annealed with c^.r^.j, Jh^^^jl^tlj^ m^'%lf^Wi\¥m. 

 directly ascertained :^jipsayiip|eni accuracy, as they were not 

 geometrically rectilinear, nave been calculated by the formula 



2jirnilafl} naawJod 89V|[93IP^^g^9:Mirfj2irn zazno owT .10 

 ,jt^,.o,r otinflni nsol bne o-tos o)bii4q8ft'io3 rf3idw tnoiJoubnilo 

 in whicbrfj oJ 9Jin9nr biow odi Bnivlqqu ,9iiv/ IfinoiJibbc arft lo 

 jncJgiio L is'thS Idnglb soughtvln'inillimetres, dj elaqmoa 

 odj bnil P the weight of the wire,: in grammes, .HI .aiiw 

 -•^^\ns lo e^ its density, i:}«:no3 Jnsnuo baoubni arfj lo aoulBV 

 -y^Ino iiofft l:lits radius. tR .'f^jf?«:»fT> of Iff// ^rro ^earjlBv \«^Usw 

 ^ (The weighings were mad^f'byiBordn's method^ bAd)widi«ft 

 excellent balance constructed bj Fortin. The radii wfer6'esti>* 

 mated by the method of enroulemeiiL The densities are taken 

 from the best tables known. i>j}y9Tb ai o'liw lojoubni sib 



ab'^Q aohnfi 8d:r iadi nv^od a 



Name o£ the wires. 



Diameter,., f h'3fe5s¥ftrf bfA?°|i^ijfi 1 



tt; 



-rr 



grammes. 



39700 

 31-404 

 66-892 

 65-705 

 200-255 

 62-050 

 21-59.3 

 28-116 

 3-599 



metres. • ^^ 



0-8641 Ilia 

 0-927720 

 1 -007 160 

 r046243'T' 

 3-184726 

 0-828810 

 11-576000J; 

 4-216216 

 9-405405 



Platinum wire. No. 2 . 

 , .:. ... No. 3 . 

 Copper wire. No. 2 ... 

 , ..., ... No. 3 ... 



-do 90 aCJiJ! ^Q_ 4 



nf fiil o1.?J : No. 5 -.;.' 

 rfiPrass wire, No. 3 .»Bv-tv/ 



,.1 No, 4..... 



Steel wire, No. 1 



22-0 

 22-0 

 8-9 

 8-0 

 8-9 

 8-9 

 =8-4 

 8-4 



7-8 



metre. 



0-00200 

 000140 

 000300 

 0-00300 

 000300 

 0-00300 

 0-00050 

 0-00100 

 000025 



61'. I constructed a rheometer with the copper-wire No. 5. 

 This wire makes only four revolutions, and is not covered 

 with silk. The two needles (selected from fifty) are very light, 

 magnetized to saturation, and make an entire oscillation in 

 10"*1-. The point of the upper needle carries a small strip of 

 bhickened paper, which allows the deviations to be read upon 

 a frame of 0™-9 diameter. 



65. In order to have sources of electricity always compara- 

 ble, and of a perfect constancy, I employed two thermo-elec- 

 tric pairs formed of metals which undergo no alteration in 

 the conditions in which they were to be placed. These met^als 

 are platinum and copper, pure and perfectly annealed. I 

 avoided all foreign metal, by rejecting the use of solder. The 

 wires of platinum No. 2 and of copper No. 3, those of plati- 

 num No. 3 and of copper No. 2, were stretched in a vice toi<« 

 length of 0™'010 ; then, after having connected these surfaces 

 of contact by a copper wire finer and better annealed, they 

 were plunged into equal quantities of pure mercury, intended 

 to equalize promptly the temperature throughout their whole 

 extent. This mercury was inclosed in two German glass test- 



