of Glass at different Temperatures. 229 



The resistances in columns C 3 and C 4 are too high, because 

 of excessive polarization ; but taking H 1? H 2 , H 3 , and H 4 , we 

 find that the effect of the first three heatings has been to raise 

 the resistance to six times its original amount. Again, by 

 reversing the battery at the end of the experiment C 4 , I found 

 the resistance at 60° about 35 x 10 11 ; that is to say, the resist- 

 ance had been increased eighteenfold by the four heatings. 



In the electrometer-jar experiment the results were similar. 

 The resistance in this case, however, can very nearly be ex- 

 pressed by the equations: — 



log R= 17-535 -0-322 1 

 when the temperature is rising; 



logK = 18-1248-0-0353* 



when the temperature is falling. 



The heating in this experiment was much more slow than in 

 the test-tube experiment : three hours was the time occupied 

 in raising the temperature from 100° to 200° C. It is possible 

 that the greater regularity of the results in this case is due to 

 the slower heating, which would admit of the permanent 

 change being more nearly accomplished for each temperature 

 before it was passed. The resistance at 113° C. at the end of 

 the experiment was twice that at the beginning; and by con- 

 tinuing the curve given by the equation down to 0°, a resist- 

 ance of four times the original amount was indicated, part of 

 which must have been due to polarization. For this glass, 

 then, the so-called permanent change of resistance produced 

 by heating and cooling is comparatively small ; but probably 

 a greater effect would have been produced if the jar had been 

 raised to a higher temperature. 



When the glass is being heated, the polarization-effect pro- 

 duced by keeping the current constantly flowing through the 

 glass is very slight at low temperatures : only a very small 

 apparent increase of resistance can be observed after several 

 minutes' electrification. As the temperature is raised, how- 

 ever, the polarization gradually increases; and this increase 

 remains to a considerable extent during the cooling. The 

 increase of polarization is not due to the length of time the 

 battery has been applied before the temperature is raised; and 

 it assumes its former value within two or three minutes if the 

 battery is reversed at the high temperature. 



The following table shows the result of a set of four experi- 

 ments on a test-tube. The first two columns give the tempe- 

 rature and resistance when the glass was being heated, the 

 third and fourth columns the same when the glass was cooling. 

 The measurements were not, however, made at so low tempe- 



