1898.] BREEDING OF THE DRAGONET. 303 



Underlying the chromatophores of the inner layer of the skin 

 in the blue bands (and elsewhere) are nests or aggregations of 

 reflecting tissue in minute particles. Such are everywhere present 

 in similar association with the black chromatophores in both sexes. 

 When the chromatophore is fully expanded its central part is 

 transparent and practically colourless, and the underlying nests 

 can be seen from the outer side. In such aspects they are not 

 refractive, but, if the preparation be reversed, they are fouud to 

 have in part optical properties similar to those of the prismatic 

 bodies. Particles of a nest will be found to glitter with the same 

 blue colour, but in parts of the skin where yellow pigment occurs 

 the latter aifects them very strongly, causing the refraction to be 

 chiefly yellow, sometimes green (from the mixed influence of 

 yellow and black pigment?). In a nest underlying a black 

 chromatophore remote from yellow pigment the bulk of the 

 reflecting matter, viewed from the inner surface, is steel-grey in 

 colour, portions, as already noted, being blue. 



Reflecting matter of a similar nature forms a more or less 

 continuous " argenteum " under the coloured pai'ts, the minute 

 elements being often, if not always, rod-like in shape. The skin 

 of a young male, in which no prismatic bodies have been developed, 

 can be cut with the microtome without the necessity of entirely 

 dissolving out the reflecting tissue. Here the black chromato- 

 phores of the innermost series are seen in section, sometimes 

 imbedded in a thickened depression of a continuous argenteum, 

 sometimes overlying masses of similar tissue detached from each 

 other. 



The skin of the white ventral surface of the abdomen in old or 

 young has a dense white argenteum and no chromatophores. The 

 argenteum may be resolved into minute rod-like particles, similar 

 to those obtained by rupturing the prismatic bodies of other parts, 

 and to those which form the much thinner argenteum of the 

 sides. The white effect appears to be due to the manner of their 

 arrangement, since, if traced out, the elements of all reflecting tissue 

 whatsoever seem to possess the same optical properties '. 



The masses of reflecting tissue underlying the black chromato- 

 phores (equally present in both sexes and at all sizes) can certainly 

 play no part in the colour-mechanism of the body, since they are 

 only refractive from the internal aspect. In the transparent 

 fin-membranes they may feebly contribute to the coloration. 

 Cunningham and MacMunn, who have noted that the iridocytes 

 of the riounder are closely embraced on the outer side by the 

 black chromatophores, offer no suggestion as to the function of 

 the former. It is difficult to see that they have any influence at 

 all, in such association, on the colour-effect. 



The results obtained by the various observers who have investi- 



* For the present I do not include as reflecting tissue the brown granular 

 matter, which I have shown to be probably a derivative of degenerate black 

 chromatophores. 



