of Edinburgh, Session 1880-81. 
49 
when very dilute solutions of the acid sulphate and caustic baryta 
are taken, whilst on the other hand the insoluble body appears to 
be the only product when concentrated solutions are employed. 
The analysis of the insoluble product led us to the very 
unexpected result that it was the oxide of tri-benzyl-phosphine. 
The formation of this body is only intelligible on the assumption 
that it arises from the breaking up of the hydrate, and that toluol 
is produced along with it — ■ 
(C 7 H 7 ) 4 POH = (C 7 H 7 ) 3 PO + C 7 H 8 . 
That this is the explanation of its formation was easily verified by 
mixing the solutions of the acid sulphate and caustic baryta in a 
distilling flask connected with a condenser, and boiling the mixture. 
An oily liquid passed over which was lighter than water, and had 
all the properties of toluol, and its amount agreed with the quantity 
demanded by theory. This reaction which we have described is 
perfectly anologous to the behaviour of tetrethyl phosphonium 
hydrate when heated* 
(C 2 H 5 ) 4 POH = (C 2 H 5 ) 3 PO + C 2 H 6 . 
But we think it remarkable that the hydrate of tetra-benzyl phos- 
phonium should break up so easily. 
Hydrate of tetra-benzyl-phosphonium . — This body is obtained 
when boiling solutions of the acid sulphate and caustic baryta are 
mixed — provided as just stated the solutions be very dilute. It 
separates from the solution filtered off from the sulphate of barium, 
(formed at the same time) after considerable concentration, which 
we always effected by boiling. 
The hydrate is very soluble, and crystallises easily in transparent 
tables of striking refractive power. Its composition was verified by 
determinations of hydrogen and carbon. 
Like other soluble compounds of tetra-benzyl-phosphonium the 
hydrate, even in very dilute solutions, is precipitated by hydrochloric 
acid, the chloride resulting. 
We are engaged in preparing other salts of tetra-benzyl-phos- 
phonium, and in investigations on them, especially the action of 
Hofmann, “Phil. Trans.,” 1857. 
