Vol. 63.~] CHANGES OF PHYSICAL CONSTANTS IN MINERALS, ETC. 157 



of the meldometer when using purified natural felspars. This dotted 

 curve represents the fusion-point of the glasses formed by melting 

 crystals of the given specific gravities : therefore, it is necessarily 

 lower than one obtained for the crystalline substances themselves. 

 I can in no way account for the wide difference between my results 

 and those of Prof. Joly & Mr. Cusack ; the purification of the 

 tminerals could hardly have so great an effect. 



In using the meldometer, I found it absolutely necessary to 

 calibrate each ribbon separately. When this was done, I found 

 that no two curves corresponded exactly; and readings from one 

 ribbon, if inserted on the curve obtained for another, gave results 

 differing widely from those obtained on its true curve. 



The melting-points that I have given are those at which the 

 particle appears completely molten, and change of form begins to 

 take place. 



The following tabulated list shows the similarity between 

 the change of specific gravities of the natural and of the artificial 

 felspars, and a comparison of melting-points : — 



Sp. Gr. 



Sp. Gr. 



Melting-point 



Melting-point 



of crystal. 



of glass. 



of crystal. 



of glass. 



*2-765 



2-700 



1532° C. 





2-750 



2-665 





1505° C. 



*2733 



2-648 



1500° O. 





*2-710 



2-590 



1463° C. 





2-700 



2-550 





1390° C. 



*2-679 



2-533 



14i9° C. 





2-677 



2-514 





1340° C. 



*2-660 



2-487 



1367° C. 



1 



2656 



2-470 





1310° C. 



*2-649 



2-458 



1340° C. 





2-625 



2-373 





1268° C. 



*2-605 



2-382 









Those marked with an asterisk are the chemically-synthesized felspars of 

 Messrs. Day & Allen. 



The greatest difficulty met with in using this type of instru- 

 ment is that due to the viscosity of the substances examined, since 

 there appears to the eye to be no abrupt change between the 

 crystal immediately before f ision and the fused substance when in a 

 viscous state. This is especially the case in minerals which undergo 

 a long period of viscosity such as albite ; and in such minerals 

 deorientation may be withheld for some time, owing to their high 

 viscosity. 1 



1 For a criticism of the two methods of determining melting-points men- 

 tioned in this paper, see C. Dcelter (5, iii) and A. L. Day & E. S. Shepherd (3 ). 



