1 89S.] PLiNETOW OF THE TABROB CHANNEL. 557 



are abdominal. The posterior region of the tail is imperfectly 

 segmented. 



Two less advanced larvae, 6-5 and 4-5 mm. in length, may be 

 taken together. The smallest, fig. 7, has about 31 myomeres, 

 some 14 overlying the alimentary tract. The tail is practically 

 diphycercal. The specimen of 6-5 mm., fig. 6, has the caudal 

 metamorphosis more advanced, and shows an early condition in the 

 development of the anal fin. The alimentary canal is much aUke 

 in both, but in the smaller the anterior part is nearly straight. In 

 the larger there is a slight post-oesophageal dilatation, presumably 

 the stomach. This is followed by a (pyloric ?) constriction, distal 

 to which the gut at once expands and is slightly bent towards the 

 left side in front. Posteriorly it tapers to the region of the 

 rectal valve. I cannot detect distinct transverse ridges, bnt there 

 are some indications of a folding of the lining membrane of the 

 wider anterior part, and I think that this condition may well 

 represent an earlier stage of the voluminous intestinal tract of 

 the more advanced larvae. A large stellate chromatophore in the 

 abdominal roof, about midway between the supposed, pylorus and 

 the anus, apparently overlies a small vesicle, not very clearly 

 outlined. This may represent the air-bladder, and there are 

 indications of its connection by a duct with the anterior part of 

 the alimentary canal. In both specimens the marginal fin-fold is 

 much abraded, but is certainly ampullate in its anterior region. 



I have not attempted in my drawings to restore it to what may 

 be presumed to be the natural proportions. The teeth are small 

 and not very numerous. The proportions of the head, rather 

 injured in both specimens, are, I think, correctly represented in 

 the drawing \ 



The general conformation, pi-oportions of abdomen, and a sufficient 

 harmony in the number of myomeres seem to reasonably connect 

 these larvae with the smallest (8 mm.) of the serieS of Scopelus 

 glacialis'^. It may be objected that in the undoubted Scopeli of 



II mm. and upwards, the proportions of snout and eye have been 

 shown to change in a manner inverse to that which obtains in the 

 two smallest larvae. I think, however, the increase of the eye is 

 a secondary condition of comparatively recent establishment. In 

 the earliest stages I imagine that the eye and snout retain the 

 proportional metamorphosis common in the development of 

 Teleosteans, the snout gradually elongating as development pro- 

 ceeds. This would go on until the attainment of a condition 

 roughly corresponding to that shown in fig. 5. Thereafter the 

 eye commences to increase in size until the adult proportions are 

 attained. Such a condition appears to me more natural than an 

 entire inversal of the metamorphosis whereby the ordinary pro- 

 portional growth of eye and snout would be reversed from the 

 earliest stage of larval development. In Arnoglossus laterna the 



1 The specimen of 65 mm. haa only one eye, wbich, whether naturally or 

 otherwise, is oblique in position. 



■^ For an intermediate specimen, see note on p. 560. 



Pboc. Zool. Soc— 1898, No. XXXVII. 37 



