310 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1921 



beyond my comprehension. Viewed in the light of our subsequent 

 investigations, however, the explanation is perfectly'' simple. The 

 larvae of the European eel which are found at Bermuda and in the 

 western Atlantic generally are, owing to the vicinity of the breeding 

 grounds, all young, belonging to the O and I groups, and conse- 

 quently as yet far from that stage of development at which they 

 seek the shores. It is only larvae of the American eel which are found 

 so far west in the full-grown and metamorphosic stages, at which 

 they are drawn to the coasts. Consequently, Anguilla rostrata is the 

 only one of the two species which " lands " at Bermuda, the hosts of 

 larvae of Anguilla vulgaris which surround these islands being only 

 passers-by on their way to a far more distant goal. 



As we have learned, the breeding gro'unds both of the European 

 and of the American eel lie west of longitude 50° W. Although the 

 larvae of Anguilla rostrata become far more numerous in proportion 

 from east to west, it is nevertheless a fact that there are extensive 

 areas where the larvae of the two species are greatly intermingled. 

 The Dana stations provide many examples of such intermingling, as 

 shown in the photograph, Plate 5, and the graph, Figure 12. 



It is very natural then to ask : If the larvae of the two species are 

 found greatly intermingled in certain areas of the ocean, how is it 

 Ihat the stock of eels in Europe is, practically speaking, pure, i. e,, 

 composed exclusively of Anguilla vulgaris and that in America of 

 Anguilla rostrata^ The question may also be formulated as follows: 

 How do the masses of larvae in the western Atlantic sort themselves 

 out, so that those individuals which belong to Anguilla vulgaris ulti- 

 mately find themselves in Europe, while those of Anguilla rostrata 

 " land " on the shores of America and the West Indies ? 



In the main, the question is no longer difficult to answer. In the 

 case of the American eel, the pelagic larval stage is terminated in 

 about one year; consequently the larvae have not time to make the 

 journey to Europe, the distance being more than they can cover in 

 that period. It is otherwise with the European eel, which takes 

 nearly three times as long over its larval development, as a result of 

 which practically all of them are far away from the western (Amer- 

 ican) portion of the Atlantic when the time comes for them, as 

 elvers, to seek the coasts. 



We can thus indicate both a geographical and an ethological cause 

 for the distribution of the two species of fresh-water eels. The for- 

 mer lies in the fact that Anguilla rostrata has its center of produc- 

 tion somewhat farther west and south than Anguilla vulgaris. The 

 latter is the different duration of the pelagic migratory stage. These 

 tvvo facts, in conjunction with the ocean currents as an aid to trans- 

 port, and later — once the earliest stages of development are past — 

 the active movements of the larvae themselves, must be rejjrarded as 



