n the solid state be cause the molecules are rig i d l y bound and will not be 9r i ented by the external fie l d . The justification for the omission of 7 .. this component of the polarization is treated by IB.nge 18 • P2 for the various concentrat ions used may be calculated from eq. 2, when it is assumed that P1 D the molar polarization .for the solvent is constant. : P12 • P1 + pl f2 (5 ) Fi na lly , si nee P2 is known from r-e.fractive data ; and P2 i s kn0wn from the extrapolated Pu~ curve obta ined from diel e ctric data, P may be calculated dire ctly f rom eq . 3 . 2 From this value or P~ , t may be evaluated using eq . 1. ( 6) (7) where T =298° K as t.lriis is the temper-a.tu.re used i n mos t of the experi mental work to follew. The veetor t,heo:ry tor the 1nterpretati•o n of the d ipole moment~ as above determined was first suggested for the case of d.1 ... substituted benzene compmz.nds by J . J . Thems o:n19 and approxi mate agreement has been obtained by several invest i gators 4 7 i n c luding Errera , Smyth & Morgm6 , s.nd Will i ams & Schw i nge l 11 The expar:tmental ev:tdenoe av&ilable up t o the present time i ndicates that i n the p-disubst i tuted. benzene molecul e , the resultant moment is EH?,.'U{:'. 1 t o the algebraic ~n.un of t he moment s of the separate polar groups and. th1$ has been verified for p.... bromotol uene within the limit$ of experimE:inta.1 error by the author o The ster1c effect present in the ortho and me t a compounds which makes the oa l cula. t ion of t he angles between the vector dipole-momen:ts i n the molecule sonewhat uncertain ., is also present in the other· molecules consider.ell in this investigation and ea ch w:111 bs discussed in the experimental se ction to follow. 3. ApP!;:rs:~us .Development: The diele ctri c constant$ of the solutions used wer e :measured by means of a high frequency resonance system similar 6 to those described by ~V illie.ms 1 and Sm.yth • The principle upon whi~.h this form of measurenwmt is basiSJo. is a very useful ona in the field of physical measurements anG. a number or applieations have been suggested by Wegsta.ff 20 , Be lz Whiddingtgn22 , Dowling 23 , sucksmith24 , Scott-Taggart 21 25 , o and others a The general pr inciple of the resonance n~thod utilizes the faot tha. t the beat frequency between tv10 oscillators which are o~rat1ng at e.pproxims.tely the same r1~equsney may be aceuPately measi1.red by comparing t his beat i'raquency with a tuning fork , vibrating reed, or other constant lew frequency generator. In the apparatus as used ; ona osc i llator utilized a quartz crystal to maJ..ntain a constant frequency of 9 088x10 5 cyelefJ par sec()nd . '!'he osaillator used for the dieleetr1o measurements waa not crystal controlled be cause !tis used to rn86tsure the changes in t1req,ueney produced by the introduct i on of' a dieleet1:~1c in the mea.suri:ng eondenr;;er. Under normal oper,e.ting conditions however its frequency ,s hould be very stable so that cor:ree.tions ,vould not have to be applied to the results o~tained on the d1ele etrie measurements . 9. Therefore t he f i rst part of the development program was t o construct a sultabl, oscillator for thi s purpose o First will be presented an analysis of the stabilized Hartl ey t ype of oscillato~ which offers considerabl e opportunities for future development , and second 0 an. e.nalys is of the Me iEJsner(> type of oscillator which was .finally adopted for" d ie1&otric rt1$asm•unnent$ ., 3 , Ao Stabilized Hi'l.rtley Oscillatol": Fcll 011v!ng the a ccidcntt:,1 d :i.seover;r by tfi" • Ros e of the Bell Te lephone Laboratories that e. certain L-nped.ance p la ced bet"Y11een the gria. arld the tuned circuit of an oscil lator would render the frequency independent of the battery voltages , Mr . F . B .. Llewellyn of t he sarre organization suggested the fo l lowing analysis for the Ha:l"t,ley os cillator . 'l'he fi>equen~y of any oscillator wi ll depend on a ll of the pti.rameters in the ci rcui t ., namel y , the se l f 1nductanoe t of th@ network, the mutua l inductance M, the capaci ty c, the resists.ne e R of the network , the p l at.a r•esiatance r p of the t ube used ., the grid resists.n ee r g of tha tube , .ru1.d ( runpl 1f 1cation .f.aeto:r for the tube . , the With careful censtrutit i on, all of t h ese para.me te:rs may be considered as ess@nt:!.al1y invariant with the excepti on of r p and rg whi ch of course deJ;end dire c tly on t he battery voltages used ., Assuming tha t r 1a s.rbit:rari ly nade ver•y large wh i ch i s an easily rea lized 6 condi. tion expe:rtw.ental l y , then any s ms.11 mange in rg wi l l have a negli g i b l y am.al l effe ct on the ft'equeney compared with the changes which wi ll occur simultaneously i n rp o Therefore the condition we wish to realize is that ~ = o, where w ~ 2 -rr t . oft.I' 10. In Fig . l e have t.hs schematic form of the circuit to be consider•ed , 57118/LIZED Hl'lfiTLE.Y CIRCUIT ments i ndicated for connecting in the ootterieaa Us 1ng the e~!"ent, dire et i ons as shm.rn by the t.trrows, we ci;i..n write the circuit equations ( eg 0 = :t1 (rp + ~x 7. - I 1(Zl . ,t + ~) + I2Zo ( 9) ❖ I2 ( Z.2 + Zm) ( 10) z 0 c z1 • z2 + z5 + 2z. Zl (8) Z1) + I2 ( z 1 + zm) = R1 + 1 w t 1 :: A + !B ZX :t-! •1/w CX z2 :.e R2 {· 1 w L 2 A ~ Rl Z3 ~ R3 ... 'l/ 1.AJ C3 B ·~ w t 1 + w ½ + 2 w M .. 1/ w C3 + R2 + R 3 Z ~ 1 1, \) M m Solvlng equat:tons a, 9, and 10 1 e liminating eg, :t1 a\"ld I 2 i> and substituting thf3 equi "J'ttlent qu8ntit:h~s for the othe:r> ,,.ariables as i ndicateil above f) we obts.1n:, (rp•1/wc,,J (A+ 11l) + (R1+1wL1 lff<r +l) R2 +R 3) + t(<t- +l )wL2 •l/w c 3)] ,;,»1w M[ t R3 ... 1f•Ct- 1•l) w M + r / {,l) c 3 }J :: 0 ( 11) The x-ea.l part of eq . ll gives t.he conditions under which osc i lla.t 1ons will occ"t:tr 1:n. the oii..euit while t.he :bnlll.gina.ry par-t g i '\"GS the conditions W'.l?,ieh determine the frequency or os ci l la t i onn and it is thi s second part tbat we are i nteres ted in. 0 11. Equating the imaginary part t o zero we obtain~ •A/ w Cx, + Brp + Ri f ( t +l) w½ - 1/ ..., c 3) - rW MR3 + w t 1 { (t" +l)R 2 = o If changes in rp 1 which are caused by battery voltage f luctuations, are not to affect tlw frequency of oscillation, then B = O is a necessary conclusion . Therefore: which leads t o ~ w = l/c 3 ( t + L + 2U ) ; 2 1 ~ (13 ) w.,. Eq . 13 proves the very important fact that if the frequency of oscillation is to be independent of the plate resistance , then the frequenc y of the oscillation will be the natural frequency of the tuned circuit . Solving t he remainder of eq . 12, substituting for w its value as obtained in aq. 13 , we obtain: ex= z C3 (Rl .,. R2 + R3 )(Ll + Lg + 2M) . L1 f(r +l)R2 + R3 ... R1} + L2 t R2 • M(2Rl + t R3} (14) stabilizing impedance which will make the frequency independent of changes in battery voltages . In pra ctice this equation can ba materially simplified. For example , let: "' Ll = L2 = Lo Therefore: Cx =: C3 and a still M = (2L 0 i Lo + M) M ( 15 ) f~ther simplification results if o. ex ::,:: 2C3/L (16) The result or eqo 16 may be checked very readily . A Hartley circuit s i milar to Flgo 1 was constructed having a natural frequency of 106 cycles per second, c3 = 300 mmf, and ; ~ 8 for the UX 201-A tube used ., These quantiti1eij when substituted in eq. 16 yield the result that ex= 75 .mmf is the capacity$ which if inserted in the plate circuit , would render the frequency independent of battery voltages. This value was found to hold experimentally within the limits of error Qaused by neglecting the other variables in the circuit., tlspeo !ally t..lie intar11al tube capacity , ar1d so may pe consid.ered a.s a complete checlt on the theory , In or(ler to adjust ex accurately to the proper value , a switch was provided which would increase the pl ate voltage by an amount 4 Ep = 4 . 5 volts o Having set ex at approx i mate ly the right value , the cl1ange t.1 Ep would cause some corresponding change in frequency A t . Va l ues of the condenser readings ex were then plotted against ~ f for this constant ~ EP and the point found where ll f = O• .o r could be made to change sign by continually varying ex on either side stabilization. of the point of The adjustment of ex t o the proper value was very criti ca l and the s lightest error in setti ng would result in enormous frequency changes for the .Ll Ep used i n the setting . This c:trcuit was not as useful as had been expe cted be cause after stabi lizat i on had once been effected. , uncontrollable variations in other parts of the circuits would , i n an hour or two~ vary the constants so that ex &s adjusted no longer had the proper value for stabilization .. When such a condition occurs , the resulting IJ f is much larger than if the i mpedance Cx had not bean used in the circuit . I t is the small changea 13 • produced in the circuit incident to the handl ing of the apparatus while making the d ielectric mea.sure:rrents which change the various circuit parameters and so cause the unsts.b ilizs.tion of' the apparatus . For this reason such a c ircuit was not used in the final measurements . although the principle on which it is based is perfectly sound . Developtrent work which has been done on this type of circuit at the Be ll Te lephone Laboratories in conjunction with oscillators t o be used in carrier frequency telephone apparatus have proven the value of the theory espec:ia lly s.t low frequencies .. The carrier channels have a minimum separation and it is absolute l y necessary that the frequency of each channel remain constant 1n order that cross-talk and other @ttenda.nt diff icul t i es may be avoided . Such an oscillator constructed to operate at a frequency of 35000 cycles per second remained absolutely stable , with changes in frequency of the order of only 0 .. 1 cycles resulting from a change in the pla 't,e voltage of 50% . At such a lo frequency, the i nductances and capacities a.re much larger less than those usaQ at radio frequencies and thus are correspondinglyA influenced by small changes in capacity to ground , tube capacities, etc. which so markedly affect osc1llato1. . s operating at radi o frequencies. I t is possible however that with quite different circuit constants , a better approach to permanent stability can ba made at the higher frequencies . 14. 3. B it lle issner Di 1 ct,ri c Oscillator: UJ<2.0IR <R.Y:JTAL CCIVTROL O :>CILLflTOR .DIELECTRIC o::ictLLl'fTOR /IMPLJ FIER rij. ~- The Meissner oselllator 26 as shown in Fig . 2a was final l y chosen for dielectric m9asurements because or its inherent frequency stability. Changes in plate voltage am filamsnt current have an almost negligible effect on the frequency . Tests were conducted and it was found that wi th varying values of the grid leak r-esistianee from 105 t o 106 ohms, the maximum b f observed was 20 cycles in 106 eycles for changes or pl.a. te voltage from 30 t o 90 v olts . Changes in filament current had a correspondingly srrall effe c t on the frequency . 1'he coils L1 , L2 , and t 3 were of a special construction suggested by Dr. Frayne who had used a similar circuit a.t tha General Rad.io company . Fifteen turns of #22 DCC wire were wound on a 3-inch bakelite tube for L1 , the grid coil, then fifteen turns for L3 , the osc11Jating coil , ·then 15 turns for !1-2 ai a.ml sa on urltil t 1 :::: 15 turns ., L2 = 30 turns , Ls= 46 turn~ . This filtagg@red arrangement pr©vides fo~ a 10 distributed coil ce.pacity with high incl.uctive coupling between the plate and oec111Qting coil$ o c1 1s a aerie~ tuning aond~nser of 250 mmf capacity.; Cs is a G&ner>al Radio Pi"@cis i on condenser , Type 222 , 0~1500 mmt . It was provided with a 25 poin t c~l1b:ro. t i on chart $0 tba t diffe!".e ncea or capacity could~ read dia:eetl y to within 0 .06 mm.f .. a oone$ntric cylinder condenser> used for the liquid diele ctri c o It was const1"ucted of heavy Wlil. lled bra s s tubing with bake lits end blocks to keep the inne:r'" cylinder accurately concentr-io with the outer cylinder * The effE,otive length of' ths cylinders was 8 .9 cme with a spacing between them of 1 .3 IllJM . Its capacity ix:1 ai:.r• was 91 .5 mmf ~ The spa ce be t,.veen the plat eE contain~d all of the d iel~ctri~ to be m.oasured so that, ine&uiurerre: l'.};t~ could be m3.d@ on as srre. 11 aa 25 ec of solution . This is a distinct ~dvantS1.g@ bi that it, reduc~s the quant:H,ies of materita.l!?l that must ~ pUJ:> i fiad to &\ minimum e The liquid. to b@ 1naaaured coul d be pC>l.lred in a t the top of th© condense?• ·thN a gl&sa tube provided ·:d th a stopcock ~ This ato19cock was closed while th~ lJ.cru. id was b~ing m@Sisu:red, thua preventing evaporat i on which would. have continuous lir change ~ the mol e fraction of substance in sol ut ion e After the measurement of the dielectx)ic constant had hsen completed , the liquid wa.a aiphoned off' thru s-. tube lea.ding out ot' the bottom et the condensel!"' .. The measuring oonda11ser, Cd, was submerged in a water bath ma1ntab1ed at a constant temperature by means of a liHH1aitlve mercury therm.ostat . This water was c:J.rcrulated continuously thru the condenser bath as well $.S thru the bGi.th us ct in con.'rle ctio:n w:1.th the density 1nea.suremserrt.s , e.nd finally thru the refra ctometer . The circulation was ~a.pid. , being ma i ntained by :a. water pun1p., s o that a ll t hree pB.l"ts of the apparatus were kept at the same temperature .. ".!'he dif'ference in tempera tm•e between the var•ious parts of the system was never greater than 0 .1°0 and the absolute temperature was kept a 'b J:tnpt at 25 ;.t . 05°c except for the measurements on p- bromotol uens 1n which case th.El tempera tu.re · was maintained at 27 .5°., All thermometers used were carefully si tandardi zed • The swit~h Sin Fig . 2a was used to disconneet Cd from the circuit . In al l cases, measuremen ts were made on C6 with s open and cl osed alternately . In this way any change in '.the fr• quency of the oscilla·tor caused by accidental ja.rring of the cases ow ld be detected. 1.nnned i ataly and t,h® di~le ctr:tc :measurements repaa ted if nooessa1"y o 'l'he oscill ator was a llowed to ope:rate continuously f or several weeks while the necessary observations were being t,aken. During this interval there were noted only sTI\Qll frequency chang®s which were probably due to a slight warping of the t~bl@ and th , circuit suppox~ta, as well as s:im.11 changes in the relative posi t 1on of the shielding with respe ct ·to the circuit o 17. During · the course of any individual dielectric determinat ion, the .frequency did not. change by more than ten eycies in a milli<tm , which was we ll withi.n the 11..mit~ of' e~perin:ental ePraor in the determinati.on. et the di0lectrio eornstant e The het©redyning beat nots oscillator show-r1 diagram- type of cireui t mainta ini a const&n·t i'requeney , dep-endent almost ·entirely on the p:ropert.ies ot the m"ystal H:,self ~ Extensive studies have been conducted in the field. or 2'7 28 piezo-elect.riei ty by G" W. Pierce , w. c . Ca.d.y , and others , and they ti..a. v@ shown that the frequency of such oscillators is constant to one part in a rni111cm. if t,h.e temperature a pressu~re on the crya tal., and load a.rs kopt constant. The temperat'UI'e CO®.fficient is a.bout 1 cycle in 50000 pc-n·· degree. Since the f'undrunental f::requenoy of the c1"ys tm.l used wa.s 9 .88 x 105 cycle~ per second , a or.i..ange 1n temperatwe of 1V c . would. oause a .feequency change of about 20 eycles . To minimize this effect , the ent ir~ :poo.m in which the apparatus was housed was thermostatically eon.trolled to within O .2°C . This precaution l.U~ew-ise p:revented slow frequency chang'ils in. the dielectric oscillator due to expansion or contraetion of the coil form.s tJ condensers, and. supports • '.nie circuit 142c 2 iw.s tuned to the f7tequency of the crystal and this a.djustnwnt could be ehecked by noting the decrease in plate cUfrent. when the circuit was adjusted to resonance . The dielectric and crystal controlled oscillators completely shielded . ere A brass rod A extended between the two ~~cillators t o provide coupling between them to the beat note a.mplifi$r o This roo ext,ended into the oscillator cases for about three inche s and the coupling thus provid~d was so eak t.hat the frequency of the two oscillators oool,d be adjusted t o with in one cycle o.f each other w-i th no synbhroniza tion taking place . ThG coupling rod A wae in turn connected to a tuned circuit c3 t 5 1n the first stage of the beat note amplifier a.a shown in Fig. 2o . This tuned circuit increased the voltage drop across the grid of t'.h.e first tube which acted as a detector to separate too auAible beat note from the radio frequency input. Thie audible fre quency 1J1rn.s then amplified in the ci:r•ou :tt shown about 20000 times and the output obtained from the secondary of the tn.nsformer , T .. R4 is a variable high resistance potentiomater used to oontroi the amplitw:le of the output voltage Which ras applied to one pair of plates on a Braun oseillographic cathode ray tutoo . The other circuit constants were: c4 = 0 . 1 mm.f fixed condenser , en= 0 . 0005 mmr , ... R1 = 0 .25 megohms , R2 = 2 megohms , R3 == 1 megohm, R5 = 6 ohm filament resistSlnoes . In orde~ to provide a comparison source of frequency for the beat note , a vacuum tube driven tuning fork drive was constructed as shown in Fig . 3 . in e. heavy steel frame . TF is &. 435 cycle fork mounted The coupling and driv·ing units to the fork were provided by two Western Electric phone units , P2 , adjusted eo that the poles of the i r permanent magnets were tuning fork . The d . o o P, component ot plate "" TO f3RAVN TU8[ 1Zur>rent t:n the l.a$t which would be TUNIN6 FOR/{ DRIVE e,ceesaive .for the phone unit , 1' • 2 ,c,j , 3 The1,.efortl a. Gen@r>al. Re.dio Speaker Fil tell"', S .F . was inserted so that only t he a .e o canpon.ent of the p late ct~rent flowed thru the plate driving coil, The output fr>om this genere.tor wa$ obtained. across e.n impe.danee P1 ~ sui trably controlled by a Vl\l. rjab le resistance R8 • This output wa.s conn®oted to the second pair o.f plates i:n th0 Braun tube . It i s output volte,g0 of 75 to 100 vol't,s on no load. could el\s:l ly be obta.ineo. , aa we.a the case also 1ui th the beat note ampl1f ier .. The othe~ c :Jroui t eo-nsta.n.ts in the tuning fork d.ri ve were, R1 = 0 .25 mego.hm$, R-z :::: 1 megohxn., Re· = 6 oh.m fil11unent <J t) re.sistol"'~ .11 R0 :r. 5 megohm variable grid le aJ.t to control too input. vo lta.ge to the grid of the second t.ube ~ R7 = 0 . 5 megohms , e.nd c6 :::: 1 mfd coupling cono.ensers. 20 .. 4.. Exper imental Procedure: A. Purification of Chemicals: All chemica ls used were obtained from the Eastman Kodak co . in a very pure form .. Benzene: Ee.oh was further purified as i ndicated . Thiophene free benzene was stirred with several portiona of eonc. r.1 2so4 until no further d iscoloration took place o It was then washed with distilled water and dilut Na.OH solution am stirred with clean mercury foro 24 hours .. 11.'he benzem, was then dried over p2 o5 , distilled , and ke p t '7Ver metallic sodium. B .P .. 745 '79 .4-79 . 6°C . Benzene which was twice .f'rozen out over clce b8£ore distilling showed no vt1riat:ton in dielectri o constant. Toluene: Toluene prepared 1':rom sulfonic acid v1as mtirred with cone . H2S04 , washed with distilled vtater and Ne. OH., and allowed to stand over mercury for two days . It was then drisd over P2 o5 , d istilled 0.nd kept over :met.all:tc sodium. B. P . 745 l07.9-108 .1°c . BromobenzenE?,l Tti..e E o K o Co . produc t was dried oviar P2 o5 and distil led . B.P.146 l52-l 53°C . J?""'Bromotoluene: The E. K. Co . prodi1e t was er-ysta llized three ·ti m.as a.rxl dried. over oone . H2so4 in a dess:leator . M. P . 26-27°c. n ... Propyl Bromide; The E . K. Co. product was dried owr P2 o5 and distill ed. B.P o 68.6-69 .4°C . 21. 4 . B. Determine.ti.on of' the Dielectric constant . In maki :ne d i e lectric constant deter minat i ons with the previously descr i bed apparatus , a pair of phones connected to the amplif:ter served to dete ct the beat note . Vlith the switch s open ., ths beat not,e was adjusted by varying the condenser C8 t o B, frequency approxi.71la. te l y equal to that of the tuning fork . R4 v1as ad.justed so that the defle c tion on th61 Br~aun tube was about 3 ems . The o.efle c tion s.t ri~t angles t o the first , and produced by the tuning fork drive ., was adjusted to about the same value by the control.a R0 am R8 • If Cs i s then var :ted carefully unt il the heat note exactly equal s that of the tuning .f'or>k, the resulting L:tssajou f i gure observed :ts either a straight line or a circle depend 1ng on the phase r-elatione between the two freque ncies . Chan~s in freqten.e y of the beat note as small as O . 05 llycles can be readi l y dete cted , correapondi ng i n the present case to a change i n frequency of tha dieleotrio oscillator o.f one part in 2xl07 ,, which approaches the erder of accura cy obtainable· vrith e.n interferometer o While such prec isl on is net required. in the msasurement cf the dielectric oonsta.nta of liqu i ds , it would be required i n the accw..a.ta dete1"1ni nat ion of the d iele atric constants of gases 111 Closing switch S i ncreases the capac ity of the resonant circuit and in order to bring the beat note ooek to its ori g ina l value , Cs is readjusted . The change in 0 8 is t.hen a measure of th~ capacity of Cd plus lead capaci ty . If a diele etrio other than air is now inserted 111 Cd " C8 is a.gain adjusted to resonance and this seco nd change i n C8 will be proportio:r.1.al. to the d iele ctri c conste.nt of the l iquid us d. . In the present work , the die lectri c constant of benzene -was Wied to calibrate the ag,ar•atus be cause it was to be used as tl e solvent for all of the liquids Whose moments were to be determined . AbsolutG determi:na:tions or foi"' benzene have 1 29 been made by Williams , HartshoPn , and others , am a va lue =2 .282 f'or benzene at 25° C . has ooen used in thi s i.l'lork. of e Si :nee the heat. frequen cy was a l ways made equt\ l t o that c£ t }-,e tuning fork , the fundantental frequency D and t hereftT' e the t.ota.1 capacity of the d:i.eleetr1c oseillator w~s r1scessarily kept conste.nt for ull i:neaaurements since no othe1-> ebJllnges oocured i n the circui t . Tb.e equations for dete:r-m! n i ng the d.1.e l e etric constant ot' any liquid i n terms o.f the d i ele ctri c consta.nt of benzene might, t hen be W:t'i tt.e:n as follows: Cs2 -= n tt u n Cs3 -· f or unknown . therefore: 0 s2 + ~l ed - K Oss + e 2cd :r. K e2 = l + (e-~1 ) J. C Solving we obtain: .., C al s3 0 s 1 - 0 s2 (17) Eq . 17 was used i n the cal culati,on of dielectr ic constants of t'he solutions in terms of e 1 == 2 .282 , the d i e l. constant cf benze1'te • A:f ter sa ch measurement , the condenser- was washed w:1 th benzene , pum.ped out for ter1 minutes , su1d dry compressed air p assed thru it to remove a ll 1m.pur1ties whiah mi ght contamina t e fut1.we so l ut ions used . All compounds were readily solubl e i n benzene ancl no <U·f fieu.lty was experience in checking readirg s . 4 . C o Measurement of Mole Fra ct i o11s, Density, Index of Refraction The compounds used were to be clissolved in benzene 0 e.nd sin~ most of these solutions d i ffer from ideal solut1ons 0 particu.le.r-ly at the l ov.ror coneem.tra t.ions , only these values wea:•~ s tud1ed. bece.us e lt is the val ue o.f the polar iza t 1o:n a t infinite dilution that :ts of interest . 111.ea.surad was carefully lreighecl. i n ~ Tt.e liquii:J. t o be volum.et:r•ic flask f i tted benzene had been &..clded to mtJ1.ke the total volume of sol ut ion approximately 50 CH1.. The mole f'r~-e tion of the polar liqu id in benzene could than be caleulsted very accuratel y o Density measu.reme:nts on the solutions were made by t,he usual pynknometer method immediately ~ft.er the determi nat ion of the rdi eleetri e const,a:nt . The Jl)ynknometer used was provided with grou:nd•in th@rmomater , and e. glass oap with ground jo i nt · to fit over the capillary stem to prevent evaporation during t he ti.me requirEle. fol' weighir..g Ii The temperatu.re of the bath in whicl1 the pynknometer was hung was regulated very carefully as has bean deseribed in a pr~vious seotion . The valu0s of dens :lty obta.1ned agr eed verry we ll with va l ues taken by other observers and with va lues calculated from dat a in t h..a I:nternat i onal Critical Tables r>· Vol. III , p 27 ·• Tabl ~ l o Benzene M@$.Stn~ed Cal e • I .c .'r . W111:tams ( l ) Smyth (30) Bromobenzene Tolu~ne 0 .8735 1 .4885 0 ;,8623 .8734 . 8731 1 .488? .8610 .8593 1.4490 1 .4479 Measurements of the index of ref:r-acti on of, the l i quids used were mad~ with an Abb~ refraot.om-eter us L"'1.& a sodium . flame for illumination . In calculating the eptice.l pols.:rization ., the i ndex of refraction should ~ extrapolated to ze1~0 frequency . In order to do th:ls s.aau.rately, det.~nmiYl...at:tons of the index of lf:!efraction would hava t o be macte l::'l.'t various wavelengths before t,hs ext.rapolm. tions could be 1'118.de with certainty. This would !~e qu1re e. Pul fri eh refra ctometer "'Jlrh ich was not a.vt>-1lable . Wi t.h this :instrun>nti, the diaperm on o.f 11.qu'◄d.$ can be measm,ed with <ions 1ders.b1e aocuraey .. 'l\ble 2 compares tha date. taken from an average of ten :r•eadings o:f' nd with those obta5.ned by W1lJ.iams 1 using a Pulfri ch refractometer . Benzene Bromobenzene Toluene Measure cl l. .49784 1 .. 55702 1 .49398 Williams 1.49780 1 .55711 1 .,49392 _l?,.• , Expe:r:tmental Restilt~, and Di acuasio1_1; Determine. tions of tli....e dipole-moment of bromobenzena and toluene 1!Jera first undert.aken to obtain the pl•obabl e value of the moment of Br and CH 3 groups singly bef-o re fH')mbina t ions of the two were studied . The f oll0tvi ng pages g3v e the results of tJ.1e determina:tions oxi. the various compounds studied with a.n ane,lysis of etich o 25 . Table 3 presents the results obtained for the bromobenzene-benzene mixture . Figure 4 shows the variation of P12 wit h the mole frs. etion of c6H5 Br in solution. Table 3 . Bromobenzene • Mola Fract. C~H 0 d25 l,0000 0 .8735 1 .49784 2 .282 . 9528 . 9068 1.50093 2 .458 29 .49 84 .75 . 9002 . 9442 1 .. 50444 2 ,.640 32 .16 80 .96 1.51094 2 ,972 36 .71 77 .24 3 0727 45 ,,85 71 .28 p2 4 .8027 . 5711 1 .1622 1 .52542 . 0000 1 ,4885 1.55702 93 . 25 .., ... 2 = 33 ,,96 t = 1.69 x 10"" 18 esu P Hojendahl 8 obtains a val ue of r=1 . 56x10•18 ; ca lcu- l at i ons from di e l e ctric constant measurements of Kerr 31 give 1 .71xlo- 18 ~ t= wh ile Williams 5obtains a value of 0 ~ ···--0.9 13R.0"1B£NZ£.NC : 6 P-13R.OMTOLUEN£ I -o.s-· ··- -·J _ j _______ _Jo.s 0.7 0.6 MOLE. FRt'ICTION BENZENE C= 1.5 l x10· 1~ These values are i n agree~ ment within the limits of experi ment a l error w 1th that obtai ned by the author .. The moment obtained for c6H5Br represents the moment of Br a lone since the remainder of the benzene ring does not contribute to the moment . The moment of benzene has been found t o be zero by several 32 033 observers using cs 2 , cc14 , c6B6 ~ and c6H14 as so l vents • . The faet that b~nzene l1a43 a zero moment agrees also with expef'im.ents on p-dini trobenzene , sym... trinitrobenzene, and. sym~tribromobenzene 8 , all of which yield dipole moments which are not larger than the ex!13 rimenta l error e In this connec tion it should be mentioned that since it is 1. 2 i s propor t i ona l t o t , a small value of P2 which is so sfflll l that P2 that is measured , and since P VS!.lue of ( CQ'Ft>esp on.ds to & it :m:ay not b~ distingu:tshs.bla from zero. P~ enters into the 2, both of which $.I"e equation as "· difference between P12 e.nd P experimentally determined quantitie.e ., and an error i :n e ithar 0 pEl.rtiov.ls.rly in P which is obtained by extrapolation , will lead 2 t<.> a large 8l"f"OI" in r for smal l va l ues of this differen ce . Tabl e 4 . Tol uen • d 2 or.; 4 n25 d e pl 2 p2 1 . 0000 ..8'735 1.49784 2 .282 26 .75 35 . 90 . 9072 . 8720 1 .49725 2 .294 27 .. 42 33 .97 .8019 . 8703 1 .49597 2 .3 14 28 . 29 34 .52 . 6025 .. 8679 1 .49572 2 .331 29 . 61 , 33 . 95 . 0000 .8623 1 .49398 2 .387 33 .75 33 .75 M .F . C6H6 P2 :: 31 .08 r ~ 0 .48 X l0•l$ 6SU For tol uene j the extrapo• 3~ f - - - - - + - - - + - - - + - - - - - i - - - = -""- - - - --1 lated value of P2 ie some~ what uncerta i n and the final value of l may b$ in e:rror by 10% • va l ue for The above t however agrees with values obta i ned by other observers • T()LU £NE. 2 5 '-------L-----'--------'--------'-----' /.C 0.8 0.6 NOL£ FR/iCT!ON 0,3/ 0.4 BEN:ZENc C Wi11i ams 2 obtains l = 0 .52xlo•lS while Smyth & Morgan6 obtained a. value of t' :;.-: O .4Qxl 0 -18 • The d:tp.0 le-m0ment. obtained for toluene P$presents the m0rn.ent ct the OH 3 gr,oup attached to the benzen0 ring . From the data obta.in~d on toluene an.d bl"'om0benzene " we · can now discuss the results obtained. for combinat ions of CH3 ·and Br in the case of the fol lowing compounds. M .F . C6H6 d 27 •5 27o5 nc!l 1.0000 . 8708 . $}550 e pl2 PS 1.49784 2 . 280 26 . 80 118.4 . 9002 1 .50010 2 . 5 14 30.64 112 .1 .. 9039 . 9363 1 . 50330 2 .765 34 ..41 106 . 0 .8352 .. 9816 1.50?80 3 .o~n 39 .13 101.6 . 5018 1 .1721 1.52619 4 .466 56 .86 87 .14 eOOOO 1 .. 3993 1.54768 -- .,. ... ... .... 4 P"2 = 24 .51 r= 2 .13 X 10 -18 esu If the steric eff@ <rt is negligible i n the para form of the di-subst i tuted benzene ring , then we would expect a stri ctly additive result for the moments of CH and Br . 3 18 (l.69 + .48) x 10"" = 2 .11 x 10... 18 by ve ct or addition t = ':;-' 2 .13 x 10- 18 measured . The agreement obtained ie remarkabl y close . The sign attributed to the group pols.rity results fr om th~ charges on the atoms forming the group and the sign of the rnoment is take.n as po$1 t i ve When the ve c tor passes thru the negative to the positive pole of the group and points outward from the center or symm$try for the molecule. 28. In brom0benzene , the bromine atom is negatively charged with respec t to the adjacent carbon atom i n the benzene ring e In toluene the hydrogen atoms are positively charged with If this is the respect t o the carbon atom in the CH3 group. /JA. ! ease , then the moment of p-CH3c6H4 Br should b A check on the correctness of this conclusion 0 may be obtained by analyzing the results of other CH3 the alge braic sum of the moments of cH 3 and Br . i W1111ams 4 obt a i ned a value i nvest i gators on similar compounds• ot t = 3 .90 x lO~la for p-nitrobenzene and t =4 .50 x lo-la f""""' p--n 1troto1uene ...,£ The difference betw en these two val ues , ie (" = .60xlo-lS should represent the moment of toluene and agre . s within the limits of error \Vi th the value obtained by Williams and, the author• for t oluene . Wh ile agreement is not as close i n a ll cases , the general theory of' ve ot(>r add itivity is seen to hold. Other- cases which ha:ve been treated with success are p-chloronitro- be nzene , p•bromonitrobenzene, and p-ni trotoluene • Table 6 . M.F . 25 25 n-Pr>opyl Bromid@ . CsHa d4 nd e p12 p2 1.0000 .8735 1.49784 2 .282 26 .75 107 .1 .9654 .8939 1 ,,49438 2 . 512 30 .0l 99 ,,85 . 9193 . 9118 1 .49199 2 .698 32 038 96 . 51 .8995 . 9209 1 .49062 2 .796 33 . 57 94 . 62 .8245 . 9567 1.48535 3 .193 '37 s93 90 045 .6734 1.0283 1 .47497 4 e0 13 45 .18 82 .18 . 0000 1.3455 1 .43197 -- -- -- P"2 = 23 .'71 t = 2 .00 X 10- 18 esu ss , - - - - - , - - - - . . - - - - - - - , - - - - - - - - , - - - ~ o i n -PROPYL IBROMIDE 6 /50- .. .. ~s L _ __ '° _ i __ __ J __ _- - - - - - ' - - - - - - - - " - - - - - - ' aq ~B ~7 ~6 0 . .5 M()LE FRAC TION BENZ E NE The struetu:re of n ...propyl bromide might be 1~apr$sented as shown o In this case we woul d not neeessar-i ly expe ct th$ CH 3 and Br grnups te l i e i n a plane • Ho,,ever tha mome :ntra of these two groups should be approx imate l y addi t i vfl and th$ val ue of r = 2.00 x l OulS i s fa i r l y i n a ccord with the swn that would be obtai ned by ve ctor addi t i on, name l y 2 .17xlo"" 18 • This difference may be par tiall y expl ained by assumi ng that there is an angle between the vector moment fo:r CH:3 and Br but the i nterpr•etat i on of thi s angl e , which is smal1 8 would. be rather uncerta i n . The onl y check on the results obtai ned for n-propyl bromi de i s :found i n a calculat i on made by Hojendahl 8 on the b~sis of di e l ectric measurements on vapors at 100° made by Pohrt 34 i n 1913 ,. Ca l cule.t i on gives a val ue of f 2 o15x1o... 18 = on the bas :ts of' these e,a rlier e.nd poss i bl y somewhat i naecura:te bri dge measurements . The check that is obtained however 1!1 close cons i dering that nne s@t of 1l'l$asurements was made on cUlute solut i ons and. the other on the vapor of' the compound e 30 . Tabl ia 7 . :t.sc-Propyl Bromide . 425 n.26 d e 1.0000 .8735 le49784 .9358 .9013 . 8988 M.F . p12 p2 2.282 26 .75 124 .3 1.49231 2 .667 32 . 07 109 ,6 . 9186 1 .48950 2.884 34 . 68 105 .1 .7932 .9660 1 .48128 3 . 53 8 41 .4 3 97 .74 . 6097 1 .0470 l .46720 4 .758 50 .76 88 .27 . 0000 1 . 3040 1.42267 ...... ...... ... ... 4 C5H5 The s t ructure of i so•propyl bromide might be represented as shown . ":' ~. /7 Ac cording te this I view , i f the angle between the I thNe vectors were ea.ch 120° and all lay i n on~ plane , then the resul te.nt moment for th, molew le would be that du ffe ctive l y to one CH3 ano. one Br group in • • 18 a lin • Thi s sum has been shown t,o be 2 .17 x 10 on the baais of the measurements on tol uene al"ld bromobenzene 0 and is in close agreement with the va l ue obtained above for this molecule . The data on ethyl bromide is present.ad on page 31. h j? -7.. . ti I /3A-~-C~ 30 I ti ETHYL /. () 0 . C/S BROMIDE 0. 90 0 ,BS MOL£ Fl?f'ICTION 13.ENZENE. Table 8 . M.F. Ceff6 d:5 n25 1.0000 e87S5 . 9514 Ethyl Bromide . e Pl2 p2 1.49784 2 0282 2S .75 101.7 .8966 1.49423 2.648 30.~W 97 ./73 119212 ..9125 1.49210 2.687 31.75 90 .23 ,8746 ~9344 1.48904 2 .-915 34 .16 87.82 . 0000 l .4490 1 .42128 d p'D == 19.09 2 t = l.9S X 10· -- _.., -- 18 eau . If the eonf'iguration a s shown on page 30 is assumed to 11e in a plane, the moment of this compound 1.vould be the ·same ... 1a as that obta.ined for p-bromotoluene ., t' 2 .13 x 10 , and = agreement is obtained within 7%,. Hojendahl calculatea the value of the dipole m@rnent for· thi/3 compound to be 2 . 09.xlo"" 18 u~ing the vapor data of Pohrt as mentioned in the case of :n-propyl bremide . ( -::; 1 .86 x 10 -18 Smyth & Morgan 30 obtain a value of tor ethyl bromide in a h0x:ane solution. They u.-sed a bridge method for maasuring the diele ctric constant and a lso investigated this compound at se ..tTeral d i.ffa:rent temperatures . The author is !ndebted to Dr . R. A . Millikan and Dr . L. c. Pauling f or their assistance in the carrying out or this woY•lq to Mr • F. J . EW1ng f or the care tha. t he t ook in making the chemieal.K purifications; and. to Mr . W. C . Bl"UO$ for experimental assistance . 6 • Summ.ary : A bri f review of · t he Debye theory of dipole-moments is p~sented with a view toward interpreting the structur8 of certa i.n organi c compounds • A discussion of a new theory f or constant frequency oscillators is p~sented which will be of i nterest in the adaptation of the apparatus to the measurement of t he dielectric constants of gases . A constant. frequency diele ctric oscillator has been developed toge ther with the associated he t erodyning oscillator, runplifiar , tuning fork drive , and Braun tub oscillographic c ircuit. The second part of the inves t i gation considers the relation between the dipole moment and the structure of the molecule . It is found that the group moments of the aliphatic compounds investigated, name l y n-cH 3cH 2cH 2Br , iso-(CH3 ) 2CHBr , and c2H5Br can all be ascnibed to the ve ctor addition of the moments obta ired for CH 3 and Bl" in the mono-substituted hydro carbon, benzene . n•propyl bromide i s a pliable molecule , ie free rotat i on can occur bout the s i ngle bond joining the carbon atoms and s o accounts for the moment obtained whict, is slight l y lower than t he algebraic sum of the separate moments . The moment obtained for p-bromotoluene i s i n excellent agreement with the ve ct or addition Which would be expected for the case of the plane hexagonal benzene ring. A 11st of the dipole-moments obtai n ed' includes: Molecul,s Bromobenzene Toluene p•bromotoluene n•Propyl Bromide iso-Propyl Bromide Ethyl Bromide t xlO 18 e. J . u. 1.69 .,48 2 .13 2 .00 2 .20 1 . 99 1;, 2 -e 3a 4• 5. th 7. a.. Will i ams , J . W. and Sr>e}'l...ma , I .. : J ..A .c .s . 8 4 8 8 "· •" " n . 49 , to H It tt 49 , " "- Schwingel , C .• : 0 50, " " Krchma , I . : " 50, Smyth, C . • P. and Morgan, s. o . " 49 , Errera , J . : Compt. Rendus , 1888, 1676, 24 16 , 362, 2 350 , 1030 , ( 1926) • ( 1927 ) • (1927) e ( 1928). (1928) . ( 1927) . 182 , 1623 , (1926) . Hojendahl , K . : StucU,e.s of Dipole Moment , 28 ., (1929) . 9.· D~bya , P .: Phys . Ze i t ., 13, 97 , (1912 ). 10 • 11. 12. : Ve:r'h . d . D . Physik . Ges •., .15 , 777 , (1913) . ; Ha.:ndbuohderRadiol og i e , (Marx): 6 , 597 , ( 1925) . Kron 1g , R ·• de L .~ Proo. Nat . Ac&A. ., •12 ., 488 , 60S , ( 1926). Mens i ng , L . & Pauli, W. : Phys . Ze i t ., 27 , 509 , (1926 ). Mannebaek , c . : Phys . Ze 1t ., 27 , 563 , (1926) . Hubbard. , ; Z o Ph . Ch ., 74 , 207, (1910 ) Philip , J ._ C .-: z. Ph . Ch . 24 0 18 , ( 1897) . 13 . ;J.4 . 15 . 16. 17 . 18 • 19. 99 c, ~orentz, H. A. : Ann . d . Phys i k , 9, 641, {1680) . Lange , L • : Z • f • Phys i k , 33 11 169; ( 1 925 ) • • Thomson , J . J . : ·Phil. Mag . 46 0 513, (1923) . • 20 . 21 . 22. Wegstaff , J .E . P . : Phil . Mag ., 47 , 66 9 ( 1924) • Be l z , M. H.; Phil. Mag ., 44 , 479 8 ( 1922) . Whiddington , R . and Hare , A . : Phil. Mag •s 4:6, 607 , ( 1923) 23 . Dowling , J . J . : Phil Mag ., 46 ,. S l, ( 1923) • Nature , 111 , 742 ., (1923) . 24 . 25 . sucksmith, w . : Phil . Mag . 43 , 22~, ( 1922) . Scot t-Taggart, J . : Ele ctri cian, 86 , 124 ,(1921 ) . 26. 27 . 29 . 30 . 31. 32 . 33 . Me i ssner ., A . t Ele c tr1oian, 73 ; 7 02 , (1914) . Pi erce , G . w . : Am . AQad . Proc., 59 , 81, ( 19ij3 ) . Cady ., w. c.~ p . I . R . E ., 83 , 10, ( 1922) . Hartshorn, L . and Oliver , D .A . g P~oc . Roy . Soc • 0 123 1 664 , ( 1929 ) Smyth , C .p . and Morgan , S . o • ., : J .A .c . s • ., 50 , 1547 ~ ( 1928) • Kerr, R .N . : Jour . Chem . Soc ., 129 , 2796 , ( 1926) . Wi l l iams , J .w. & Ogg, E ..F . : J .A . c .s ., 50 , 94 ,. ( 1928) ., Krcnma , I . & Williams , J . w .D J.A . c .s ., 49 , 2408 , ( 1927) . 34 . Pohrt, G. sa . " " 49 , 113 , ( 1925) Ann . d . Physik, 42 8 569 , (1913) .