284 ME. E. W. L. HOLT OK TfiB [Apr. 19, 



of a fully mature male, such individual differences as are met with 

 in the number and distribution of the bands being of no importance 

 here. The bands of the first dorsal are less frequently a pale 

 blue, the colour manifesting itself rather often in the detached 

 markings near the base of the fin. The eye, reddish brown or 

 cupreous in females and young males, becomes in large breeding 

 males of a brilliant metallic blue-green, most readily comparable 

 to the palest tint present in the ocellus of a peacock's " tail " feather. 

 On the ventral surface the throat, the underside of the pelvics, 

 and the anal fin become adorned with black pigment, the anal 

 occasionally showing an additional bluish tinge. 



The colour-change of the eye appears to be of a seasonal 

 character. The other changes are of a more permanent nature, 

 though the bx'illiant coloration shown in the drawing (Plate XXVI.), 

 especially in so far as regards the yellow bands of the body, is 

 of comparatively brief seasonal duration. As in the case of the 

 structural characters, the full development of the coloration-change 

 is not necessarily achieved before the fish is fully mature. Thus 

 Dr. G. Johnston (Zool. Journ. iii. p. 336) records a " Sordid " * 

 Dragonet with milt, and Mr. Gr. A. Boulenger has shown me a 

 male, with ripe milt, in which the secondary characters are only 

 slightly developed^. 



This is by no means surprising, since males of Armglossus 

 laterna^ and Coris julis^, which, as I beheve, undergo asexual 

 dimorphism exactly parallel to that of Oallionymus, are frequently 

 mature in the undifferentiated livery. It is, I suppose, unnecessary 

 to adduce instances among the higher animals in which the sexual 

 maturity of the male is accomplished before the full development 

 of the secondary differentiations. 



The smallest sexually mature male Dragonet which has come 

 under my notice measures 16 cm. (6| inches). The extremity of 

 the first dorsal ray reaches the base of the seventh ray of the 

 second dorsal fin. The blue and yellow coloration is present, 

 but the yellow especially is much less brilliant than that of large 

 males taken at the same time. There is very little black pigment 

 on the throat, and only the hinder rays of the anal are somewhat 

 dark. The testes are small, but contain many advanced and a 



' It is perhaps unnecessary to state that Uie Sordid Dragonet, C. dracun- 

 culus, comprising females and undiflPerentiated males, was once held to be 

 distinct from the Geuimeous Dragonet, C. lyra; the latter term being reserved 

 for fully-developed males. 



- Day (Fish. Gt. Brit. i. p. 176) quotes an observation of the Rev. G-. Harris, 

 ■who stated that he had found hard roe in a " Gemmeous" Dragonet. I can 

 only associate myself with the compiler's comment that this observation is 

 interesting, if correct. 



^ Cf. Cunningham, Proc. Zool. Soe. 1890, p. 540 ; Holt and Calderwood, 

 Sci. Trans. E. Dub. Soc. ser. ii. v. 1895, p. 488. 



* The specific identity of C.julis and C. giofrcdi has been denied by Goiirret 

 in the most positive manner (Ann. Mus. Mars. iv. 1893, no. 3). I have re- 

 examined the question at Marseilles, where Gourret's material was procured, 

 and can find no support for bis conclusions. A note dealing with the subject 

 is now in the press (Ann. Mus. Mai"s., ser. 2, Bulletin i.). 



