GENERAL DISCUSSION. 9 



affected by longer and longer intervals, the length of these last being 

 dependent on the duration of the exposure of the species to the lower 

 temperature and minor factors. In this more or less lengthened increase 

 of time required for development of eggs or growth of young appears 

 an effective agent in isolation and in differentiation of the isolated species. 

 What the effect of the retardation may be in regard to a tcnrlency to ovo- 

 viviparity is problematical. 



Some of the deep sea fishes evidently, as suggested by Agassiz, have 

 been vertically derived from pelagic Berycoids, Scopeloids, etc. ; otliers, as 

 is shown with tolerable conclusiveness in the collection before us, are 

 descendants of forms living on the bottom in shoal waters, forms that have 

 gradually traversed the slopes from the shores to the abysses. The larval 

 fishes secured by the tow-net at intermediate depths are mostly young of 

 pelagic species and young deep sea Scopeloids and others descended from 

 a pelagic ancestry. 



Reasons for concluding tliat many of the less powerful swinuners among 

 the bottom fishes have worked their way down, slid down as might be said, 

 are seen on comparison of shoal and deep water forms in such cases as those 

 of the Oncocephali [MaUhe, Cuv.) and the Halieutoids, of the Lophioids and 

 the deep sea Pedicnlates like Ceratias and allies, of the Liparids of the shoals 

 and the deep sea Discoboles, and of the shoal water Pleuronectoids and their 

 abyssal relatives. Similar evidence is seen in the Eaiidte, the Zoarcidi^, the 

 MuraenidiB, the Myxinidae, and others. 



It was expected that by means of the tow-net near the surface and at 

 various distances below it, something might be learned concerning the young 

 of species frequenting the great depths. Such expectations have not been 

 realized. Numerous young fishes were obtained in the net, but they are 

 identified with pelagic Berycoids, Scopeloids, and others of the intermediate 

 iipper waters, and in greater part with the shoal water fishes, Sebastoids, 

 Percoids, Pomacentroids, Scomberesocoids, Squamipinnes, Plectognaths, 

 Pleuronectoids, Mursenoids, etc., and together they give no light on the 

 breeding habits of bathybial species. This experience is similar to that of 

 the " Challenger," which also by towing secured a fine lot of specimens of 

 pelagic species, but among them few or none of those of the bottom. Among 

 the deep sea specimens at hand there are small ones, more often apparently 

 taken in depths somewhat less than those given for the large individuals of 

 the same species. In these depths the temperatures were commonly some- 



