TIO

SC IENCBS ACADEMY

requires!' 'le distinguish de rio inhretises influences r{ui paraissent intei'venir et dont les actions, coi "1' r{ii('es des orientations, se combinent réciprofJtic men t: influen cc direct stu' the corner not }'ar a kind of 'rubbing or d'en traî meme ri t et in11 ucnc'- sur l'ai iriantiitioii }x opi'e bles ;ii quilles due au dép1acern ent in pidc diins le c)iam J , sans li:irlcr des ccii rati ts engendrés
pro bablcment dans l'i ctt vc r}u i j' ue en incrnc te mps tin rùle d'(Ecran and can partly mask these ell'ets.
Mais, cont iiic- on le conçoit faci le mien t, tlans bles conditions telles c}ue celles où nous opêrions (soit 'la ns un wagon .d'u ri train rapide ou sur une machine isolée) ct qui ue }icrine tien t de fii ire toutes les manœuvres désiratiles, par suite dii mia{; n/'tisiiie sous-}ieriiianent dcx elo}'pè }iar les vilii "utions et the router in1, .iir)if 'iue of the installation where we are at high speeds, it was not }iossible to detei'iiiinate the diveiv-es magne- data. ti'Jues and even less to dissociate ct d'É'tti il ier these different actions. Ces sans doute nc se iii1'len t }'as i 'csentcr uti int('r^' l' itir}ti'* iiii mt- diat, mais elles oc sont toute l'ois poin t sans port ce gcnerale, et nous mous propose rl'cn continue lii rechei'clic.

O P'I'IQUE. - Les syslênies optique.s en rnoiireritrnf et la ircinslalioii Ile la Ten e. Note by Mr. G. Saosac, presented by Mr. Lippmann.
1. E,frt fr rlémeiitaii-e movement. - J'ai expliqué einématiijuement : l'enti.i i ncuient [Partiel des ondes luini iieuses pai' l'eau en niouveine ri t Comptes ren lus, t. I 2ñ, p. 8 i 8; Soci'éld fi'ançaise de NÂJ sis-^' '°99) i = Yelttnann's principle and astronomical aberration studied with a system optique t{uelcont}uc Comptes rencliis, t. l4l , I go5, p. i cao). Mes raisonnements supposent que l'etlicr du vide n'est pas entraîné tout dans la traduction de la matière (hypothèse de Fresnel), ou, dii moins, que la vitesse r du s3-stème -rlique }i.ir rapport fi l'ether du vide est uniforme aii.x points of the system. But whatever the distribution of the vector v in the scope of the system, it is periiiis to keep in the following form the principle of the clcmentary motion effect, which I established in I°99 ( toc. cit. ) and which will serve as a breeze for a more cineniatic theory. genéi'ale.
Sur char}ue elê-nien t de loiig'ucur cll lie à u ri système optique, la trans- lation du systeme fait varier la durée de propag ation des ondu lations lumi- neuses de u dl (effet dc mouvement eleiræn taii e); u désig'ne la composante,

SESSION OF Ô FEBRUARY i g i i .

3 i i

according to 'll, of the velocity v of the element Cf of the system with respect to t h e vacuum cther; V, designates the velocity of light ':lnns the tyree, iT1''ine if the ele- ment ch is coin}aris clans l'un des rn il ieur matériels du systèiTle o}atirjiic.

2. E,fel. toni hltlonnaire maligne. -° *Pl°-lle ainsi la variation AT que la dur ec de propagation sur le périmr-tre du cirçtiit subit sous.s l ' influence du relative movement cle this invariable circuit and ether 'lii vide. This is the
sum of clem on taii'es Vé c tendue fi tons les elena ents 'll alu circuit. Or, la somme des aletit s Ile rr r/f r-1-r:sente (lorcl I(e1vin) la r:'rro/"iir'n C cle l'e ther le lon g du circti it on ( i3jerl'nes) 1'/n/erzsifr ':Iii lourl'i'llon corres}iondan t, through the circuit. 1ii ti'ofliiisons la valeur ur moyer ne /' d u vecteur
de Bjerl'nes, ou densité tlu tourbillon, P-' l'endiciilaireinent à la siii'face S dlI Cil'CUit StlppOSt }1l élfl . J-'C fPt tOtl1')JÎl1Ol1f1ilirC Optit{tle 8 }JOUl' Villetl1'

C bS

If the density of the tourlaillon is toiijotir.s ri i:lle, in other words, if the relative moirre- iiient tle l'etlier is irrolalu'itnrl, the valcNr of AT is zero and we }ieu t apply Vel tinan ri's theorem (/'c. "i/.).
If, on the other hand, the relative slack is rotated nell the delay dT {" ocliiit a variation 'J- 1° ''ise (L, long'iierir d'onrle) :

then interferer bleu x systems of light ondiilations '{ii i have traversed in opposite directions the circuit o} tir}ue Ile large surface S (see inès Notes, Comptes i eii':lus, t. 160, I t) I O 1-° ^°* °* ' 7 )
L'effet tourlaillonnaire alterera de zm 1'i difference Ile phase des deux ondulations inverses, car il rcsul te rl'cffets de mouvement elu }" einier ordre c}ui cliaiig'ent Ile sens avec la }="°1'"3'ation Ile la lumière.
3. Limi "te siipériew-e Ile l'r.ntt'cn "iienienl rl,e l'élliei- rlcins In trcmslnIi'oii 'ie In Tew'e. - Si l'étlier est su}'}iose entrainé aii oisina ge du sol, la vile.sse relative r de la Terre et clv 1'et lier aug'iiiente de fr r{uanrl l'altitude croit 'le As et ne devient ég'ale 'i lit vitesse r, de ti.iii.slation Ile 1:i Terl'e r}u'à l'altitude où cesse Ï 'entrainenien t.
Around noon (or midnight), the speed v is Jaarall!ele at the horizon, the
vector ô is horizontal, close to the inericlian, and has the value ps if we

neglect the

3 a

ACADÉM IE DES SCIENCES.

curvature of ether flux lines with respect to - Under these conditions,
the value (*) of m applies to the surface S of a vertical circuit oriented east-west.
From noon to midnight, the direction of propabation o f each ripple is reversed i n space, the 2s variation of the phase di fference is inverted and t h e interference bangs must move by d m rows.
In the course of observations that I will describe elsewhere, I found that the position of the central fringe of my 'i inverse beam interferometer (loc. cit., }i. i G26) did not depend on the time of day. my pointing accuracy made it possible to determine an upper limit of m corresponding to , "" of wavelength for a circuit of 3o'° contour, inclined to the horizon; vertical projection zo"". According to the formula ( *), ù or Ar then admits the
radian per second. In other words, for a vertical ascent of I'", the relative
velocity v increases by no more than a fraction 3 i o" of the Earth's velocity
v. Taking up the theory of star alignment Comptes rendus, ' 9º°.
loc. rit. ) in the liy}iotliùse of an entrainment of the etlier r' ù s t h e ground, one sees easily c{u'elle subsiste, à condition de définir l'aberration par la vitesse relati -e r of the globe and the étlier att place o f observation. As the value of the Earth's velocity v, makes the oliservated value of the alierration retron ver, to the a}'rroxiiriation of , is that the entrainment v e l o c i t y (r, - v) near the ground adrriet Jr, as the su}ierieure limit. The result of these observations completes the r-ecedent. Moreover, it shows c}u il faiit réduire 1ieaucou}i la upper limit Ile the drive speed (r - r,) if you d o not want to This speed is still noticeable at great altitudes.

6. Ef,fet toui-hillounnire optic]ue a,agiilaire. - Let two Jiinettes be directed towards each other at a great distance D. On t h e area (D . /) of the section (diameter f) of the long beam l uniineu.x which separates the glasses,

l'efI'et tourliillonnaire 1'' d nit le i et'ird ET ou D . / en entre les deux vilx a-

tions elénientaires p- i'agées suivant les l'ordS °1°1'Oscs du faisceau. To 'iue

the focal synclironisine is ret,abli, the image of the foyet' of a telescope in the

focus Ile l'autre 'loit é tre deviée tl'ti n anÿ'1c t tel '}iie l'avance géométrique

c1 compensates for the geoinetric}ue reta rd V, ET. It is easy to deduce that if the

sharp lii L_ is exactly pointed at the lunet te L, , the latter is depointé-. 1'-' --rP '

' I " of angle 2t or AD

V,

rEy 6 SESSION

9'

3 i3

According to the upper lirrlÎte of ù, c}iie inès observations ont ('tablie, the effect

tonrbillonnaire angulaire * e admet la limi te supérieure 3 i o ' " ' -
c
To determine this limit directly, it would have been necessary to set the precision of the recipror}ue tips of the two glasses at less than o", i through a mosplieric layer i oo' ' long, or liien at less than o", i 'a

15° Island distance.

TH ER MODYùtAMIQU E. - A pylicnl.i'on dif I*'-*'-cij'e Ile heu - cmæ ju/iènornenes yti accomyamet lci charge des coztdrtisnleurs. Note by il. A. EDLti, presented by II. E. Bouty.
I will first summarize, in a single section, the dernonst ations relating to .t these 1 lic-nomena given by Pellat ( ' ) and by Sacerdote (') using Massieu's remark.
Soi t u n co nrfenscifeii /' fer mê don t les a i in atit res in i n ces son t coll'ies su t le d ielec- triq ue. L'a rma t u re ex te rn e B c-ona m u ri iq ue où l'e ricei n te, ta n dii q u e l'i ri tern e est }iortce a u pot en t tel V > o.
The state of this condensate t1/ depends on V, its tent perat u re 'I" and 'the p ressi on s u ii iforms P and p qu i ni g n e, one at the extü rien r, the other at the i ri tc ri eu r.
Let's assume 1° co nstant. Sous 1'i ri fl uen ce des vari al ions d V, d'1', d/', la clia rie 31 de .\ a u, in e ri te de rf\f, le dié lectri c{ue reçoit une qua ri ti té de cli ale u r/Q, le s ol u nde in tc ri cu r ( ca vité )' aux ir e ri te de dv e le vol u ir e ex rérie ur v' de di''. The energy increase o f the condensate in this process is
If this transformation takes place in a way that is similar to il' le, rfU, d(SIY ), rt( /ai ) and c/ ( 1° i") are d i Her en tielles ex actes, and it is Ile iué in e de
Poso ns
( 3 )

(' ) I*. P'r.£ituoTc, ' lit.e cle rloclora I, y. fi. Pa i'is. i Sgg.