ON THE MODULUS OP TORSIONAL RIGIDITY OP METAL WIKK8. 



29 



TABLE V. Platinum Wire. 



Between the observations on December 5 and those on December 9 the wire 

 was heated several times to above 100 in an endeavour to obtain the temperature 

 126, but owing to the amyl alcohol in the boiler having absorbed some water, the 

 mixture boiled considerably below that temperature. I had therefore to suspend the 

 observations fora few d.-iys while I purified the amyl alcohol. During this interval 

 some more observations at 100 were taken, as will be seen from the table. The first 

 observation at each temperature is plotted in Diagram IV. 



