176 PROCEEDINGS OF THE AMERICAN ACADEMY 



metbods : the distillation may be made directly witb an excess of 

 acetic acid ; or the hydrochloric acid may be first removed by means 

 of silver nitrate, and the distillation of the filtrate proceeded with at 

 once, or after precipitation of the excess of silver salt by means of 

 sodium hydrate or carbonate, care being taken to acidify again suf- 

 ficiently with nitric acid after the removal of the silver. Of these two 

 modes of proceeding, I incline to the treatment with nitric acid and the 

 removal of the chlorine by precipitation ; and this method has been 

 used with success by others as well as myself, for some months, in the 

 analysis of waters carrying boric acid, and natural borates. 



The process in either modification is fairly accurate and easily exe- 

 cuted, and admits of very wide application. Insoluble compounds in 

 which the boric acid is to be determined may be dissolved in nitric 

 acid at once, or, if necessary, first fused with sodium carbonate ; and, 

 fortunately, nearly everything which is volatile in the subsequent 

 treatment and capable of forming with lime compounds not easily 

 decomposable by heat may be removed by known processes. The 

 combination of fluorine, silica, and boric acid is perhaps most difficult 

 to treat; but the precipitation and removal of the first as calcium fluo- 

 ride from the aqueous solution of a fusion in alkaline carbonate may, 

 it is believed, be effected with care, and the mode of procedure from 

 that point is simple. 



The number of distillations necessary depends, of course, upon the 

 amount of boric acid treated. To remove 0.2 grm. of boric anhydride 

 completely to the distillate, six charges of methyl alcohol, of 10 cm.^ 

 each, proved, as we have seen, to be ample. 



The apparatus by the aid of which the distillation processes which, 

 have been described were carried out has found useful application in a 

 number of other processes. In the determination of free and albu- 

 minoid ammonia in wafers which can be boiled quietly with difficulty, in 

 the methods of estimating hydrofluoric acid which involve the expul- 

 sion of silicon fluoride from a mixture of the fluoride with sulphuric 

 acid and silica, in the separation of iodine from bromides and chlorides 

 by distilling with ferric sulphate and sulphuric acid, and of bromine 

 from chlorides by means of permanganic acid, it has proved of value, 

 and will doubtless be found convenient in many analytical processes 

 in which quantitative separations by the distillation of liquids liable 

 to spatter or boil explosively are involved. 



