392 NATURAL SCIENCE. June. 1897. 



is a migration sea-wards, but the eggs are small and laid comparatively 

 close inshore. 



The next stage in this process is probably seen in the eel (Fig. 4d.), 

 in which the fresh-water habitat has become more pronounced, and, 

 indeed, the normal one for the adult. It is possible that the egg is 

 pelagic, and, after the curious leptocephaline postlarval stage, the 

 adolescent forms (elvers) migrate up the rivers. 



In the demersal series, the herring (Fig. 46) lays its eggs on the 

 sea-floor in moderately deep water, and the larva remains in the same 

 habitat. The postlarval stage, however, moves upward, and passes 

 its time in the pelagic water, eventually coming in to the littoral region. 

 Here the demersal condition is seen in its earliest stage, and from the 

 other littoral fish can be selected a series (Fig. 4f) in which the 

 ontogenetic migration is gradually given up, till, in such as the 

 viviparous blenny and the stickleback, the young forms immediately 

 assume the habits and habitat of the adult. 



Certain of these littoral fish have migrated to the fresh-water — 

 e.g., salmon (Fig. 4g), — and these, having in their young stages no 

 adaptations to pelagic life (as in the case of the eel) to contend with, 

 the eggs are laid far up the rivers, secure from the many dangers of 

 the teeming littoral region, though the young salmon, in obedience to' 

 ontogenetic exigencies, come down the rivers and complete a migra- 

 tion in the sea. The extent of this is not quite known, but salmon 

 have been caught in the pelagic water far out at sea, so that it is 

 possible that the cycle is fairly well repeated. In the case of the 

 freshwater fish, this cycle is finally given up. 



The katadromous type is essentially that of a pelagic-spawning 

 fish with secondary freshwater leanings, and the anadromous that of 

 a demersal-spawning fish with the same proclivities. 



It is instructive to note that the ontogenetic migration is 

 eliminated by a process of shifting environmental necessities to a later 

 stage of development. In the herring the sojourn in pelagic waters 

 devolves on the post-larval form, and in the salmon it is further 

 relegated to the adolescent. A stage further back eliminates the 

 migration from the life-cycle altogether. In one case, the angler 

 (LopMus piscatoriiis), there are good grounds for believing that an 

 inshore pelagic spawning-habit has been acquired. The eggs float 

 about in huge masses, and their structure and that of the larvae is 

 distinctly of the demersal type. The fate of the larvae on hatching 

 is unknown. 



It must not be supposed that, by these types being placed in 

 gradation, there is necessarily any genetic connexion implied between 

 the adults. In this, as in other groups of animals, the closest allies 

 exhibit remarkable diff'erences in their ontogeny, and the gradual 

 transition in ontogenetic processes may quite well be worked out from 

 the life-histories of a series of widely divergent species. 



University of St. Andrews, N.B. Arthur T. Masterman. 



