m GENERATION OF FISHES AND THEIR FECUNDITY 79 



the fertilisation of the eggs but the fact that the male keeps 

 guard over the eggs until they are hatched, and until it is 

 explained how twenty-five females can supply enough spawn to 

 keep one hundred males occupied in paternal nursing duties we 

 can scarcely feel confident that the figures are correct. 



It is obvious that the above considerations have no application 

 to the case of' differences in the proportion of milt to roe among 

 fishes which agree in producing spawn of the buoyant kind. 

 Other things being equal, we may reasonably suppose that the 

 disproportion will be greater when the eggs are larger. For the 

 quantity of milt required will depend on the number of the 

 eggs, not on their size, and therefore if the eggs are larger the 

 roe will be larger while the size of the milt remains the same, or 

 if the size of the roe remains the same the number of eggs will 

 be smaller, and the milts will be diminished in size. But 

 according to this reasoning the milt ought to be larger in the 

 sole than in the plaice, instead of much smaller as is actually the 

 case, for the eggs in the sole are smaller and more numerous than 

 in the plaice. The most important consideration is certainly the 

 rate of spawning, that is to say, the number of ripe eggs shed at 

 the same time, or in one act of expulsion. Here again more 

 exact determinations are required, but it is a matter of general 

 observation that the cod and allied fishes shed their spawn much 

 more rapidly than the flat-fishes, and among the latter the plaice 

 flounder, and turbot more rapidly than the sole. 



The differences in the rate of spawning become very obvious 

 when one is pressing ripe eggs from the various fishes for the 

 purpose of artificial fertilisation. In most cases, for instance in 

 those of the haddock and plaice, a large proportion of the con- 

 tents of the ripe roes can be squeezed out and fertilised at one 

 operation. On the other hand, when I was studying the eggs 

 of the sole I found always the greatest difficulty in obtaining a con- 

 siderable number of eggs from a number of spawning fish, while 

 when squeezing the males I never saw large drops of milt escape 

 on pressure "as in other species, and had to effect fertilisation by 

 cutting out the milts and mincing and pressing them in a vessel 

 of sea-water. The differences in the rate of spawning are also 

 evident from the condition of different roes when they are 

 minutely examined in the ripe condition. In such roes as those 

 of the cod and plaice a large proportion of the eggs are found 



