62 



NATURE 



[March 9, 191 1 



myomeres cannot be counted with certainty, but the 

 pigmentation is sufficiently characteristic. 



In the accompanying table (Fig. i) I give a 

 graphic summary of the Conger larv.i; tal<en on the two 

 cruises of the 'thor. Without further explanation, it will 

 be evident from this table that the youngat stages, of 

 1-4 cm. in length, have only been taken on the summer 

 cruise, the older intermediate stages, of 5-9 cm., only on 

 the winter cruise, whilst the full-grown Leptocephali, of 



ft 

 /Z 



9 

 8 



7 

 6 



♦ 



Z 

 4 



o • • • 



o o o • • 



OOOoOOOOCtt 



o o O o • • 



o o o o o • 

 00000 



OOOO0OO0O9 



0000000000 



oovoooooooooooo 



0000000000000 



oooaoooooooo 00000000 



oooeooooooooooo 



0000000000000 



0000000000 



Fig. 



yo is M W~ 



Numbers of Larvae. 

 -Conger larvae 



IT 



• Summer cruise (vi, vii, viii). 

 o Winter cruise (xii, i, ii). 



ca. 12-13 cm. in length, were taken on both cruises, though 

 mainly in the summer. From the sharp distinction between 

 the youngest specimens taken on the summer cruise and 

 the larger of the winter cruise, we are obliged to conclude 

 that the spawning of the Conger takes place mainly, if not 

 exclusively, at a certain definite time of year, namely, in 

 spring and summer. We may further conclude from 

 the material that the larval group of about 7 cm. is about 



half a year old, and that most of the 



larvae of 12-13 cm., some of which 



are in process of transformation, are 



about one year old. 



.All the larvae recorded in the table 



were taken in the Mediterranean or 



in the Atlantic off the Straits of 



Gibraltar. In earlier years (1905-8), 



when the Thor was working in the 



northern parts of the Atlantic, west 



of the British Isles, we also found 



several Conger larvae during the 



period May to September, but these 



were all older stages, more than 12 



and up to 16 cm. in length. The 



observations made on our last two 



cruises in the Mediterranean throw 



a new light on these older dis- 

 coveries, for we now know that it is 



extremely easy to take the earliest 



Conger larva; in the upper layers, 



when we know just where to find 



them. We may, in fact, conclude, 



with a high degree of probability. 



that the Conger does not spawn in 



the region examined by the Thor to 



the west of the British Isies, but 



further south in the Atlantic (Fig. 2, 



off the Straits of Gibraltar). 



A more detailed account of the 



distribution of the different develop- 

 mental stages will be given later, 

 but on the accompanying chart I endeavour to sum- 

 marise the occurrence of the youngest specimens, from 

 1-4 cm. long, all of which were taken on the summer 

 cruise. It will be seen from this chart that these earliest 

 developmental stages were mainly taken over very great 

 depths, outside the 2000-metre line (or near this), but not 

 in the many hauls which were made in the shallower 

 waters. It appears, further, that the largest hauls (in- 



NO. 2158, VOL. 86] 



eluding newly hatched larv») were, made at the deepest 

 places the Thor had visited, namely, over depths of more 

 than 3000 metres or about 3000 metres, in the Levant, the 

 Ionian and Tyrrhenian Seas, as also in the deep ba.sin 

 between .Sardinia and the Balearic Isles, where we have 

 taken twenty to sixty larva: in quite short hauls with 

 pelagic apparatus. 



These discoveries show that, when the time for repro- 

 duction arrives, the Conger seeks out from the coasts to 

 great depths, where it spawns mainly in the deepest and 

 most central parts of the basins. 



In an earlier paper (1906) I suggested that the youngest 

 larval stages of the eels might be bathypelagic, that is to 

 say, living at great depths below the surface. Perhaps 

 our most important discovery now is that the earliest 

 pre-leptocephalous stages of the Conger, as well as of the 

 other four to five species the earliest larvae of which I 

 know, really belong to the upper layers. This can b*- 

 seen from the hauls at any of our stations where such 

 eel larvae were taken in quantities. Our practice at each 

 station was to make a series of hauls at varying depth-;, 

 the length of wire out being 25, 100, 300, 1000, and 2oo(» 

 metres, and the large quantities of the fry and eggs were 

 always taken in the first, but few, or none at all, in th<' 

 others. As the depth fished in could not have exceedf'd 

 15 metres, we must conclude that the youngest stages and 

 the eggs belong normally to the uppfermost layers of 

 water. The older stages (Leptocephali), on the other 

 hand, are also found in greater depths, 100-200 metres 

 below the surface, as I have already shown in my earlier 

 papers, and the life-history of the eel larvae is thus no 

 exception to the general rule applying to fishes with 

 pelagic eggs, namely, that the earliest stages are passed 

 at or near the surface, and that as development proceeds 

 the larvae sink down into greater depths. 



The renowned Straits of Messina have also become 

 famous in connection with our present subject as the first 

 place where eel larvae were found. -So far as I know, tho 

 eel larvae have never before been discovered in any 

 quantity in the open waters of the Mediterranean, and 

 this is the reason, I believe, why the reputation which the 

 Straits of Messina obtained through the discovery of the 



Fig. 2. — Chart of the Mediterranean and .\tlantic, showing where the Conger larvse were taken. 



Leptocephali there was. to a great extent, undeserved ; in 

 any case, the theories which the Italian observers founded 

 on this discovery regarding the bathypelagic or demersal 

 mode of life of the pre-Leptocephali and Leptocephali can- 

 not be maintained. The conditions in the Straits where 

 deep-sea forms and surface forms occur together at the 

 surface are exceptional and abnormal, and the phenomena, 

 biological and physical, we find there are not at all indi- 



