like crystalline forms indicate like Chemical Formulae. 409 



some observations by my friend Mr. Richard Phillips'*, as if 

 it in some way opposed the general doctrine of isomorphism. 

 The second member of the formulae for these two minerals 

 is identical, and may therefore be neglected ; the first mem- 

 bers are Be Si"^ and Ca^ Si respectively. If from each we 

 take an atom of silica, then Be S should be capable of re- 

 placing Ca^ ; and if we suppose that glucina is not a sesqui, 

 but a protoxide, " the supposition," according to Mr. Phil- 

 lips, " most favourable to the doctrine of isomorphism ;" and 



that the Be of the formula should be written Be-^, replacing 



Ca% we should still have an equivalent of silica Si in excess, 

 in the formula for euclase. It does not appear therefore that 

 the formulae in the eighth group are at pi'esent to be re- 

 conciled. Were this the only case of such disagreement in 

 formulae, we might suspect that Mr. Brooke had mistaken 

 one or other of the minerals, euclase or zoisite; but the num- 

 ber of examples which the table contains tends to remove 

 such suspicion from the mind of those who are not so well 

 acquainted with Mr. Brooke's accuracy and knowledge of 

 minerals as I have had the opportunity of becoming. 



9. If the forms of silica and chabasie be identical, their re- 

 ceived formulae appear irreconcileable. 



10. Mohsite has not been analysed, but it is not easy to 

 suppose its known constituents to be present in quantities, 

 such as to produce a formula analogous to that of eudyalite. 



11. The only relation observable among the formulae in 

 the 11th group is, that the second and third added together 

 make up the first, or that the baryta harmotome may be sup- 

 posed to consist of an equivalent of stilbite united to one of 

 lime harmotome. In what way the formulae for these two 

 minerals are to be reconciled does not appear. 



From this discussion then of the mutual relations of the 

 formulae by which the substances contained in the above 

 groups are severally represented, we have arrived at three 

 conclusions : 



1. That some of them appear to be altogether irreconci- 

 leable. 



2. That those which may be reconciled require the adop- 

 tion of hypotheses which are much at variance with received 

 opinions; and, 



3. That these hypotheses may be directly the converse of 

 each other. 



• Lond. and Edin. Phil. Mag., vol. xii. p. 407. 



