— 8 — 



the number of specimens of A. japonica would, to judge by the experience gained 

 in all other similar countings, limit the P. F. A. for this species, so as to let the 

 P. F. A. of the two species fall clear of each other. 



We have thus seen, that the average number of anal rays, like that 

 of the vertebrae, reveals a distinct difference between the European, 

 American and Japanese eels. We have now to consider whether local »races« 

 or »smaller species« can, by the aid of this character, be proved to exist within 

 each of these species. This point has been investigated only in the case of the 

 European eel, owing to the inadequate amount of material available for the other 

 two species. 



The samples selected for investigation were taken from the most widely differing 

 localities whence it was possible to obtain such; viz; a Baltic (Sealand, Denmark), 

 an Atlantic (Bristol Channel, England) and a Mediterranean (Cette, France). The 

 detailed results are given in Table III, 1 — 3. The averages for the three samples 

 are, for Denmark 215.176, England 215.257, and France 215.24. Considering the 

 extremely vide range of variation, 72, 62, 71 for the three samples, the averages lie 

 surprisingly near each other; so near, as to render it immediately obvious that no 

 racial difference can be shown to exist among the European eels by 

 means of the number of anal rays. Examination of the P. F. A. for the three 

 samples, as well as the fact that the average for the Mediteranean sample lies 

 midway between the values for the two from northern Europe, both prove this 

 to be the case. 



The anal rays are, like the vertebrae, present or distinctly forming in their 

 full number at a very early stage in the life of the individual. This may be seen 

 from the very extensive material of growing larval stages procured by »Kommis- 

 sionen for Havunders0gelser« in the course of a cruise made in the Atlantic by 

 the M/S »Margrethe« in 1913. This material shows, that the larvae of the European 

 ell have already at a length of some 4 cm. their full number of anal rays^ 



The number of dorsal rays has not been subjected to closer investigation; 

 although there is hardly any doubt that they may be used for the characterization 

 of species in the same way as the anal rays. That they have not here been thus 

 employed is due to the fact that the still greater number of rays in the dorsal fin 

 would render the work even more difficult than in the case of the anal, in addition 

 to which, the foremost rays in this fin are small, and frequently difficult to count 

 with certainty. 



' The following data regarding the larvae of the European eel may here be given: One specimen 

 of 30 mm length exhibited a ver}' large part of the fin where forming rays could not yet be distin- 

 guished. In the case of three specimens of 36, 37 and 38 mm were counted 201, 183 and 209 rays 

 respectively, in the foremost part of the fin, however, the rays (or interspinous rays) were not yet 

 forming, or at least, not to be counted with certainty. In the following 5 specimens, all the rays 

 were apparently forming; length: 40, 41, 43, 47, 50 mm, with 215, 226, 213, 205 and 223 rays respec- 

 tively. It may be added, that the full number of segments is present at an even earlier stage. This 

 it was possible to determine, both in the case of A. vulgaris and A. rostrata, in larvae of abt. 2 cm 

 in length, not, however, in those 1 — I'/j cm long. 



