ON THE SPECIFIC HEAT OF WATER. 



243 



It is clear that the thenuoid effect must be considered in relation to current 

 density. We may therefore call ft in this last expression the thermoid coefficient for 

 win- of a given material and HI/I-. 



In the following table an- given the particulars of the three platinum wires above 

 mentioned, and the values of a and ft which result from our experiments. These 

 were of the same nature as mentioned in Section 6 of the paper, i.e., the wire was 



THERMOID Effect in Pure Platinum Wires. 



measured in a bridge arrangement against a mercury resistance, the current being 

 varied from zero to about five amperes for the largest wire and two for the smallest, 

 and the bath kept at a nearly uniform temperature. The thermoid effect, as 

 expressed in the coefficient ft, is seen to be roughly proportional to the radius of the 

 wire or more nearly proportional to r 4 ' 3 . It may also be expressed approximately as 



ft = O'OOO? V + O'OGSr 1 . 



These comparative results for ft are only approximate as the drawing of the wire 

 affects its structure so that the results in the different wires are not exactly 

 comparable. The ordinary temperature coefficient of each wire was observed and is 

 placed in the third column of the table. It appears that each successive drawing of 

 the wire affects its structure in such a way as to send up its temperature coefficient. 

 Probably the heat conductivity would also be affected by the successive drawings. 



2 I 2 



