694 REPORT OF COMMISSIONER OF FISH AND FISHERIES. 
with ether before using. We have found it convenient, however, to 
duplicate the apparatus, using several at once. A rectangular box of 
zinc or galvanized iron, 120 cm. long, 20 cm. wide, and 25 cm. high, and 
provided with eight pipes of block tin which serve as worms, makes a 
very good cooler for the ether. The radiation of heat from the water is 
perhaps sufficient to keep it cool enough to condense the ether, but we 
generally have a current of water running through the cooler when in use. 
The point at which the extraction is completed we have not always 
found easy to determine. The methods ordinarily recommended, of 
evaporating a drop of the ether after it percolated through the sub- 
stance, either on a piece of paper or on a clean w T atch-glass, and noting 
whether a transparent spot or a residue remains, have not in our 
experience been satisfactory. We prefer to continue the extraction 
for such time as experience has indicated to be usually sufficient, and 
then remove the flask and substitute another, repeating this latter 
operation until the new flask shows no gain in weight after the extrac- 
tion has been continued for some time. When the extraction is believed 
to be complete, the apparatus is taken off from the condensing tube, 
the inner tube holding the substance is taken out, a test tube substi- 
tuted, the apparatus put back in place, and the ether in the flask is 
again warmed, allowed to condense and run back into the test tube, 
and thus most conveniently recovered for subsequent use. The small 
portion of ether still adhering to the extract in the flask is removed by 
heating in a current of hydrogen in the apparatus used for drying. 
What other substances besides fats are thus extracted from the flesh 
of fish, oysters, etc., is a matter which I have not investigated, contenting 
myself for the present with simply calling the material extracted by 
ether, ether extract. As the analytical details show, the material used for 
fat determination was dried in the hydrogen before extraction with ether. 
By this means two possible sources of error are avoided, to wit : 
(1) Certain fatty substances, as is well known, are rendered insoluble 
or very difficultly soluble in ether by being heated in air. This altera- 
tion, due, I suppose, to oxidation, is very marked in many vegetable 
substances. Thus I have found the‘larger parts of the fats of linseed, 
maize, and grasses to be rendered insoluble in this way. To what ex- 
tent any of the fats of the animal tissues which we have analyzed would 
be thus affected is a matter which I regret to have been unable thus far 
to study. That the error by drying in air would be large seems to me 
improbable. 
(2) When the material extracted contains considerable quantities of 
water, or when commercial ether containing water is used, substances 
other than fats may possibly be dissolved. When and how greatly 
the presence of water would affect the results I can not tell, though I 
do not believe the error would be great in such substances as we are 
here dealing with. 
The ether extract was nearly always more or less colored. How far 
