Mr. Brooke ow Isomorphism. 1 65 



respect of stilbite and paranthine are not so in relation to 

 amphibole; and if they are so, then amphibole should consist 

 of 1 atom of tri-silicate of X, 



or tri-ahiminate of X, 



and 1 atom of hi-silicate of X, 

 or bi-aluminate of X. 



X being any or all the eight elements given above, and to 

 which others might be added, in any imaginable proportions. 



M. Beudant, who admits the doctrine of isomorphous sub- 

 stitution, has in his Mineralogy limited the general formula 

 expressing the theoretical composition of amphibole within 

 still narrower limits than those of Berzelius ; for he restrains 

 the composition of actinolite to single atoms of tri-silicate of 

 lime and bi-silicate of iroti, and that of hornblende to tri-alu- 

 minatc of lime and bi-aluminate of iron. 



There is not, however, a single analysis of any variety of 

 amphibole that I have yet seen, and there are nearly twenty 

 published by Leonhard in his Handbuch, and by Bondsdorf, 

 as given in the Annals of Philosophy for October 1822, which 

 affords the slightest ground for M. Beudant's imaginary 

 composition of hornblende. Nor is there a single analysis 

 that corresponds with exactness to any of the other formulas 

 of Berzelius or Beudant. The reader may therefore satisfy 

 himself, by comparing the theoretic formulas with the pub- 

 lished analyses, that the theory of isomorphous substitution 

 is not in respect of amphibole supported by the observed facts; 

 the evidence of which ought, on the contrary, in reference to 

 so fundamental and important a change in the doctrines of 

 chemistry, to have been both distinct and conclusive. 



It must also be recollected that the chemical formulae which 

 the doctrine of isomorphism has been brought forward to 

 reconcile and support, are in very many instances, as their 

 author Berzelius has candidly acknowledged, little more than 

 theoretical assumptions foimded upon the actual results of 

 analysis ; the essential together with the accidental or foreign 

 matter, being all taken, and parceled out in such proportions 

 as are required by the atomic theory for producing definite 

 compounds. But in doing this there is generally some in- 

 tractable surplus of one or more of the elements, which is left 

 out and is therefore regarded even by the new theory as in- 

 trusive matter. The results of analysis may also occasionally, 

 as Berzelius has reniarked, be etjually well represented by 

 two or more diflereiit formuhe, according to tlie proportions 

 which may be taken of the actual ingredients, and the manner 



in 



