128 
Proceedings of the Royal Society 
results is more than I can explain. Seeing that baryta was not a 
reliable reagent for the titration of ethers, I tried a solution of car- 
bonate of sodium, hoping that it would not appreciably act on hard 
glass, but without success. 
A solution of the carbonate, when heated in a sealed-up tube, took 
up enough of alkali from the glass for its titre being increased by about 
? ^th. At last I found that, of all the available alkaline solutions, 
caustic ammonia was the only one which could be safely heated in 
sealed-up glass tubes. In order to see now if ammonia also was 
capable of completely decomposing compound ethers, a series of 
experiments was instituted, in which measured quantities of stan- 
dard ammonia (containing about 17 grms. NH 3 per litre) were 
heated with acid acetate of ethyl, containing a known quantity of 
acetic acid. Such mixtures were procured by heating to 130° — 
140 J C. &c., in sealed-up tubes, weighed quantities of glacial acetic 
acid, with a large excess of alcohol, a direct experiment having 
shown that, after a few hours heating, 60 % of the acid em- 
ployed was etherified. The amount of ammonia saturated by 
the acid, after it had passed into the state of ether, was com- 
pared with that which the same quantity required when mixed 
directly with the alkali. Of the results obtained in this manner 
some came surprisingly near the truth, others differed by a per 
cent., and some even more ; this I can only explain by the fact 
well known to chemists who are familiar with titrimetric analysis, 
that in the saturation of a weak acid with ammonia the point of 
neutrality is not by any means well marked. It may be, also, that 
in some cases the ethers had not been completely decomposed. 
Although the conditions under which this method would be sure 
to give exact results might, perhaps, be established by an extended 
investigation, I preferred to go back to using baryta- water, as re- 
commended by Berthelot, substituting only for the sealed glass 
tubes a copper digestor, soldered with sterling silver, and provided 
with a conical copper stopper, which could be securely fastened by 
means of a screw. The stopper kept perfectly tight, even under a 
pressure of two atmospheres, as was shown by charging the diges- 
tor with water, weighing the arrangement, keeping it at 120° C. for 
about four hours, and weighing again. There was not any loss of 
weight observable by a balance turning with two centigrammes. 
