DR T. THOMSON on Asbestus, Chlorite, and Tak. 361 

 This is equivalent to 



21 1 atoms silica, 



7.2 atoms magnesia, 



3 atoms lime, 



0.7 atom protoxide of iron, 



0.05 atom protoxide of manganese, 



0.17 atom alumina. 



If we take the protoxides of iron and manganese along with 

 the magnesia, it is evident that the atoms of magnesia are al- 

 most three times as numerous as those of lime. So far there is 

 an agreement between the composition of the Bernera asbes- 

 tus and amphibole ; but there is a slight deficiency in the silica, 

 even if we add to it the alumina which the mineral contains. 

 For, if we divide the atoms just given by three, we have, 



1 atom lime, 



2.65 atoms magnesia, protoxide of iron and manganese, 



7.3 atoms silica and alumina. 



But, to judge of the deficiency accurately, we must take a lit- 

 tle of the lime, and add it to the magnesia, that the atoms of 

 magnesia may be thrice as many as those of lime. This will 

 give us 



1 atom lime, 



S atoms magnesia, lime, and protoxide of iron and manganese, 



8 atoms silica. 



To form tersilicate of lime and bisilicate of magnesia, we would 

 require 9 atoms of silica and alumina ; while the mineral con- 

 tains only 8 atoms. So that, to constitute 



