126 NOTES ON THE REPRODrCTION OF TELEOSTEAN FISHES 



Unidentified Labroid, resembling Coris. 



Holt, E. W. L., Sci. Trans. E. Dub. Soc, S. II., iv., 1891, p. 467, Pis. 



XLVIII., LI., Sj). v., Coris-like. 

 Holt, E. W. L., Avn. Mus. Marseille, v., 1898, Fasc. II. Egg and larva, with 



figure. 



An egg which appears to have unquestionable Labroid affinities 

 was taken in the Plymouth district in July and August, 1897. Our 

 records show that it occurred on more or less off-shore grounds. I 

 have no doubt but that it is identical with ova already described from 

 tlie west coast of Ireland, but the parentage remains in doubt. 



The Irish specimens measured from -80 to '83 mm. in diameter, the 

 oil-globule measuring '15 mm. Those taken at Plymouth measure from 

 •78 to "81, the oil-globule from 13 to 15 mm. The yolk is homo- 

 geneous, the oil-globule colourless, but, in the Plymouth examples at all 

 events, very dark. A larva, measured very soon after hatching, is 

 2"21 mm. long. A somewhat more advanced specimen from Ireland 

 measured 2*44 mm. The conformation of the larva bears a striking 

 resemblance to Coris julis (vide liaffaele, Mittheil. Zool. Stat. Neap., 

 viii., 1888, Tav. II., Figs. 18, 19). The yolk is pyriform, its narrow 

 end, having the globule at the apex, projecting boldly in front of the 

 head. The rectum is separated by a considerable interval from the hind 

 end of the yolk. The marginal fins are of moderate width, the dorsal 

 arising behind the head. My notes distinctly state that the edges 

 of the marginal fins are not serrated as in Coris. The notochord, 

 in Irish specimens, is of a peculiar type. For the most part arranged 

 in a double series, the vacuoles are occasionally unicolumnar. This 

 appears to me a strong indication of Labroid affinity, since both Coris 

 and Ctenolabrus exhibit a notochord intermediate in character between 

 the unicolumnar and multicolumnar conditions, though the approach to 

 the former condition is much more marked than in the form before us. 

 I must add that I did not find any unicolumnar cells in a I'lymouth 

 example which I examined ; it is probably a variable feature, but, on 

 account of its rarity, in so far as my knowledge extends, in other groups 

 of fishes, not the less useful. 



The pigment is all black, and has the same distribution as in 

 Coris, but resembles perhaps even more closely that of Mullus. 

 Indeed, save for the presence of cortical yolk segments in Mullus 

 and for the separation of the yolk and rectum in our unidentified 

 larva, the two forms are extremely alike. 



I believe that the Labroid affinities of the parent are fairly well 

 demonstrated by the characters of the embryo and larva. The diiliculty 

 is to find a Labroid parent. Labrus, Crenilabrus, and Ctenolabrus are 

 naturally eliminated. Centrolahrus exoktus spawns, at least in great 



