300 ME. E. W. L. HOLT ON THE [Apr. 19, 



They are then seen to be stellate in character, but much smaller 

 than the black ones. Cunningham and MacMunn, who record an 

 approach to this condition ofdiifusion {op. cit. p. 77*3), provisionally 

 suggest that the coloured pigment in the flounder and plaice 

 diffuses from the connective-tissue cells in which it is deposited '. 



In tlie case of Callionymus, comparing the young and adult 

 conditions, no other conclusion seems possible. A complete net- 

 work of granular matter certainly represents the black chromato- 

 phores present in the younger stages. 



The white bands and their transparent margins have become 

 blue, with a border of grey, but only black chromatophores are 

 present. The latter are abundant in the grey area, aud are here of 

 a dendritic nature. In the blue part I find them less numerous 

 aud, in microscopic preparations, much less expanded, the radii 

 being very short and the centre very dense. In both cases they 

 are frequently, if not always, associated with underlying masses of 

 a granular matter. The brilliant blue colour is derived from a 

 dense network of bundles of small, somewhat bean-shaped bodies. 

 The latter are yellow by transmitted light, but intensely blue by 

 reflected light over a black ground, such as is afforded by a black 

 chromatophore. They occur immediately below the epidermis. 

 It is difhcult to isolate them, as in the process of teasing out they 

 are readily ruptured and resolve themselves into minute rod-like 

 crystals. They appear to correspond to the iridocytes of Pouchet 

 and of Cunningham and MacMuiui, but are very minute, and are 

 certainly not associated with the chromatophores in the same 

 manner as is described by the last-named authors in the case of 

 the flounder. I have not been able to detect a circular aperture, 

 and have entirely failed to obtain sections. It will be perhaps 

 more convenient to term them, provisionally, "jsrwwaiic bodies" 

 instead of iridocytes. 



They may probably prove to be identical with some reflecting 

 substance found in small quantities by Cunningham and MacMunn 

 {op. cit. p. 773), in Siphonostoma, but not, apparently, in other 

 forms examined by those observers. They appear to be repre- 

 sented in Gohius, a form closely allied to Callionijmus, and are 

 there termed by Heincke " chromatophores filled with small discs 

 of a metallic lustre." These " chromatophores " were supposed 

 to contain pigment, but Heincke, who examined them only in 

 the living fish, acknowledges that the pigment may have been 

 really external. I find that the blue colour of the dorsal fin 

 in the male G. minutus is identical in mechanism with that of 

 Callioni/mus. 



We have seen that these bodies are yellow by transmitted light. 

 The rod-like particles into which they may be broken up are also 

 yellow, but, though highly iridescent by reflected light, their 

 iridescence is usually yellow, sometimes green, and only rarely 



' The yellow pigment of Carassius was usually met with by the same observers 

 in diffuse condition. 



