OF ARTS AND SCIENCES. 175 



than 0.0010-|- grm. Throughout the entire series of experiments the 

 tendency to yield figures slightly larger than the truth is manifest, but 

 the error is quite within legitimate limits. The greatest care was 

 taken to secure similarity of conditions under which the crucible and 

 lime were weighed before and after the evaporation and absorption of 

 boric acid, and the weight after ignition was taken in every case after 

 cooling over sulphuric acid during a definite period of ten minutes in 

 order to eliminate as far as possible the effect of atmospheric condensa- 

 tion upon the large surface of platinum. Ignitions were always fin- 

 ished over the blast-lamp, and constancy of weights secured. 



The results of both modes of treatment are on the whole satisfactory, 

 and equally so. 



In the presence of chlorides, it is of course impossible to employ 

 nitric acid to free the boric acid. Oxalic, citric, and tartaric acids also 

 liberate hydrochloric acid to a considerable extent from alkaline chlo- 

 rides. It was found, however, that when acetic acid was distilled over 

 sodium and potassium chlorides only traces of hydrochloric acid passed 

 into the distillate, and experiments (21) to (23) were made to deter- 

 mine whether these amounts are sufficient to vitiate the separation of 

 boric acid from alkaline chlorides by distillation in presence of free 

 acetic acid. The details of treatment were identical with those of 

 experiments (17) to (20), excepting only the addition of 0.5 grm. of 

 sodium chloride to each portion before distillation. 



The mean error of these results is about 0.0009-|- grm., and it is 

 plain that the presence of sodium chloride does not materially change 

 the conditions of the experiment. There seems, therefore, to be no 

 reason why boric acid may not be separated by distillation from alka- 

 line chlorides in presence of free acetic acid ; but it was found that the 

 presence of any considerable amount of potassium acetate is disadvan- 

 tageous. Sodium dcetate to a reasonable amount does not interfere 

 with the favorable progress of the separation ; but potassium acetate 

 appears to require a much higher temperature for the expulsion of its 

 water, and longer distillation. 



When, therefore, chlorides are present in the salts from which boric 

 acid is to be removed by distillation, the choice is open between two 



