﻿Cooling exhibited by Glass and by Steel. 



191 



the decrements corresponding to the first and the second phases 

 of annealing, respectively. 



Tables VII and Till show that both in the case of glass and 

 of steel the mean strain effect of sudden cooling is dilatation 

 throughout the substance of the quenched material ;* that the 



Table VII.— Density relations of tempered glass relative to soft glass 

 (mean results for shells). 



A 3 = 2-491 ; - A — = 





No. of 









k-? 



P. R. 



shells 



h 



& 



A'. 



Drop No. 



dissolved 









A* 



3 



3 



2-473 + 0-006 







0-0076 



5 



3 



2-479 + 0-008 









0-0052 



7 



7 



2 477+0-003 



. _ _ _ 



2-495 



0-0060 



8 



7 



2-474 + 0-004 







2-498 



0-0072 



9 



7 



2-470 + 0-003 







2-501 



0-0088 



10 



2 + 2 



2-477 + 0-008 



2-489 + 0002 







0-0064 



lif 



2 + 2 



2 485 + 0-029 



2-492 + 0-026 







0-0028 



1 



2 





2-497 + 0-006 









2 



2 





2-484 + 0-000 











6 



2 



.... 



2-496 + 0-008 











Mean 



(H) 



2-476 



2-492 



2-498 



0-0063 



Table VIII. — Density variations of tempered steel relative to soft steel. 



No. 



2p 



A/-A s 



A/ -'A' 

 A c 



A'-A s 



A c 



A c 



I 

 II 



61 to 63 





 21 to 29 



1-90 

 0-58 

 0-13 

 023 



0-08 



0-0048 

 0-0155 

 0-0171 

 0-0133 

 00150 



0-0018 

 0-0060 

 0-0068 

 0-0080 

 0-0033 



0-0029 

 0-0095 

 0-0103 

 0-0053 

 0-0116 



Mean 





0-0131 



0-0052 



0-0079 



amounts of mean dilatation for glass and for steel are of like 

 order; that the strain in glass exceeds that of steel. We find, 

 in general, that the practically measurable value of density is 

 not a satisfactorily sharp datum for discerning the primary 

 causes of the electrical, the optic and the magnetic variations 



* If " quenching " means sudden cooling, then if the solid quenched be massive 

 and thick, the inner layers cannot be quenched in virtue of their position. In the 

 above Tables a general increase of the shell density from the surface inward is 

 indistinctly apparent. But the evidence is not sufficient. 



f In view of the large value of probable mean error of this measurement we 

 thought seriously of rejecting it. But we were unable to find any error in the 

 work and hence the datum has been retained. Its effect is principally apparent 

 in the ratio, depressing it in a way adverse to our inferences. 



