390 NATURAL SCIENCE. June. 



allied species may be found spawning close together in one confined 

 area, so that if fertilisation is a pretty certain process, the chances 

 appear to be nearly even that cross-fertilisation would also take 

 place. Experiments have shown that the pelagic eggs of one species 

 may be readily fertilised with the sperms of another, not necessarily 

 closely related, though the resulting hybrids usually perish at an early 

 stage.' The diversity in specific spawning-season and of size of egg 

 may be largely accounted for as a partial safeguard against the evils 

 of cross-fertilisation. 



The enumeration of these factors of pelagic spawning points out 

 the lines along which future scientific fishery- work should proceed in 

 the investigation of the life-histories of our food-fishes. Only when 

 we know exactly their spawning-areas and the spawning-seasons 

 can we attempt to close these areas with a maximum of beneficial 

 effect and a minimum of trouble and difficulty to the industries involved. 

 Only when we know the exact spots in the inshore waters most suited 

 for the fry, can we successfully crown the labours of artificial hatching, 

 which attempt to eliminate the natural depleting ravages of wind and 

 temperature, not to mention the fatal effects of non- and cross-fertilisa- 

 tion. Those who deprecate the attempts at artificial propagation now 

 being made would do well to look into this subject of the conditions 

 of pelagic existence. In every important district are required countless 

 meteorological observations of wind and temperature, close investiga- 

 tions of the surface-currents, and a study of the distribution of eggs 

 and fry in the surface- and mid-water. 



Reasons have already been given for regarding the pelagic-spawn- 

 ing habit as the more primitive, and this is borne out by the further 

 study of the specific ontogenetic migrations. These may be expressed 

 graphically in cycles (Fig. 4), showing the gradual transition between 

 the pelagic type (plaice) with a full ontogenetic migration and the 

 demersal type of littoral fish with no migration at all, such as the 

 stickleback. 



In Fig. 4a there is represented the typical cycle of an ideal 

 ancestor of the fishes. Starting at (P) the pelagic water, we may 

 suppose that this form migrated inwards to the pelagic water 

 inshore (PJ, returning at the breeding-season to its offshore spawn- 

 ing-area, so that the drift inshore of the eggs and larvae repeated in 

 ontogeny the phyletic history. The same principle held when, from 

 the pelagic inshore water, the littoral region was reached, and later 

 the deeper offshore bottom-areas. In each case, the pelagic eggs 

 formed a starting-point for the ontogenetic migration. 



The plaice (Fig. 4b) will be seen to follow this migration almost 

 exactly, its ontogenetic migration being well-marked throughout. 



The flounder (Fig. 4c), on the other hand, has complicated 

 matters by itself becoming partially adapted to a fresh-water habitat 

 (F), so that the young forms migrate up the rivers. In this case there 

 1 Cf. Rep. Fish. Beard Scotland, \\., p. 317, 1S91 ; and T. Scott, op.cit., vii., p. 3S2, 1SS9. 



