— l62 — 



canal was empty of food, from which must be concluded that the eel larvae had not 

 been feeding at the present stage. By this reason it seemed highly probable that 

 the appearance of the larvae must have undergone some change during the 2 — 3 months 

 since we first got them ; in fact this must necessarily be the case. As I thought it highly 

 desirable to determine the question, we went out, when returning from Iceland to Den- 

 mark, to the first station (between St. Kilda and Rockall) where the Leptocephali were 

 found in June and made several hauls there with the young-fish trawl on August 31st 

 and September 1st. In all 8 specimens were taken, when a western gale stopped our 

 work here over the deep water. Just as supposed all the 8 specimens appeared to 

 be in the process of metamorphosis. (Cf. PI. VIII, Fig. 3 — 9.) 



After taking further hauls in shallower water near the coasts of Lewis and Scotland 

 without finding any specimens we went through the Pentland Firth to the North Sea and 

 proceeded to Copenhagen. 



3. The larvae (Leptocephalus brevirostris) of the eel and their biology 



It was not difficult to prove that the Leptocephali we found in the Atlantic are really 

 the larvae of the eel {Leptocephalus brevirostris). This is evident in the first place 

 from their agreement in all essential points with the specimens described and figured by 

 Grassi and C.ilandruccio. In the second place I have found, as the figures on Plate VIII 

 show, all the transition stages from the typical broad Leptocephalus appearance, which 

 has been well compared to a Nerium leaf, to the well-known, cylindrical, narrow elver. 

 In the third place it appears that the number of segments in the Leptocephali is in re- 

 markable agreement with the number of vertebrae in the common eel {Angmlla vul- 

 garis), as can be seen from a comparison of the following list, compiled as the result of 

 an investigation by Cand. mag. A. Strubberg, with the numbers given in Sect. II, Chap. 9. 



Number of segments in the Leptocephalus hrevirostris taken by the "Thor" in the Atlantic'. 



Number of 

 segments 



117 

 116 



"S 

 114 



"3 

 112 

 III 



Number of specimens 



0000 



000000000000 



0000000000000000000000000000 



oooooooocooooooooooo 



000000000000000000 



00000000000000000 



000000 



As already mentioned, our larval eels were taken partly in the middle and second 

 half of June 1905 and partly on September ist of the same year. The June material 

 consists of ca. 265 specimens, all essentially at the same stage of development, that stage, 

 namely, which has for many years gone under the name of Leptocephalus brevirostris 

 and which in form resembles a Nerium leaf (see Fig. 1—2, PI. VIII and PI. VII). On the 

 other hand there are 8 specimens composing the September material which are consider- 



' The segments were counted under the microscope on material preserved in formaiine. Sometimes the 

 limits between the segments are rather indistinct, which may perhaps cause a few of the numbers to be i loo low. 

 This could have been avoided by clearing the specimens (in xylol oi- something similar), but this seemed unneces- 

 sary as it was only desired to prove that the Leplocephali belonged to the common eel {Anguilla vulgaris) 

 and not to study races when each single number has to be absolutely certain (see later on, Section II, Chap. 9). 



