290 



DK L. N. G. FILON ON THE 



TABLE III. 



For the glass 935 a correction yW s was applied to W, y being taken +0'001 for 

 pressure and O'OOl for tension. In (22) we have then to substitute W-f yW a for W. 

 This glass is badly annealed and does not seem well fitted by the formula. 



The glass 2783 had to be reduced differently. This is a lead glass, a specimen of 

 which had been examined under simple pressure and whose behaviour had appeared 

 peculiar (see ' Camb. Phil. Soc. Proc.,' vol. XII., Part V., p. 323, where the glass in 

 question is described as O 154). 



There are two sets of tension observations, denoted by A, B in Table IV., and two 

 sets of pressure observations denoted by C, D. The values of A r , ?-AX/AW r (see 17) 

 are given in Table IV. 



