Spodwnene and its Alterations, 



325 



It has been usual to consider the relation Na : Li, which is 

 here 1 : 9, and, according to other analysts, 1 : 12, varying from 

 1 : 4 to 1 : 20 ; bat it seems proper to group together all the re- 

 placing elements in R 1 , i. e. 



Mm Mg. Ca. (=2 R 1 ) + Na. Ka: Li : : 1 : 3. 



The corresponding ratio in Pisani's analysis of Spodumene 

 from Brazil is 1 : 6, and it is therefore evident that no definite 

 relationship holds in the replacements within R 1 



Spoclumene is thus shown to consist of one molecule of the 

 normal silicate of aluminium, plus one molecule of the normal 

 silicate of lithium, and its true formulas must be 



Empirical, Li 2 AF Si 4 O 12 



■p , • ! ( (Li. Na) 2 Si O 3 

 Rational j ^j ^ Q8 



Generally accepted.* 

 R 6 Al 8 Si 15 O 45 



3 (Li. Na) 2 Si O 3 



4 At Si 3 O 9 



This may be graphically represented, but more simply than as 

 given by Dr. K. Haushofer,f adopting the Augite-type, thus : 











/O Al=X=Si Ox • 



n q-/ ° \q- °- Li 



\ O / 0— Li 



\ Al= u =Si / 







From my oavh analyses, the theoretical constitution of the 

 American variety of this mineral may be calculated as follows: 



* Rammelsberg (Handb. der Min. Chem., 423, 1875), Roth (Allgem. u. Chem. Geol., I, 382, 

 1879), etc. 

 t Die Const, der natiirl. Silicate, 97 (1874). 



