410 TRANSACTIONS OP SECTION D. 



of the eels is fully unravelled, but I am able to discuss some of the previously 

 unsolved problems on which our investigations have thrown light : for example, 

 the mode of life of the larval stages and their distribution as compared with the 

 distribution of the older fish. These investigations can best be discussed in 

 chronological order, and this method will also show how we have been obliged 

 slowly and gradually to abandon the prevailing views of the larvae as larvae 

 which belong to deep water, or live in the sea-bottom itself, or near to it :— 



1904. — In the month of May, as mentioned, we found a Leptocephalus 

 brevirostris in the Atlantic, west of the Faeroes. 



1905. — In June 1905 we continued our investigations with the ' Thor ' from 

 the Faeroes in a southerly direction along the continental slope west of 

 the British Isles. Larvae of the eel and conger were then found in quanti- 

 ties to the west of the 1,000-metre line, right from the Faeroes to Brit- 

 tany. On the other hand, none were found to the east of this line, though 

 the same apparatus was used. Thus they were absent from the North 

 Sea, Channel, the Skager Rak, and Danish waters. The natural conclu- 

 sion I drew from this was that the two species of eels mentioned do not 

 spawn at the British coasts, nor in the North Sea or waters further east, 

 but in the Atlantic to the west of the 1,000-metre line. The stock of eels 

 in these waters must therefore come from the Atlantic Ocean. This was 

 sufficient to show that the conditions in the northern waters was something 

 quite different from those prevailing in the Straits of Messina, where the 

 larvae of the conger and eel live at the same places as the older fishes. 



These investigations of June 1905 also showed that both the conger 

 and eel larvae are true pelagic animals, having nothing whatsoever to do 

 with the bottom. They live, in fact, in the uppermost 200 metres, and 

 are often found just under the surface. The larvae found were all fully 

 developed, those of the eel about 75 mm., and those of the conger about 

 double this size — measurements that agree fairly well with the data of the 

 Italian observers. As we did not find the younger developmental stages, 

 the eggs or earliest larvae, I suggested, influenced by the Italian state- 

 ments that the eels come into the world and pass the first period of their 

 lives in deep water, that the youngest stages would probably be found 

 deep down in the sea, perhaps at the very bottom. This was one ex- 

 planation, at any rate, of why we had not obtained them. Our later investi- 

 gations have shown that this suggestion was not correct. 



1906. — The investigations in 1906 were also made in the Atlantic, west of 

 Europe, but extended further to the west out over greater depths than in 

 1905. We found that the larvae occurred not only in the belt between 

 1,000 and 2,000 metres, as in 1905, but also everywhere to the west of 100 

 metres. They were even found in quantities over the greatest depths of 

 about 5,000 metres. In May the larvae occurred further out than in 

 September, and the majority of the eel larvae were in process of meta- 

 morphosis in September, whilst none were so far advanced in May. This 

 does not hold good for the conger, however, as the metamorphosing stages 

 were mostly found in May. I was obliged to conclude from the investi- 

 gations that the eel larvae come from places where the depth is very 

 great, at least 4,000 to 5,000 metres, and from there move in towards the 

 continental slope. During this movement the metamorphosis of the larvae 

 begins, that of the eel in the autumn, of the conger somewhat earlier. 



After finding the larvae of the eel over depths of about 5,000 metres, 

 I could no longer believe that this fish spawned at the bottom, and I put 

 forward the suggestion that reproduction takes place far out in the ocean 

 and independent of the bottom. In the Atlantic, north of Spain, we have 

 up to the present found about eight hundred larvae of the eel and about 

 fifty of the conger. For the Mediterranean, on the other hand, the rela- 

 tive abundance is reversed. 



1907. — The year 1907 was used for the collection of material regarding the 

 distribution of the fresh-water eel. The result was that the fresh- water 

 eels are only found in the Northern Atlantic, where their distribution 



