22 



DR. F. HORTON ON THE EFFECTS OF CHANGES OF TEMPERATURE 



coefficient of the rigidity modulus of different specimens might be more constant 

 than the absolute value. This, however, was not found to l>e the case. 



Copper. 



The wire used was made from electrolytically purified copper. The observations, 

 the results of which are given below, were all taken with the apparatus in its final 

 form. I had previously experimented on the same wire in the first form of apparatus, 

 and had ascertained that at a constant temperature the period of torsional vibration, 

 and therefore the rigidity, was constant. The periods given below are each the mean 

 of many observations, they have been corrected for clock rate and for the expansion 

 of the wire and vibrator. 



TABLE I. Copper Wire. 



The numbers in the preceding table show that after the wire has been heated the 

 rigidity at the lower temperature does not come back to its original value, but is 

 slightly greater than it was before the heating. The mean of the first two 

 observations is plotted together with the observations at the higher temperatures in 

 Diagram I. 



It will be seen that the points fall on a straight line, showing that the rigidity 

 modulus is a linear function of the temperature. From the diagram the value of n at 

 15 C. was found to l>e 4 '3693 X 1.0 11 dynes per sq. centim., and the coefficient ft to 

 be '0003877. The modulus at t is therefore given by the equation 



, = 4-3GJ3 X 10" J 1 -0003877 (t 15)j. 



