414 Dr. Thomas Thomson on the 



Opaque. Lustre semi-metallic, inclining to resinous. 

 Fracture imperfect conchoidal. 



Hardness 6 or 6-25. Specific gravity by my trials 6-0380. 

 This is exactly the specific gravity given in Haidinger's 

 translation of Mohs's Mineralogy, shewing that his descrip- 

 tion is of the same mineral with that in my possession. 



The phenomena, before the blow-pipe with columbite, are 

 the same as with torrelite. The presence of iron and man- 

 ganese being indicated. 



I analyzed 20 grains of Bohemian columbite precisely in 

 the same way as I had already analyzed torrelite. But as 

 my quantity was limited, I was at more pains to prevent 

 any loss. The consequence was (as is frequently the case 

 under such circumstances), that I obtained a slight excess. 

 It is proper to mention also, that when the peroxide of iron 

 extracted from the mineral was ignited, and then digested 

 in muriatic acid, it left 0*2 grain of columbic acid. Thus, 

 shewing that columbic acid, before ignition, is not only 

 slightly soluble in ammonia, but also in muriatic acid. 

 The result of my analysis was as follows : 



Columbic acid, .... 15-93 79*65 



Protoxide of iron, . . . 2*80 14*00 



Protoxide of manganese, . 1*51 7*55 



Oxide of tin, 0*10 0*50 



Moisture, 0*01 0*05 



20*35 101*75 

 Dividing these numbers by their atomic weights, we 

 obtain. 



Atoms. 



Columbic acid, 3*09 or 1*987 



Protoxide of iron, . . . . 3*11 or 2* 

 Protoxide of manganese, . . 1*67 or 1*07 

 These numbers approach very nearly to 

 2 atoms columbic acid 

 2 atoms protoxide of iron 

 1 atom protoxide of manganese 

 which evidently shews the constitution of columbite to be 

 1 atom dicolumbate of iron 

 1 atom columbate of manganese 

 The formula indicating its composition is, f- CI -f mn C\. 



It differs from torrelite by containing half an atom of 

 columbic acid more. Torrelite consists of li atom columbic 

 acid united to 3 atoms of oxides of iron and manganese. 



