1898.] PLANKTON OF THE FAEROE CHANNEL. 553 



? S. scoticiis, Giiather, Chall. Eep., Zool. xxxi., Pelagic Fishes, 

 p. 31. 



S. mulleri, Collett, JSTorweg. N. Atlant. Exped., Zool, JFisli. 

 p. 158. 



Benthosema mulleri, Goode & Bean, Ocean. Ichth. p. 76. 



Young, imperfectly characterized stages of the Scojjelidce have 

 been a source of much labour to the various observers who have 

 had occasion to name collections of this group, since it has been 

 quite impossible to determine, in the absence of sufficient material, 

 whether certain differences have a systematic or merely a develop- 

 mental value. I think I am right in saying that the reproduction 

 is quite unknown, and I can find no description of the early larvse 

 of any species. With regard to one, the efficiency of Dr. Eowler's 

 vertical net appears to have supplied this want, as I find in his 

 collection what appears to be a nearly complete series of Scopehis 

 gJacialis. 



Though the method has its disadvantages, it appears necessary in 

 the present instance to describe the different stages in the inverse 

 order, commencing a^ ith the most advanced. This is a specimen 

 of o8-5 mm., exclusive of the lower jaw and the caudal fin-rays'. 

 It has the adult characters peculiar to the species and, except that 

 most of the scales have gone and some of the fin-rays are broken, 

 is in good preservation. No description is necessary except for 

 ontogenetic comj^arison. The radial formula is £). 13, A. 18. 

 The eye is nearly three times as long as the snout, and is y%- of the 

 length of the head (12 mm.), which is equal to the height of the 

 body at the shoulder and a little more than J- of the total length. 

 The anal commences nearly opposite the middle of the dorsal. 

 Adequate figures of the adult stage, which is practically exemplified 

 in this specimen, are given by Goode and Bean and by Smitt, but 

 in some copies of the Hist. Scand. Fishes the printing is very 

 imperfect. A clear diagram of the photophores is given by Liitken. 



Figures 1 to 7 (Plates XLYI. & XLVII.) represent younger 

 stages in Dr. Fowler's collection. The most advanced of these, 

 fig. 1, measures onl}'' 14-5 mm., and has no scales ; but the condition 

 of another specimen indicates that the body is covered under 

 natural conditions with dark-coloured scales. The part shaded in 

 my drawing remains, in formol, a bluish grey. The photophores, 

 having the formula of S. glacialis, are intact. The radial formula 

 is D. ] 2 or 13, A. 18. The proportions of the head, eye, and snout are 

 respectively as 31 (= 4 mm.), 10, 7. The eye is thus much smaller, 

 relatively, than in the specimen of o8'5mm., and the snout longer. 

 Considered in the light of the ordinary ontogenetic changes of 

 these parts in Teleosteans, this condition would appear to prove 

 that the two individuals belong to different species, since as 

 a general rule the eye decreases and the snout increases as develop- 

 ment advances. In Scopehis, as I shall show, this condition is 

 reversed during some part of the metamorphosis of the larva. 



' This limitation is implied in all measurements of total length in this paper. 



