178 Shepard's Treatise on Mineralogy. 



Vol. XXXII, p. 341, and its station as a columbate of lime, accord- 

 ing to one of Shepard's analyses, confirmed. Dana's Mineralogy, 

 one of the arrangements of which is crystallographical, although 

 in the last edition entering into every possible detail on these 

 two minerals, singularly enough omits even an allusion to the 

 above circumstance, notwithstanding its being so remarkable an 

 instance of the power of crystallography, to indicate error in 

 chemical analysis, even in hands like those of Wohler. 



" This mineral is an excellent exemplification of the difficul- 

 ties which at present surround the natural arrangement of mine- 

 rals, although chemical analysis is unquestionably hereafter des- 

 tined to be its basis. The analyses of the dark colored crystals 

 give as ingredients, columbic acid, lime, manganese, iron, tin, 

 lead, uranium, &c. ; whereas the minute transparent yellow crys- 

 tals are probably pure columbate of lime, or perhaps, even ob- 

 taining their color from a slight admixture of oxide of uranium, 

 as this color differs much in intensity in crystals of the same 

 size. These small transparent crystals are generally modified on 

 the edges and solid angles of the octahedron ; in the large dark 

 colored crystals, these modifications are often nearly obliterared. 



" My largest crystal of pyrochlore from the Chesterfield locality 

 is three eighths of an inch at the base of the octohedral pyramid." 



We have seen above in remarking upon the labors of Mr. 

 Hayes, that in the opinion of Berzelids nothing had been shown 

 which could support the reference here made of microlite to py- 

 rochlore, so far as the chemical evidence might be supposed to 

 weigh. Let us now see what fresh proof from natural properties 

 Mr. T. has adduced. It is all contained in the first sentence of 

 the paper, where he reasserts, that in point of color and crys- 

 talline structure the two minerals are the same. But where min- 

 erals afiect, as in the present case, forms of invariable dimen- 

 sions, nothing can be shown from structure ; which leaves him to 

 fall back upon his first ground, viz, similarity in color, as his sole 

 point dfappui in the discussion. Color however proves abso- 

 lutely nothing as to the identity claimed. 



The whole mineralogical question then, if there still be any, 

 comes back to the point where I had left it — viz. to the difference 

 in regard to gravities. Pyrochlore has a gravity = 4-20 . . . 4-32. 

 Microlite has a gravity=5-48 . . . 5-646.* As soon as Mr. T. 



* A fresh determination on a crystal weighing 463 grains, gives 5646. 



