141 
by a liquid film when stretched and kept at a constant tem- 
perature. Taking ’08 of a gramme weight per centimetre of 
breadth as the capillary tension of a surface of water, and 
therefore T6 as that of a water bubble, I calculate (as you 
may verify easily) that a quantity of water extended to a 
thinness of 2 to-o5o;dto of a centimetre would, if its tension 
remained constant, have more energy than the same mass 
of water in ordinary condition by about 1,100 times as much 
as suffices to warm it by 1 ° cent. This is more than enough 
(as Maxwell suggested to me) to drive the liquid into va- 
pour. Hence if a film of ro-o^b 070-00 of a centimetre thick 
can exist as liquid at all, it is 'perfectly certain that there 
cannot be many molecules in its thickness. 
The argument from the Kinetic theory of gases leads me 
to quite a similar conclusion. 
I need not trouble you with it at present, as I am writing 
a short sketch of those of the results of Maxwell and 
Clausius which I use in it, to form part of an article on the 
Size of Atoms for “Nature.” 
Mr. R. D. Darbishire, F.G.S., exhibited a number of 
beautiful specimens of electroplate reproductions of some 
of the Roman plate lately found at Hildesheim, and gave a 
summary of Mr. Wieseler s very interesting discussion upon 
this find. 
“ On the Determination of Phosphoric Acid,” by William 
Carleton Williams, Student in the Laboratory of Owens 
College. Communicated by Prof. Roscoe, Ph.D., F.R.S. 
Of the many methods proposed for the separation of 
phosphoric acid from the alkaline earths, few are better 
than the one devised by W. Reissig, founded upon a process 
originally described by Reynoso. This method, although used 
in many German laboratories, has, strange to say, found but 
little favour among English chemists. This is probably 
