564 iiK. E. ^y. l. holt on the [June 21 , 



with advance of general development. In tlie number of myomeres 

 both Paeroe and Iceland forms agree well enough with Mallotus. 



The latter has not been recorded from any point nearer to Ihe 

 Faeroe Channel than the coast of Xorway, but appears to be a fish 

 of pelagic habit, approaching the coast only for the purpose of 

 spawning. Tlie ova are demersal, and it may be objected that our 

 Faeroe larvas are too young to be found so far from land. This 

 objection depends for its validity on a knowledge of the rate of 

 growth, which is not forthcoming. 



Altliough I think I haAe demonstrated the possibility of 

 connecting the Faeroe larvae, through intermediate stages as 

 represented by the Iceland series, with the adult form of MaUotus 

 villosus, I do not think we are justified in considering the question 

 settled. The fact is that we know next to nothing of the 

 development of man}' marine forms and especially of the );elagic 

 and bathybial species, nor can it be supposed likely that a few 

 sporadic cruises Jiave furnished us with an even approximately 

 complete list of the fish-fauna of the Faeroe Channel. In all proba- 

 bility there is a strong resemblance between the larvae of man}' 

 physostomous fishes, however widely they may be separated in the 

 adult condition. Of the method of reproduction of bathybial 

 fishes, whether by pelagic or deuiersal ova, we are in most cases 

 ignorant. The characters of the Faeroe ]ar\ a, though piol ably 

 sufficient to exclude it from the Muranidse, are such as might 

 occur equally in a Salmonoid, Scopeloid, or Clupeoid. Any 

 Clupeoids known as inhabitants of the region may be eliminated, 

 since we know the larval stages of all of them. The same remark 

 applies, as I think, to Argentina sj^jhyrona ; specimens of 37 Kim. 

 have already acquired the adult conformation \ though only about 

 13 mm. longer than the Faeroe example, Mhich is still practically 

 undift'erentiated. The size-interval does not appeal" sufficient, and 

 I imagine that this species of Argentina has a shorter larva, with, 

 of course, Icwer myomeres. A. sihin has 65 to 68 vertebrae and 

 is a much laiger fish. It may conceivably pass through a larval 

 stage like the Faeroe form if its pelvic fins undergo an anterior 

 migration. Among the Scopeloids Sfoinias is an elongate form, 

 and S. ferox has been recorded by Giinther from the Faeroe 

 Channel (Chall. xxxi. op. cit. p. 31). 



However, the example in question, though capable of even 

 specific determination, was again only 37 mm. in length ; w hile I 

 can find in the Faeroe lar^a of 24-5 mm. no trace of the barbel 

 and enlarged teeth of Stomias. 1 have already referred to the 

 characters of the Faralepidcf, and the enumeration might be 

 prolonged but always without bringing us, for the present, any 

 nearer to a definite conclusion. 



Dr. Fowler's specimens w ere taken as follows : — 



13 i. 60° 2' K., 5° 49' W. 100 to fathoms. Two, 19 and 

 24o mm. 



20 c. C(f IC' X., 5" 49' W. 400 to 300 fathoms. One, 17 mm. 



' Holt & Caklerwood, Trans. E. Dubl. Soc. ser. 2, v. 189i", p. 5C9, Bg. 3. 



