206 Pen field and Howe — Chemical Comjiosition, etc. 



Authors. Sjogren. 



Chondrodite M<r 3 [Mg(F . OH)],[SiOJ s MgTMg(F . OH)] 4 [SiOJ, 



Humite Mg 5 [Mg(F. OH)l [SiOl Mg 8 fMg(F. OHmSiOJ, 



Clinohumite Mg 7 [Mg(F . OH)],[SiOJ< Mg,[Mg(F . OH)] a [SiOj 3 



Sjogren was correct in assuming that hydroxy 1 is present in 

 the minerals and that it is isomorphous with fluorine, but he 

 failed to obtain the correct formulae because he made direct 

 comparisons of the silica percentages calculated from his 

 formulae, with the actual figures yielded by the older analyses 

 which are somewhat low, while he neglected to take into con- 

 sideration the fact that magnesia is partly replaced by ferrous 

 oxide, which from its higher molecular weight still further 

 lowers the percentage of silica. Thus it happens that, without 

 making any correction for ferrous iron, the actual determina- 

 tions of silica in the older analyses of humite and clinohumite 

 agree very well with the formulae proposed by him. To 

 account for chondrodite, the most basic member of the series, 

 Sjogren adopted a formula containing the Mg(F . OH) group 

 four times. If in this F : OH = 1:1 then the calculated 

 silica, 33 "33 per cent, agrees very well the average found in 

 the older analyses, which is about 33 "5, but this should be 

 nearly 35-0 when FeO is calculated to its equivalent of MgO 

 and the analyses are brought to 100. Also the calculated per- 

 centages of F(7'04) and H 2 0(3*33), taken together, are far in 

 excess of anything observed in chondrodite. In the authors' 

 formula the ratio of RO : (F + OH) = 5 : 2 while in Sjogren's it 

 is 5:2-5. 



That fluorine is isomorphous with hydroxyl, or plays the 

 same part in a molecule, seems too well established to need 

 special comment. The results of our analyses indicate clearly 

 that in the minerals of this group this isomorphism exists and 

 if we abandon the idea a satisfactory explanation of their 

 composition cannot be given. The theory that fluorine can 

 replace oxygen, which formerly was so frequently advanced, 

 is similar to our present idea that fluorine and hydroxyl are 

 isomorphous, and the latter would seem to be the more natural 

 for both fluorine and hydroxyl are univalent, the one having 

 an atomic weight of 19 and the other a molecular weight of 17. 



The formulas thus proposed as the result of this investiga- 

 tion are simple and rational, they agree in a very satisfactory 

 manner with the results of our analyses as well as with those 

 of others and they constitute an extremely interesting chem- 

 ical series, which is related in a remarkable manner to the 

 crystallization of these minerals. 



Laboratory of Mineralogy and Petrography, 



Sheffield Scientific School, December, 1893. 



