1898.] BREEDING or THE T)TIAC401S'ET. 301 



blue. It is therefore evident that the prismatic body owes its 

 property of cerulescence to the manner in which its component 

 parts are arranged. Owing to the minute size of the whole body, 

 I have not been able to ascertain its internal structure. The 

 bundles of prismatic bodies (iig. 1 a, p. 289) ' are arranged in a 

 direction roughly parallel to the axis of the fin-ray, the individual 

 bodies transversely. Typically they seem to be arranged in a 

 single layer, but often they overlap one another. The whole net- 

 work lies in a slightly higher plane than the chromatophores, or 

 at least passes over them when they are encountered in the same 

 vertical plane. Sections of the fin are possible only when the 

 bodies have been entirely removed by the action of acids, and I 

 cannot find any trace of their associations with the connective 

 tissue. Presumably they occupy the interstices of the latter, as 

 suggested, for the iridocytes, by Cunningham and MacMunn. 

 Their component rod-like particles may be simply deposited in 

 a regular relation in such interstices, or may be held together 

 by some matrix. None of these bodies occur in the grey margin 

 of the blue baud, where the colour-elements diifer only from 

 the young condition in the much greater abundance of black 

 chromatophores. 



In a male examined shortly after the breeding-season the yellow 

 of the body has faded to a golden brown, while that of the fins 

 is paler than at the time when pairing was in full progress. 

 Examination of the yellow bauds shows that the diffuse pigment 

 is reduced in quantity. In the blue part are noticed aggregations 

 of brownish granular matter, superficial to the prismatic bodies, 

 sometimes alone, often in relation to a black chromatophore. I 

 believe that they are derived from the degeneration of the black 

 chromatophores, as in the case of apparently similar matter in the 

 yellow bands. A reduction of the black chromatophores of the 

 blue band would of course result in a diminution of the blue colour, 

 since the prismatic bodies are blue only when backed by black. 



It will be noticed that in the drawing (Plate XXVI.) the bands 

 of the first dorsal fin are white with grey margins. This is the usual 

 condition, but occasionally they have a certain blue tinge, especially 

 near the base of the fin. The white appearance is derived from a 

 network of reflecting matter, very similar to that of the blue of the 

 second dorsal, but the individual prismatic bodies are considerably 

 smaller, and there are very few black chromatophores in this region. 

 In several specimens I find that the latter are superficial to the 

 reflecting tissue. In an example in which the bands of this fin are 

 distinctly though not very brilliantly blue, I find the prismatic 

 bodies as large and as numerous as in the second dorsal. Black 

 chromatophores are somewhat less abundant than in the other fin, 

 but have the same relation to the reflecting matter, which is fully 

 cerulescent over a black surface. The deficiency of colour is thus 



' Fig. 1 a is only a sketch. The minute size of the bodies renders the use of 

 the camera lucida impossible in their individual delineation. 



