Site of the Ancient City of the Aurunci. 231 



ral admits of a considerable diversity of composition, whilst 

 still retaining its own peculiar crystalline form.* 



But later researches have carried us a step further ; they 

 shew that crystallographers had often confounded several 

 minerals which analysis proves to be distinct ; thus Henry 

 Rose of Berlin, and other chemists of the school of Berzelius, 

 proved, that whilst one kind of mica contains a certain defi- 

 nite amount of potass, in another a portion of that base will 

 be replaced by lithia, and in a third by magnesia. 



In the two first, the mineral, when examined by polarized 

 light, exhibits only one axis of double refraction, or one set 

 of rings, whilst in the last it displays two such sets, so that 

 the magnesian mica deserves to be distinguished as a sepa- 

 rate species on crystallographical grounds alone. 



In like manner, Henry Rose has asserted, on the faith of 

 chemical analysis, that mineralogists had designated by the 

 name of felspar several distinct substances, a statement which 

 the more exact examination of their respective angles of cry- 

 stallization has since shewn to be correct. 



There is, indeed, such an analogy between them with re- 

 spect to their chemical composition, as well as to their ex- 

 ternal characters, that they may, perhaps, be conveniently 

 considered as merely different species of a genus to which the 

 term felspar is applicable, the new terms which have been 

 affixed to them by way of distinction being regarded as de- 

 signating the species. 



Thus the genus felspar may be defined as indicating a mi- 

 neral, the primary form of which is an oblique rhombic prism, 

 and its chemical constitution that of two compounds of silica 

 united together ; the first consisting of silica united with any 

 base which contains only one atom of oxygen, the second com- 

 pound of silica, one where that body is united to a base in 

 which the proportion of oxygen to the radical is as three to 

 two. 



* Rammelsberg, a disciple of Rose, has proposed a chemical classification of 

 siliceous minerals founded on this principle, which I shall give in the Appen- 

 dix, as it seems to me the best attempt that has been made, to reduce to order 

 the confused mass of species containing various proportions of those earths 

 which mineralogy presents. 



