116 



Profs. W. E. Ayrton and J. Perry on 



box, the three pieces of glass were removed and thicker pieces, 

 0'324 ceutim. thick, put in their place. The cover was again 

 cemented to the box, this being, in fact, always carefully done 

 after each occasion on which it was necessary to open the 

 box. 



III. Temperature being 2°'4 C, an electromotive force of 

 0'0087 A^olt gave (after waiting eleven minutes to allow the 

 current to overcome any previous polarization) a deflection of 

 212, corresponding Avith a resistance of 355,000 ohms per 

 cubic centimetre. Allowing the temperature to fall slowly, 

 we found the readings to remain pretty steady for thirty-four 

 minutes Avhen the temperature had become — 0°'6 C. From 

 this time there Avas a moderately rapid increase of resistance 

 up to 14 megohms per cubic centimetre after forty-three mi- 

 nutes, the temperature now being — 3°'8 C. The deflection 

 Avas now 54 ; and Ave increased the battery ten times, when it 

 Avas found that this increase of the battery did not increase 

 the deflection proportionately (although subsequent expe- 

 riments shoAved that the deflection Avas A^ery approximately 

 proportional to the current) ; in fact, making the battery 

 ten times as large (that is, now using an electromotive 

 force of 0*087 A^olt), only gaA^e a deflection of 129. The 

 battery Avas noAV immediately reduced to an electromotiA^e 

 force of 0*0087 as before ; but the deflection was now only 28. 

 Leaving on this electromotive force, the deflection was ob- 

 S3rved to steadily diminish until at last it became negatiA^e ; 

 and it eA^entually required an electromotiA^e force of 0*0435 

 volt to bring the spot of light to zero. This electromotiA^e 

 force in the ice, Avhich appeared to be developed by the em- 

 ployment of an electromotiA^e force of 0*087 A^olt for only 

 about one minute, remained quite constant, and continued to 

 balance an electromotiA'C force of 0*0435 A^olt for about half an 

 hour, AA^hen, the apparatus having been cooled doAvn, a larger 

 electromotiA^e force Avas now employed. 



An electromotive 



force of 



0*087 volt 



0*174 „ 



0*261 „ 



gave a deflection of 

 9-5 



19*5 



27*7 

 diminishing to 23*5 

 in thirteen minutes 



at a temperature of 

 -5°*1 C. 



-6°*4 C. 



These deflections are nearly proportional to the electromotive 

 forces. 



