— 156 — 



eels on the surface at Naples. Why, under the circumstances, he concludes that "they 

 remain at great depths in the sea and only exceptionally, for unknown reasons, some of 

 them mount to the surface" is not apparent. Being lighter than sea water, having oil- 

 globules, and being in all respects typical pelagic eggs found on the surface, we must 

 conclude that "if fertilisation takes place at great depths" it must be "only exceptionally, 

 for unknown reasons", that they remain at great depths. The fact that Rai-i-aele never 

 secured eggs younger than when the gastrula was well formed would favour the sup- 



J mm. 



Fig- 5. Egg and larva of a species of eel, perhaps Conger vulgaris. A. Egg with embryo. B. Larva 

 some days old from a similar egg. C. The head of a larva hatched from a similar egg, greatly 

 magnified. These very small larvae are seen to have the projecting teeth so iiharacleristii: of the Lep- 

 tocephali (cf. Fig. 6 a). Reproduced from Eigenmann (Bull. U. S. Fish Commission, vol. XXI for 1901, 



Fig. I, p. 40, Fig. 14—15, p. 44). 



position that they were fertilized at a great depth and rose slowly in the water" (Eigen- 

 MANN 1. c. p. 39). 



Whilst it appears therefore determined that the Muraenoid eggs hitherto known are 

 pelagic, it is thus still an open question as to whether they normally belong to deep 

 water or to near the surface, or whether they move passively during development in a 

 vertical direction from deep water up towards the surface. Data for the understanding of 

 this question will be given later in the present paper (see p. 179 — 182). 



