127 
of Edinburgh, Session 1866 - 67 . 
in themselves a qualitative analysis on a large scale, the exactitude 
and reliability of which go far beyond what could be obtained by 
ordinary testing. This, however, does not apply to many organic 
preparations, and is eminently inapplicable in the case of com- 
pound ethers. If in the preparation of such an ether, we start 
with a pure alcohol and a pure acid, and carefully distil, wash, and 
redistil the ether produced, the product ultimately obtained may 
contain considerable quantities of alcohol, free acid, and water. 
The purity of a compound ether can be established only by direct 
experiment. Free acid is easily tested for, and even quantitatively 
determined by the well-known acidimetric processes ; the presence 
or absence of water may be proved by an elementary analysis, but 
an admixture of alcohol is not easily established. An elementary 
analysis would, in a majority of cases, show very little, because the 
difference between the percentages of carbon in the ether and al- 
cohol respectively, is usually too small to admit of basing upon a 
small fraction of that difference anything like an exact calculation. 
Thus, for instance, acetate of ethyl contains 54‘54 per cent, of 
carbon, and alcohol 52T8 % — the difference is 2*36; a decrease of 
OT °/ 0 in the carbon would therefore correspond to an admixture of 
about 4 per cent, of alcohol. 
Many years ago F. Mohr proposed a method for quantitatively 
testing the officinal “ asther aceticus,” which consisted in heating 
the ether in a closed vessel, with an excess of standard alkali, and 
determining the amount of alkali saturated. The method was in 
1863 extended and improved upon by Berthelot, who seems to have 
employed it extensively in his researches. The standard alkali 
used by Berthelot was caustic baryta, which he heated with the 
ether under examination in a glass tube. On attempting some 
time ago to employ Berthelot’s method for testing, what I had 
good reason to look upon as very pure formiate of ethyl, I was sur- 
prised by obtaining perfectly inexplicable results; and on subject- 
ing his process to a critical examination, I found that the discre- 
pancies in my results arose from the action of the baryta solution on 
glass. When baryta water was heated to 120° C. in sealed-up tubes, 
and kept at that temperature for about twenty-four hours, the glass 
was found to be strongly attacked, and the titre of the baryta solu- 
tion considerably altered. How Berthelot managed to obtain exact 
