1898.] BREEDING OE THE DRAGOKBT. 285 



few ripe and active spermatozoa. Another male, taken in company 

 with the last, is 15-4 cm. (6g in.) in length. In structural 

 differentiation it is nearly as advanced as the last, the first dorsal 

 ray reaching the fourth of the second dorsal fin ; but the 

 coloration differs from that of two females of about the same 

 length only in that the dark spot of the first dorsal is bounded 

 anteriorly by the third instead of the second ray, a moi-e or less 

 constant sexual distinction. In this male the testes contain no 

 advanced spermatozoa. It would appear, therefore, that the 

 coloration is more intimately associated with the maturity of the 

 genital organs than is the structural differentiation. 



Another sexually immature male, 18'3 cm. (7f\ in.) long, 

 supports the same conclusion. The coloration is quite undifferen- 

 tiated, though the first dorsal ray nearly reaches the eighth of the 

 second dorsal. The variation of size in relation to sexual maturity 

 is in no way remarkable. 



Smitt (Hist. Scand. Fish. ed. 2, i. p. 273) notes that the male, 

 which reaches a length of 30 cm., is much larger than the female. 

 The latter has not been observed, by Scandinavian naturalists, to 

 reach 25 cm.^ This statement is in general accordance with my 

 own experience of the species. As a matter of fact the difference 

 in size, among the larger specimens, is to some extent due to the 

 greater length of the snout and caudal fin in the male. Thus a 

 male and female measure respectively 24 and 21*7 cm., but in 

 regard to the distance from the front of the eye to the origin of 

 the caudal fin are of exactly the same length. 



In the mature condition, at all events, males appear to be much 

 more abundant than females. On two occasions a trawler has 

 brought me what purported to be his whole catch of Dragonets. 

 In one catch there were 95 males and 21 females, in another 

 72 males and 1 female. The actual proportions may be to some 

 extent obscured by the smaller females escaping through the 

 meshes or being overlooked by the fishermen, but the great 

 discrepancy cannot be altogether accounted for in this way, and a 

 general preponderance of males is borne out by my own "trawling 

 experience. A similar numerical proportion of the sexes obtains, 

 as first pointed out by Cunningham, in Arnoglossus laterna, at 

 least among the large specimens on the offshore grounds. Here 

 also the male is the larger fish. Among Teleosteans the female is 

 nearly always the larger and the more abundant sex. The relations 

 of size are here reversed, and, if food competition severely taxes the 

 weaker individuals, it is not unnatural that the relations of 

 number should conform. Without the necessity of supposing 

 (without proof) that the feebler members are actually starved out 

 in the early stages of life-history, we arrive at the same result if 

 we allow that sex is largely influenced by nutrition. It would 

 appear that a plentiful nutrition favours the production of female 

 individuals. In the case of Callionymus, if the sex is determined, 



^ The smallest ripe female which I have seen measured 16'8cra. (6| in.). A 

 specimen of 14'2 em. was very nearly ripe. 



