448 C. Bams and V. StrouhaC — Strain-effect of sudden 



Professor Rood* informed us that the polarization figures could 

 be quite wiped out by annealing such ordinal glass as ex- 

 hibited them only as far as the temperature of melting zinc. 

 We infer that at the end of the first phase of annealing we 

 have in hand a hollow glass globule practically free from 

 strain. 



During the second phase of the annealing phenomenon (500° 

 to 1000°), we observe a very pronounced change of the densit}' 

 of the Prince Rupert drops, corresponding to the above result 

 for steel. But the explanation is here readily at hand : at in- 

 cipient redness the enclosed bubbles disappear or are reduced 

 to mere specks. The large increment of density in question is 

 therefore nothing more than an expression for the collapse of 

 the viscous hollow globule in virtue of atmospheric pressure. 

 This important observation enables us to interpret all the phe- 

 nomena of annealing satisfactorily. It must therefore be care- 

 fully examined. 



We will endeavor to prove that the bubbles are vacua ; that 

 they are not accidental inclusions of gas or aqueous vapor. 

 The temperature from which glass is quenched is certainly less 

 than 1500°. The temperature at which glass is sufficiently 

 viscous to yield easily to atmospheric pressure is certainly 

 greater than 500°. Suppose now that the changes of volume 

 of the bubble were the result of thermal expansion of an in- 

 cluded gas. Let v im and v m be the volumes of the gaseous 

 inclusion at 1500° and 500°, respectively, under normal pres- 

 sure. Let v h and v s be the volumes of the (gas) bubble for the 

 quenched-hard and annealed-soft states. Then we deduce, a 

 fortiori, 



^ ^o = 2 .g (1) 



with this as a point of departure the following little digest, 

 Table IV, of mean results has been prepared. Here m is the 

 mean mass of all the drops examined ; d A , J s , V h , V s , their mean 

 density and total volume for the hard and the soft states, 

 respectively; v is the mean volume, r the equivalent mean 

 radius of the bubbles. In the second horizontal row V h has 

 been diminished one-half per cent to refer all volumes to the 

 beginning and end of the second phase of annealing. 



* Eesults obtained by Professor Rood during his experiments with high vacua. 



