of the Fishery Board for Scotland. 



219 



region of the skull. This dimension is not always satisfactory, because the 

 supraoccipital spine is liable to injury. The second point on the base of 

 the skull is fairly well-defined by the markings on the parasphenoid bone, 

 which are left when the pharyngeal teeth are removed. These are 

 attached to the expanded part of the parasphenoid under the basioccipital 

 by strong ligaments, and when the latter are taken away a rough patch 

 on the parasphenoid forms a good end-point. The line joining this to the 

 commencement of the supraoccipital spine passes at right angles to the 

 longitudinal axis of the skull. 



Table XV., p. 234, shows that this dimension does not alter either with 

 maturity or with growth. The averages remain practically stationary 

 throughout, and the slight differences there indicate neither increase 

 nor decrease. The depth of the skull, therefore, grows at the same rate 

 throughout relative to the skull-length, but the latter, as has been shown, 

 decreases relative to the body-length by '75°/ O0 , when the total length of 

 the body increases by 1 00mm. Hence, relative to the body-length the 

 depth of the skull decreases by , 26°/ 00 approximately. 



It has been shown, therefore, that for the three dimensions of the skull 

 whose variability is recorded here a decrease in the rate of growth takes 

 place relative to the body-length between 300 and 400mm. If the rates 

 of decrease of the three dimensions considered be taken relative to the 

 total length of the fish, the amounts become *5, '45, and *18°/ 00 for an 

 increase in total length of 100mm. The rates of decrease, therefore, for 1mm. 



50 45 18 



of increase of total length are jfo, and or 



10 4 , 10 4 , 10 4 . 



If the skull were a regular six-sided figure, with length, breadth, and 

 depth corresponding to the actual dimensions of the skull, this would 

 mean — neglecting the products of the above fractions — that a relative 

 113 



decrease in volume of or about 1 °/ 0 0 , took place. For such a box this 



10 4 , 



relative decrease would be very small, but in the case of a skull whose 

 volume is only a fraction of that of the box the relative decrease is 

 considerable. It is not simply the volume that is affected, but the weight 

 or mass also, and this must be nf some moment. As already mentioned 

 in connection with the relative decrease in skull-length, the reason of the 

 decrease in the skull probably lies in the necessity for a certain constant 

 balance between the weights of head and body always being preserved. 

 On account of the greater mass of the skull relative to the mass of the 

 body, if the dimensions of the head and body increased uniformly 

 together, the head would become relatively much heavier than the body. 

 This would disturb the balance and form of the animal — the qualities 

 most fixed in the species, if anything can be regarded as fixed. In the 

 section on sex-variability it has been shown that the weight of the skull 

 is probably greater in the females than in the males, and this was 

 also explained by the necessity of preserving the balance between head and 

 body. In the herring, again, Heincke has shown that the relative posi- 

 tions of the fins, as well as the length of the bases of the fins, vary during 

 growth, and he has correlated these changes with the movements of the 

 animal, and consequently with the interbalancing of the separate 

 structures and organs. 



If we take up now the question of maturity, or the ripening of the 

 reproductive organs, in the light of this conception of the balancing of 

 organs, we see that we have at one and the same time a physiological 

 cause of the changes in the relative positions and sizes of organs during 

 growth, and, on the other hand, a reason for the changes. It has been 

 o 



