Properties of the Garnet Group. 39 



dite + spessartite = 7 ; almandite + andradite = 4 ; grossu- 

 larite + pyrope = 3 ; almandite + pyrope = 1 ; almandite 

 + grossularite 1 ; andradite + spessartite = 1 ; grossnlarite 4- 

 spessartite = ; pyrope + spessartite = 0. Of the possible 

 different combinations with three constituents the following 

 were observed : grossularite 4- andradite + almandite •= 28 ; 

 grossularite + andradite 4- pyrope = 25 ; grossularite 4- an- 

 dradite + spessartite = 15 ; grossularite + almandite + pyrope 

 = 15 ; grossularite 4- almandite + spessartite = 13 ; pyrope 4- 

 almandite + andradite 10 ; pyrope 4- almandite ■+• spessartite 

 •-= 7 ; andradite 4- almandite + spessartite = 6 ; andradite 4- 

 pyrope + spessartite = 2. 



The question naturally arises in this connection, whether this 

 limited miscibility of the different garnet molecules depends 

 upon inherent characteristics of the garnet group and that in 

 a certain case only a definite amount of one molecule can com- 

 bine with another, or whether it is brought about by the nat- 

 ural restrictions of the ordinary mode of occurrence of that 

 garnet. In other words, this limited miscibility may not 

 depend upon chemical restrictions, but rather upon restrictions 

 imposed by the conditions of the origin of the garnets. This 

 is a very interesting question and might possibly find its 

 solution in either of two ways. It might be possible to make 

 experimental mixtures of the different garnet molecules in all 

 proportions and determine the limits, if any, of their misci- 

 bility. Another method would be to make an exhaustive study 

 of the modes of occurrence of the members of the group, 

 together with their associations, etc. From this latter method 

 it would probably be possible to discover whether or not the 

 manner of origin of the garnets had such a positive influence 

 upon their composition. 



In addition to the above study of the variation of the refrac- 

 tive index with change in the chemical composition, a similar 

 study of the relations between the specific gravity and compo- 

 sition was made. Of the analyses studied, the specific gravity 

 of the specimen was recorded in 72 cases. Three of these 

 analyses were rejected because of obvious errors. In five other 

 cases the error between the observed and calculated specific 

 gravities was greater than 0*1. These five analyses were also 

 rejected, leaving 64 analyses upon which the calculations were 

 based. After considerable experimenting it was found that 

 the following specific gravity values were most satisfactory : 

 pyrope = 3-510 ; grossularite = 3-530 ; andradite = 3*750 ; 

 spessartite = 4-180, and almandite = 4-250. By the use of 

 these values the theoretical specific gravity of each of the 64 

 garnets was calculated from its analysis. The average differ- 

 ence between the measured and calculated specific gravities of 



