90 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 
A paper thimble was perforated by a hole made by a cork 
borer and a little loose plug of cotton wool was placed at the 
bottom. The flesh was put in the thimble, with another loose 
cotton-wool plug on top. The thimble was closed by a per- 
forated rubber cork and placed in a wide glass tube contained 
in a brine solution boiling at about 103°C. Air or hydrogen 
or carbon dioxide, previously dried and heated by passing 
through a coil of lead piping contained in the boiling pan, was 
then drawn or forced through the paper thimble. 
The results (to which I refer later on) were not always 
consistent and must be repeated before any conclusions can be 
made. But it was evident (1) that there was some oxidation 
of the oil, (2) that the latter was not rendered insoluble, and 
(3) that increase in the weight of oil need not be taken into con- 
sideration, at all events not in results intended for dietetic 
studies. : 
More important was the great saving of time. By em- 
ploying CO, (because of its higher specific heat) constant weight 
may be obtained in 4 hours, and an apparatus which would be 
very convenient and could deal with a number of samples 
could easily be fitted up. 
However, all results quoted here were obtained by drying 
(at 97°-98° actually) in a steam-oven for about 24 hours at least. 
Extraction of the oil. 
This was carried out in an ordinary Soxhlet apparatus, 
using carbon tetrachloride as solvent. The process was always 
very simple and easy. Rarely, even in very dark oils, was there 
any trace of colour in the solvent after the third siphoning. As 
a rule the extraction was continued for 4 hours but, evidently, © 
long before that time had elapsed the process had been finished. 
At first the last lot of solvent distilling over was tested by 
evaporation on a piece of white paper, but later on a definite 
routine was followed. 
Whether or not appreciable quantities of substances other 
