NOTES ON TELEOSTEAN OVA AND LARVAE. 39 



had a length of 2-42 mm., from the snout to the anus being I'O mm. 

 The pigmentation of the former consists of a very conspicuous mid- 

 post-anal patch, made up of a large dorsal and a large ventral chro- 

 matophore (the latter accompanied anteriorly by a relatively small 

 pigment spot), a small hypural patch, while pre-anally there is a group 

 of chromatophores on the side of the body in the pectoral region, a 

 series dorsal to the gut, one large dendritic chromatophore below the 

 anus and extending partially over the oil-globule, and some pigment 

 in the eye and on the head. The pigmentation of my second larva 

 at a slightly older stage, when most of the yolk had been absorbed, 

 differed only from the above in being somewhat more densely aggre- 

 gated. Two large dendritic chromatophores — one dorsal and one 

 ventral — with outgrowths extending laterally as far as the notochord, 

 constituted the mid-post-anal group, a continuous line of dendritic 

 pigment ran along the whole peritoneal region, one large dorso-lateral 

 patch of pigment occupied that part of the trunk above the base of 

 the pectorals, two small chromatophores were on the head, and the 

 iris was now completely black. Holt (lib, PI. YI, Fig. 53) has 

 figured a larva (" Species III (Motella ?) ") somewhat similar to this, 

 but having an anterior group of post-anal pigment spots much re- 

 sembling that of 0. mustcla, which it also approaches in having many pig- 

 ment spots over the oil-globule. Ehrenbaum (5d, p. 278) considers this 

 may possibly be 0. tricirratus, Bl. The egg had a diameter of 0"72 mm. 

 (and less) and an oil-globule of 17 mm., and the larva in its first day 

 was 2"20 mm. long. My larvae certainly have a closer resemblance to 

 Raffaele's 0. tricirratus, BL, than this of Holt's. I may mention, how- 

 ever, that Ehrenbaum {ibid.) warns one not to attach too much weight to 

 Raffaele's identification of the parent fish as 0. tricirratus, BL, suggest- 

 ing that the Naples observer did not wish to designate this species as 

 distinguished from 0. mcditcrrancus (L.), but merely referred to the 

 common tricirrate form of the Bay of Naples, which appears to be 

 0. mediterraneus (L) ( = 0. tricirratus, Briinnich), and not 0. tricirratus, 

 Bloch { = Onos vulgaris, Yarr.). I do not see, however, that there is 

 any solid ground for doubting Raffaele's identification in this case. 



Onos, Species B. 



A much more problematic form of larva, which I will term " Onos, 

 sp. B," was hatched from a rockling egg taken in the young-fish trawl 

 2 miles S. of Rame Head on 2nd June. The egg had a diameter 

 of 0-68 mm., and its oil-globule 0145 mm. When the embryo had 

 developed a short free caudal portion, it was marked with large 

 black pigment spots, and the stellate chromatophores in the pellicle 



