Properties of the Garnet Group. 



37 



exceed this amount except in four cases, namely in analyses 2, 

 8, 9 and 10. In these four cases the discrepancy is large, but 

 considering the close agreement shown in all the other cases it 

 would seem as if there must be some explanation to account 

 for the variation in these particular cases and that they ought 

 not to carry much weight against the accuracy of the assump- 

 tions upon which the table was made. The average discrep- 

 ancy between the calculated and observed values for the entire 

 series is only '007, and if the four cases mentioned above are 

 omitted this is reduced to - 0027. From this table, therefore, it 

 becomes evident that the indices of refraction of the members 

 of the garnet group vary directly with the variation in the 

 composition and that the theoretical values assumed for the 

 indices of the different simple garnets are very closely correct. 

 Such close agreements in such a large series of cases could 

 hardly be accounted for on any other assumption. 



Analysis Pyrope 



Pyrope 100.00 



Grossularite 



1 _. 1-29 



2 55-30 



3 61-66 



4 2-12 



5 54-78 



6 57-53 



7 



8 2-12 



9 9-33 



10 



11 3-57 



Spessartite 



12 



13 17-97 



14 



15 12-40 



16 26-90 



17 0-67 



18 7-88 



19 



Almandite __ 



20 3-75 



21... 



Uvarovite 



22.. 2-57 



23 0-69 



Andradite 



Grossu- 

 larite 



ioo : 6o 



94-66 

 5-96 

 531 



91-93 

 8-61 



79"-19 



V-68 

 1-93 



5-39 

 1-25 

 4-03 

 2-47 

 1-48 

 5-21 

 0-22 



rY-33 

 23-11 



i"i5 



4-06 



Spessar- Alman- 

 tite elite 



2 49 

 0-75 

 1-15 

 0-69 

 0-88 



0-30 

 3-01 

 3-16 



65-42 



91-29 

 100-00 



93 66 

 3-61 



85-77 

 2-71 



72-59 



50 : 44 



0-65 

 0-79 



0-99 



28-83 

 24-04 



30-99 

 36-77 



85-09 

 72-96 

 31-37 



3-63 



73-03 



12-98 



80-76 



64-39 



25 26 



86-91 



49-34 



100-00 



5-76 



1-78 



Uvaro- 

 vite 



9-16 



7-48 



4 : 74 



59 

 19 



73-16 



ioo : 66 



4-18 



Andra- 

 dite 



0-57 



5-26 



5 : 70 

 20-51 

 9-78 

 6-87 

 111 

 5-14 



2 : 71 



6-42 



75-11 



94-04 



89-09 



100-00 



re, meas. re, cale. 



Error 



1-736 

 1-741 

 1-742 

 1-744 

 1-750 

 1-760 

 1-763 

 1-778 

 1-782 

 1-792 

 1-800 



1 : 801 

 1-801 

 1-806 

 1-807 

 1-808 

 1-811 

 1-813 

 1-820 



l"-838 

 1-865 



l"-888 

 1-888 



■705 

 •735 

 •738 

 •759 

 ■742 

 ■743 

 •755 

 •762 

 •768 

 •833 

 •815 

 •806 

 •801 

 •800 

 •804 

 •801 

 •803 

 •807 

 •802 

 •806 

 •815 

 •815 

 •830 

 •838 

 ■857 

 •870 

 •887 

 •885 

 •895 



•002 

 •018 

 ■000 

 •001 

 •005 

 •002 

 •005 

 •055 

 •033 

 •014 

 •001 



: 003 

 •000 

 •003 

 ■000 

 ■006 

 •005 

 ■002 

 •005 



: 666 



•008 



-•001 

 -•003 



If then the garnets were composed of but two isomorphous 

 molecules it would be possible, after determining through 

 qualitative tests which two molecules were present, to establish 

 the composition of the specimen by means of its refractive 

 index. Unfortunately, however, the majority of garnets con- 



