CHEMICAL COMPOSITION OF FOOD-FISHES. 
697 
ence of large amounts of sodium chloride, provided the proper precau- 
tions are observed. The experiments satisfied us that the presence of 
the sodium chloride in the determinations of sulphur, as we conducted 
them, did not interfere with the accuracy of the results. 
CHLORINE 
was determined by burning the partly dried substance in platinum 
evaporating dishes, as in the determinations of phosphorus and sul- 
phur, and estimating the chlorine in the fused mass with ammonium 
sulphocyanide by Yolhard’s process.* Dr. J. H. Long, then assistant 
in this laboratory, but now professor of chemistry in the Chicago Med- 
ical College, by whom the determinations were made, carried out, at 
my request, a series of experiments which led to the observance of 
certain precautions and showed that with them very satisfactory results 
may be obtained in the determination of chlorine in such substances 
as those we were working with. The precautions are contained in the 
following statement of the method followed in the determinations of 
chlorine in fish flesh. 
(1) £he substance was very slowly fused with a mixture of potassium 
nitrate and sodium carbonate so as to avoid any possible loss by spirt- 
ing. It was observed that when the platinum capsule, in which the 
ignition was carried on, was covered with a watch glass, small portions 
of the substance were thrown out against it unless the operation was 
conducted very slowly. With care, however, there was no consider- 
able loss, as indicated by either the appearance of the watch glass or 
actual determinations with sugar and a known amount of chlorine. 
(2) The fused mass was dissolved in chemically pure nitric acid, which 
was made quite dilute, since by using a stronger acid some chlorine 
could easily be driven off, as was found by experiment. 
(3) To the solution thus obtained an excess of silver nitrate solution 
(decinormal or half decinormal) was added, and the whole boiled on 
tho water bath for about 2 hours. This long boiling with excess of 
nitric acid, added after the silver nitrate, was found necessary to expel 
nitrous acid coming from the reduction of the nitrate in the fusion. 
(4) After the boiling, the solution was allowed to become quite cold 
before titrating. 
As the details of the experiments on the determinations of phos- 
phorus, sulphur, and chlorine are somewhat extended and will prob- 
ably be printed in full in another place, they are not inserted here. 
PROXIMATE INGREDIENTS DIRECTLY DETERMINED. 
In a number of the specimens determinations were made of the ingre- 
dients soluble in cold and hot water, and of the portion not dissolved 
by water, alcohol, or ether. The object was as much to test the rneth- 
* Leibig’s Annaler, 190, 1. 
