754 REPORT OF COMMISSIONER OF FISH AND FISHERIES. 
lime, through which the products of the combustion were obliged to pass. When, 
however, only the fine soda lime was used the colored products passed over without 
being decomposed. 
Almen insists upon the finest practicable pulverization and careful 
mixing of the substance with soda lime, which latter had been care- 
fully tested by combustion with sugar to prove its purity : 
As an indication that the nitrogen determinations were reliable and correct* 
within the limits required for their purpose, I may state that two determinations 
with fish meal gave, respectively, 12.17 per cent, and 12.21 percent, of nitrogen, and 
two others with fresh flesh of cod gave, respectively, 2.63 per cent, and 2.72 per cent, 
of nitrogen. The disagreement of duplicates thus did not exceed 0.1 per cent, of 
nitrogen or 0.5 per cent, of protein. 
In speaking of our own attempts at determining the nitrogenous 
ingredients directly, Section A, Division 2, I referred to the fact that 
these determinations were made as much for the purpose of testing the 
methods, and for comparing the results with those obtained by other 
investigators, as for the sake of the results themselves. I had especially 
in mind the work of Almen, who has made a considerable number of 
estimations of albumen, other substances soluble in water, gelatin- 
formers, and the extractive matters. The descriptions of his methods 
as well as his results seem worthy of somewhat detailed statement here, 
since this work with that of Kostytschef, which I presume to have been 
conducted by similar methods, constitutes the most extensive contribu- 
tion we have to the knowledge of this particular subject. 
Almen lays especial stress upon the comparison between the compo- 
sition of beef and that of the several kinds of fishes which he analyzed, 
the analyses of the beef and fish being made by the same methods. 
The albumen soluble in water, the gelatin-forming substances (gelatin- 
oids as they are sometimes called), and the extractive matters were 
estimated as follows : 
The determination of these several constituents were madq in each case in a single 
portion of the flesh, generally about 33.3 grammes, which had been finely divided 
and well mixed as previously stated. The flesh was placed in about 250 grammes of 
distilled water, stirred and allowed to stand from 8 to 12 hours, with frequent stirring, 
and then filtered through paper by means of a Bunsen filter pump. The uudissolved 
portion was then extracted in like manner a second time, with the same quantity of 
distilled water, and washed once on the filter. The filtrates and the washings, which 
together made about 600 cubic centimeters, were then boiled down in a porcelain 
dish to about 70 to 100 cubic centimeters. In this process the albumen separated out 
in coarse fiocculent masses which were easily filtered. They were placed upon the 
filter, washed with hot water, completely dried at 110° C., and weighed between two 
watch-glasses. In working with some kinds of fish, it happens that in the second 
extraction with cold water the insoluble portion assumes a gelatinous form similar 
to syntonin, and is difficult to filter. By use of the Bunsen pump and a goodly amount 
of patience, the filtration always succeeded. 
The longer the extraction was protracted with new portions of water, the more 
gelatinous the residue became, without, however, any considerable quantity of solu- 
ble albuminous matter passing into the solution. This was particularly the case 
with perch. Whether this peculiarity of the flesh of some of the fishes depended 
