1877.] 



Dr. E.J. Mills on Electroslriction. 



505 



Daniell's cell, having a porous diaphragm, and the circuit included a gal- 

 vanoscope. Observations were made at definite intervals of time, imme- 

 diately after stirring the liquid ; and the difference between the two scales, 

 after suitable reduction, was registered as electrostrictive effect. The 

 temperature was in all cases the unrestricted temperature of the labo- 

 ratory. 



Copper. — The thermometer (" 454 ") first used had a cylindrical bulb, 

 which was slightly concave towards its vertical axis. The electrolyte 

 consisted of the ordinary acidified solution of cupric sulphate, containing 

 zincic sulphate. 



In the following Table the column below x contains the time in hours ; 

 that below y the number of scale-units (" degrees " C.) remaining to be 

 traversed. The latter are compared with a series of values calculated 

 from the equation — 



y =9-859 (-96363)* + 1-358 (-71841)*. 

 Table I. 



X. 



y- 



y calc. 



Temp. 



1 



10-520 



10-476 



12-2 



2 



9-917 



9-856 



12-9 



3 



9-316 



9-325 



13-9 



4 



8-822 



8-863 



14-7 



5 



8-467 



8-452 



15-5 



6-5 



7-825 



7-907 



16-1 



8 



7-359 



7-427 



16-7 



9 



7-276 



7-133 



17-4 



10 



6-979 



6-856 



18-1 



48 



2-302 



1-665 



12-2 



Mean temperature 15'0 



The difference between experiment and calculation in the last value 

 of y is doubtless due to the considerable amount of stratification observed 

 in the electrolyte, which had not been stirred subsequently to the last 

 observation. Omitting this result, the probable error of a single com- 

 parison of the calculated with the found values of y is -056. The calculated 

 total ascent is 11*22 scale-units. By continuing the preceding experi- 

 ment, but with two Daniell's cells, for 23|- hours longer, 10*933 units 

 were actually attained. 



A thermometer (" 502 ") having a very nearly spherical bulb was next 

 employed; its polar axis (a line representing the prolongation of the 

 mercurial thread) measured 11*73 millimetres, its equatorial axis 11*55 

 millimetres. This instrument was prepared in the same manner as the 

 preceding, immersed vertically in the electrolyte, and turned halfway 

 round every half-hour. Even deposition on a sphere is, under such cir- 

 cumstances, out of the question, most of the metal being precipitated on 



