696 REPORT OF COMMISSIONER OF FISH AND FISHERIES. 
29° C., two precautions are necessary: (1) the use of a large excess of 
the molybdate solution (unless nitric acid aloue should perform the 
same office), and (2) allowance of ample time for precipitation. Neglect 
of these precautions involves risk of loss of phosphoric acid The 
effect of higher temperature was not tested. Practically, we have used 
25 cc. of molybdate solutiou and allow from 36 to 48 hours for the pre- 
cipitation. 
SULPHUR. 
About 1 gramme of the partly dried substance was oxidized as for the 
determination of phosphorus. The cooled mass was dissolved in hydro- 
chloric acid, and the sulphuric acid determined by precipitation with 
barium chloride. 
It is well known that the presence of salts of the alkalies, including 
sodium chloride, when present in considerable quantities, may affect 
the precipitation of sulphuric acid by barium chloride, the precipitate 
of barium sulphate bringing down with it under some circumstances 
more or less of the alkaline salt. The ordinary means for avoiding 
this consist in diluting the solution, precipitating hot, and washing the 
precipitate with hot water. In view of the large quantity of sodium 
chloride present in these determinations, it seemed desirable to study 
the conditions under which pure and impure barium sulphate precipi- 
tates are formed a little more closely than had, so far as I am aware, 
been done. For this purpose a series of experiments was devised. 
The details were faithfully carried out by Mr. J. P. Bartlett, then a 
student in this laboratory and now chemist in the Maine State College 
Agricultural Experiment Station. 
The plan consisted in taking a solution of sulphuric acid and deter- 
mining the amount of acid both in the solution alone and after adding 
different quantities of sodium chloride. Comparative trials were made 
with concentrated and dilute solutions and by precipitating hot and 
cold. The sulphuric-acid solution used was the same as employed for 
nitrogen determination with the soda lime. The sodium chloride solu- 
tion contained about 25 per cent, of sodium chloride, and the barium 
chloride solution about 20 per cent, of the ordinary crystallized salt. 
Without going into details of the experiments, it will suffice to say 
that, although when precipitated cold from concentrated solutions the 
barium sulphate is apt to be too heavy, i. e., to bring down sodium 
chloride, yet, when precipitated hot, or even precipitated cold from 
dilute solutions, it was uniformly pure. This is in accordance with the 
previous observations on which general practice is based. Our deter- 
minations served simply to show the limits of concentration and amount 
of sodium chloride within which it is safe to work. It should be noted, 
however, that the solutions contained in all cases a small amount of 
free hydrochloric acid, and that only a small excess of barium chloride 
was used in the precipitation. Practically, the ordinary way as recom- 
mended by Fresenius is accurate for these determinations, even in pres- 
