SEA-FISHERIES LABORATORY. 97 
traction was incomplete and that there was still some oil in the 
dried residues. That would have depressed the proteid value, 
which is calculated on the wet weights. But renewed extrac- 
tion of the residues, reduced to fine powders, gave no appreci- 
able amount of oil, and had there been such it would have 
become apparent by forming a soap in the process of distillation, 
durmg the Kjeldahl estimations. If the fat is erroneously 
estimated it ought to be an error in excess due to a possible 
extraction of some substance such as lecithin, or to incomplete 
drying of the oil (the carbon tetrachloride being adsorbed to 
- a slight extent), or to oxidation of the oil during the drying in air. 
It is certain that such oxidation occurs. The oil extracted 
is always dark brown in colour, but if the drying of the wet 
flesh is carried out as suggested above, in an atmosphere of 
hydrogen or carbon dioxide, and if the subsequent drying of 
the oil is also carried out, in a similar indifferent atmosphere, 
then the oil obtained is light yellow in colour, with a crystalline 
substance separating out on cooling. That fish oils oxidise in 
air at high temperatures is well known, and the principal means 
of obtaining a colourless, medicinal cod-liver oil that does not 
produce eructation when administered, is the exclusion of air 
during the process of preparation of the oil. An indifferent 
atmosphere of CO, is, in fact, used. Further there are signifi- 
cant variations in the refractive indices of the herring oils ob- 
tained in various ways. Thus we have :— 
Extracted and dried in air ... Refractive index = 1-388. 
Extracted and dried in CQ,... ms » = 1378. 
Kipper oil, air-dried... ... if » = 138. 
_ Red herring oil, air dried... ” 55 == 1-455. 
The previous oxidation of the kipper and red herring oils 
raise the index of refraction. 
But the effect, with regard to weight, is very small, as will 
be seen by drying a sample persistently in the air oven and 
observing the rise in weight. It is, probably, a lipochrome, 
