1877.] 



Dr. E. J. Mills on Electrostriction. 



509 



instrument for such a purpose), at pressures varying from 49*4 to 133*6 

 atmospheres. By these it was found that a scale-unit corresponds, in 

 this case, to 9*44 atmospheres, with a probable error of 0*1 atmosphere 

 for a single determination. 



The above data have led to the construction of the following table. 



Table VIII. 



Extreme effect 



Metal. 



Therm. 



Effect. 



in scale-units. 



in atmospheres 





454 



Distension 



0-274 



2-3 









0-749 



6-2 



Nickel 





Compression 



2-314 



19-2 







55 



2-193 



18-2 







55 



8-006 



66-4 







55 



10-933 



90-7 







55 



12-87 



106-8 





502 



55 



14-707* 



74-0 f 





455 



55 



11-50 



108-5 



It thus appears that the extreme force at work on the bulbs must have 

 been very considerable. 



Subsidiary Experiments and Remarks. 



(1) An instrument having a general resemblance to a thermometer was 

 constructed with a cylindrical caoutchouc bulb, about 30 millimetres long 

 by 11 millimetres wide. The bulb was rendered conductive, and then 

 made a kathode in the copper bath already referred to ; it was frequently 

 turned half round. In the course of a day it had become coated with a 

 sufficiently thick layer of metal, and had then undergone most palpable 

 collapse. We have here, therefore, a real case of local constrictive effect. 



(2) Two flat microscopic slides of the usual size were silvered chemically 

 and then made the kathode in a similar circuit. After twenty-four 

 hours, the deposited copper on both slides was found to be arched down- 

 wards and inwards, towards the anode ; the deposits were in fact, owing 

 to their contiguity, to a great extent united. The upper parts of the 

 deposits had torn themselves away from the glass ; and two close parallel 

 lines cut through the silver of one of the plates had the copper between 

 them curled inwards like a gutter. Erom this evidence it was inferred 

 that the cylindrical deposit, if cut in the direction of its length, would 

 spring open. Such was found to be actually the case even with the 

 copper deposited on the caoutchouc bulb ; and to a still greater extent 

 with copper overlying some of the same caoutchouc, where that material 

 was stretched over a glass shoulder for the purpose of fastening, and 

 there did not admit of collapse. The opening backwards of nickel and 

 iron films has already been noticed. 



* Calculated. t This number may be wrong in either direction by 1| atmosphere. 



