Penfield and Foote — Composition of Tourmaline. 115 



change the numbers of the hydrogen ratio ± 0*27 in the one 

 case and ± 0*38 in the other. If the silica were one-half of 

 one per cent high and the water correspondingly low, the effect 

 upon the ratio would be to change it from 4 : 1 : 20 to 4 : 0*99 : 

 19*35. The evidence is therefore convincing that, with the 

 exception of analysis No. 53 (brown tourmaline from Grouver- 

 neur, N. Y.), the analyses of Riggs are very exact, and also 

 that the material he analyzed was very pure.* Referring to 

 page 111, it will be observed that analysis No, 53 is the one 

 which compares most unfavorably with Rammelsberg's cor- 

 rected analyses. If the silica in this analysis were 1*25 per 

 cent lower and the A1 2 3 correspondingly higher, the agreement 

 with Rammelsberg's corrected analysis would be quite satis- 

 factory, and the ratio would correspond with the theory. Leav- 

 ing out of consideration this one analysis, which may be con- 

 sidered as either defective or made upon impure material, the 

 average of the ratios of Riggs' analyses becomes 4 : 0'95 : 19*91. 



The analyses of Riggs were very severely criticized by 

 Rammelsberg, who characterized tourmaline analysis as "Kein 

 Thema fur Anf anger" but in the light of our present inves- 

 tigation we find the results very accurate, and it may justly be 

 said that we are indebted to Riggs for the best series of tourma- 

 line analyses that has ever been made. In fact, our conclusions 

 regarding the composition of the mineral might readily have 

 been deduced from his results alone. 



Review of the analyses of Jannasch and Kalb. — 

 Nine analyses were made, from which the following ratios have 

 been calculated: 













Total 











Total 



No. 



Si0 2 





B 2 3 





hydrogen 



No. 



Si0 2 : 



B 2 3 





hydrogen. 



56. 



4 



• 



0*96 



• 

 • 



19*7 



61. 



4 



1 0*95 



• 



20*2 



51. 



4 



• 



0*99 



• 



19*8 



62. 



4 



: o*80 



• 



20*00 



58. 



4 



• 



0*95 



j 



20*4 



63. 



4 



: 0*98 



• 

 • 



19-7 



59. 



4 



• 



0*92 



; 



18*8 



64. 



4 



; 0*84 



• 

 • 



20*01 



60. 



4 



• 



0*88 



j 



20*4 



Average, 



4 



: 0*92 



• 

 • 



19*9 



These analyses, like those of Riggs, bear every evidence of 

 having been made with great precision, and the ratios, with the 

 single exception of No. 59, approximate closely to 4 : 1 : 20, 

 thus furnishing additional evidence that the acid from which 

 all tourmalines are derived is II 90 B 9 Si 4 O 91 . The analyses do not 



* In a personal communication from Professor Riggs the following statement 

 is made concerning the quality of the material investigated by him: "The 

 material analyzed was of excellent quality, selected with great care. The color- 

 less, pink and light green varieties were transparent, gem-like crystals, and the 

 material of the rose-colored, brown and black varieties was, in my opinion, 

 equally pure." Dated, Hartford, January 4th, 1899. 



