128 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 
100 % as the deviation from the theoretical sum of the constitu- 
ents present. 
Most probably this deviation is to be accounted for by the 
presence, in variable quantity, at different seasons, or in 
herrings variously cured, of some substance in which the nitro- 
gen is present to a greater or lesser extent than 16 %. What 
this substance may be has still to be determined, and the in- 
vestigation obviously has much theoretical importance. 
The emperature ‘Fa chore 
Inspection of the Tables on pp. 93-4 will show that the 
percentage of fat in herring flesh varies in a regular manner. 
It is least (for the Manx fish) in May (when the sea-temperature 
is rising from its minimum in March) ; it increases to its max1- 
mum in August (when the sea-temperature is also maximal) 
and it then decreases as the sea-temperature begins to fall 
towards its next minimum. The two variables, fat-percentage 
and sea-temperature, can be plotted on the same time-scale of 
abscissae, making the two ranges of ordmates cover the same 
part of the vertical scale, and it will then be seen that there is 
a certain degree of correspondence in the curves—the variables 
are correlated. 
The fat-percentage is, of course, a function of several 
independent variables, sea-temperature being only one of the 
latter. The sexual cycle is a function such that the fat-con- 
tents of the “ flesh’? accumulate, while the ovaries or testes 
are growing in mass, and also such that the percentage of fat 
diminishes rapidly just before and after the act of spawning. 
The large increase of oil in the flesh may be regarded, from one 
point of view, as an excretion. While the gonads are growing 
in mass the proteid percentage in the flesh is decreasing, and it ~ 
may be suggested that proteid is being transferred from muscles 
to gonads, becoming rearranged, with the elimination of fatty 
acid residues, and that the latter accumulate in the flesh. This 
