1898.] BREEDING OF THE DRAGONET. 299 



is as abundant in young examples as in old. I have found neither 

 chromatophores nor reflecting substance in the epidermis. The 

 numerous large epidermal alveoli of the mucous S3'stem need not 

 here concern us, since, though apparently acting to some extent as 

 condensers, they do not alter the effect of the underlying colora- 

 tion-elements and are alike in both sexes. 



The second dorsal fin is one of the parts most conspicuously 

 coloured in the mature male. lu the young undifferentiated male 

 the markings are sombre. If a young specimen be compared with 

 the drawing (Plate XXVI.), it will be seen that the yellow area 

 (of the adult) is brown, the blue lines are opaque white, and their 

 grey margins are colourless and transparent. Sections of this fin 

 show that the skin consists, internally to the epidermis, of loose 

 connective-tissue cells overlying a thin fibrous layer, apparently 

 representing the chorion. Except where the rays intervene, this 

 layer is closely apposed to the corresponding element of the skin 

 of the other side. The chromatophores lie in the loose con- 

 nective tissue already mentioned. 



In the young male the brown bands are found by micro- 

 scopic examination to derive their colour from very numerous 

 yellow and black chromatophores. In connection with the latter 

 are frequently seen underlying masses of finely granular matter, 

 of a brownish colour by transmitted light. As the chromato- 

 phores contract it becomes evident that there is a large quantity 

 of apparently similar matter arranged in a continuous network 

 resembling strands of dendritic chromatophores. By reflected 

 lig t this network takes on a pale yellow colour. It has hardly 

 any iridescence. 



Passing towards the transparent areas which border the white 

 bands, one observes scattered chromatophore-like aggregations of 

 the same substance, some of which contain a little black pig- 

 ment, while intermediate conditions lead up to perfect black 

 chromatophores. It appears, therefore, that these bodies are merely 

 degenerate black chromatophores. The transparent areas are 

 simply devoid, or nearly so, of any sort of coloration-element. 

 The white bands have a few black chromatophores, but the degene- 

 rate structures are much more numerous. The opaque white 

 appearance is derived from a granular reflecting matter, arranged 

 in an irregular network, appearing steel-grey in colour by reflected 

 light over a black surface. So far as I can see, it has no connection 

 with the chromatophores, though it may be of the same chemical 

 nature as the granular matter associated with the latter. 



In the adult breeding male the brown bands are brilliant yellow. 

 This result appears to have been achieved (i.) by the reduction of 

 black chromatophores, which are now much less numerous tlian in 

 the undifferentiated stage and sometimes entirely absent, (ii.) by the 

 excessive development of yellow piguient. 



In the fresh condition the ground-colour of the yellow bands 

 is a diffuse yellow ; no separate yellow chromatophores can be 

 discerned until the diffuse stain has been extracted by a reagent. 



