— l82 — 



and belonged really to the surface. Eigenmakn reasons thus, that if the eggs of the eel 

 are lighter than the ocean water (which however, has not been proved) and further have 

 oil-globules they must rise up from deep water to the surface. 



As we have now learnt that quantities of large pelagic fish eggs really occur in the 

 deep water of the Atlantic, where they normally pass through their development without 

 rising to the surface, and which in structure show an extraordinary resemblance to the 

 eggs of the eels, there seems to me every possible reason to expect that the latter occur 

 normally under similar conditions. The experiments with my large bathypelagic eggs 

 showed indeed that they are not, as Eigenmann says of the muraenoid eggs, lighter than 

 sea-water, but on the contrary have almost the same specific gravity as the bottom-water 

 and will therefore under normal conditions not rise up to the surface. We 

 must accordingly rather consider the occurrence of such eggs at the surface as occasional, 

 as by far the greater number undoubtedly complete their whole development in deep 

 water ^. If this holds for the eels as I have shown it does for other deep-water fish 

 (e.g. Argentina sihis), then many things and much otherwise puzzling in the biology of 

 the eels will find therein a natural solution (cf our almost complete lack of knowledge 

 of the preleptocephalic larval stages of the eels). 



4. Distribution of the larvae of the eel (Leptocephalus brevirostris) 



in the Atlantic Ocean 



a. Our discovery of the larvae 



The distribution of the larvae of the eel as shown by our investigations with the "Thor" 

 can be seen from the Chart PI. X, where the stations at which we have found the eel 

 larvae are marked in red. In contrast to this the stations marked in blue indicate where 

 no eel larvae were found, in spite of the fact that the apparatus used and the conditions 

 were such that the eel larvae, had they been at all present, would certainly have been taken. 



It is necessary to say a few words to explain why these negative results are also 

 included on the chart. If we had made our investigations with the same, small apparatus 

 of little fishing capacity which until a few years ago were the only ones used, I should 

 have hesitated to lay so great weight on the negative results, as the eel larvae would 

 often be quite able to escape the pursuit of such apparatus. Fortunately we have now 

 for several years possessed a very good fishing apparatus, namely, the so-called young- 

 fish trawl, constructed by Dr. C. G. JoH. Petersen ^ and thanks to this, we can also 



I During the "Thor 's" investigations in Skager Rak (58°05' N., 8°24' E., depth: 520 meters) in July 1906 

 I made several hauls (each of one hours duration), from which it appears very clearly that the eggs oi Argen- 

 tina situs belong to the deeper water-layers as can be seen from the foUoving list: 



(i) 50 meters wire out, gave: i egg 

 (z) 100 — — — — : 2 eggs 



(3) 150 — — — — : 9 — 



(4) 200 — — — — •24 — 



(Note added during press.) 

 2 Amongst the many improvements in the sphere of scientific fisheries technique, with which Dr. C. G. JOH. 

 Petersen has done good service, the construction of the young-fish trawl is not the least important. An extensive 



(5) 300 meters wire out, gave: 126 eggs 



(6) 400 — — — — : 165 — 



(7) 600 — — — — : 236 — 



(8) 700 — — - — : 918 — 



