Prof. Odling on Chemical Notation. 119 



The formulae of the extreme perissad compounds M CI and 

 M CI 7 being indisputable, that of the intermediate perissad com- 

 pound M CI 3 would seem to be at any rate not improbable. 



I will venture to add yet one more argument on this side of 

 the question, based, I must confess, rather upon a personal im- 

 pression than a demonstrable fact. I am inclined to think that 

 alumina is an analogue of boric anhydride; that boracite and 

 spinelle belong to the same type ; and that in several complex 

 aluminous silicates a portion of the alumina is replaced isomor- 

 phously by boric anhydride, or rather that a portion of the alu- 

 minium is replaced by boron. If this view is correct, the mole- 

 cule of aluminic chloride ought probably to be represented by 

 the formula Al CI 3 , similar to the well-established formula for 

 the molecule of boric chloride, B CI 3 , despite the want of relation 

 between the observed vapour-densities of the two compounds. 

 I say " ought probably," because isomorphism does sometimes 

 obtain between elements of different atomicities, as instanced by 

 silver blende, Ag' 2 S, and galena, Pb" S, for example ; whence it 

 is quite possible that B'" 2 O 3 may be isomorphous with All^ O 3 . 

 In reference to this point, it is perhaps worthy of notice that the 

 difference between the atomic weights of boron (11) and alumi- 

 nium (37*5) is substantially the same as that between the atomic 

 weights of fluorine and chlorine, oxygen and sulphur, nitrogen 

 and phosphorus, carbon and silicon, glucinum and magnesium, 

 lithium and sodium, and sodium and potassium respectively. 



1 have put together these few remarks, not at all with the 

 view of deciding the matter, but merely of calling attention to 

 the difficulties by which it is beset. Indeed my own mind is far 

 from made up on the subject • and if I continue to use the time- 

 honoured triatomic formulae, it is not from conceiving them to 

 have a proved superiority, but because their inferiority has not 

 yet been established beyond all question, as well as because they 

 are simpler and more familiar to me. 



XXIII. Remarks on Chemical Notation. 

 By William Odling, M.B., F.R.S., Sec. Chem. Soc. 



To the Editors of the Philosophical Magazine and Journal. 



Gentlemen, 



IN addition to a perfectly gratuitous meddling with me, and 

 an uncalled for sneer at the feeble intelligence of chemists, 

 Mr. Waterston's original letter to you upon the subject of che- 

 mical notation contained an authoritative enunciation of certain 

 supposed new views a long while since adopted, and of other un- 

 doubtedly old views a long while since abandoned by modern 



