Mr. Shepard^s reply to Prof. Del Rio. 317 



iiig experiment, will abundantly compensate the ultimate loss of 

 time experienced in acquiring the natural orders. 



To Prof. Del Rio's objection to my frequent division of the 

 species, so that members of one and the same species appear in all 

 the classes, I adduce some remarks contained in the Treatise, which if 

 they had met his eye- might have saved him the wholly inapplicable 

 suggestion he has made in relation to the subject. 



" It may require an explanation, v.hy a mineralogical method, 

 should, unlike the systems in zoology and botany, make provision 

 for any but perfect or crystallized minerals. In, the vegetable king- 

 dom, it is well known, that no object is considered as classifiable, 

 unless possessed of the parts of fructification ; or in other words, of 

 the highest degree of perfection in its characters, under which it is 

 capable of appearing. And altliough tl)e majority of plants, ordi- 

 narily under our observation, is imperfect in these respects, no seri- 

 ous inconvenience arises from the fact, since they are all possessed 

 of an active principle, whose operation will at length advance them 

 to maturity ; in addition to which, we have no difficulty in finding 

 other individuals of the same species, already in possession of the 

 requisite perfection to enable us to accomplish their determination. 

 But it is otherwise in the mineral kingdom. Semi-crystallized and 

 uncrystallized minerals constitute by far the largest part of those re- 

 quiring determination, and they are wholly destitute of any tenden- 

 cy towards a higher degree of perfection. As we find them, so 

 they remain, (unless, indeed, they become, as sometimes is the case, 

 more imperfect still, from external agencies ;) and unlike the deter- 

 mination of imperfect plants, by the aid of those v^hich are more 

 perfect, it is seldom possible to determine them from their associa- 

 tion with crystallized individuals of the same species. From this 

 we see, that a method which should omit to provide for such mine- 

 rals as are not fully perfect in their properties, would be extremely 

 imperfect in general practice. 



" As a consequence of tins nccesniij of providing means for the 

 determination of imperfect minerals has arisen the frequent divis- 

 ion of the species. Thus portions of the species Fluor are found 

 in all of the classes, according as the individuals are crystallized, 

 cleavahle, or massive. It is to he remarked, hojvever, that this di- 

 vision within the species (unknown in the other departments of 

 natural history,) never takes place in the crystallized individuals 

 of the mineral kingdom; among which only should we expect to 



