302 MB. E. W. li. HOLT ON THE [^pr- 19> 



evidently due to the comparative scarcity of black pigment. I have 

 never seen the red tinge shown in Smitt's figure (Hist. Scand. 

 Pish. ed. 2, pi. xiv.), but it is a fact that prismatic bodies, when 

 seen by transmitted light in dense masses, tend to exchange 

 the ordinary pale yellow for a warmer tint. In the hinder part of 

 the blue margin of the axillary ocellus is an intensely brilliant 

 region. Here it will be found that some of the prismatic bodies 

 have a crimson colour in perfectly fresh preparations. Individu- 

 ally they are violet by reflected light over a black surface. The 

 action of glycerine rapidly reduces them to the usual pale yellow 

 tint, and I cannot say by what cause the crimson colour is 

 produced. 



With regard to the coloration of the body, the differentiation is 

 achieved in a manner parallel to that which obtains in the dorsal 

 fius. We need therefore only discuss the colour-mechanism of 

 the adult male. The skin of the body differs from that which 

 constitutes the fin-membrane in that the chorion is thick and 

 tough and is associated internally with a further layer of loose 

 connective tissue. In CaJlioni/mus, as was noted in the case of 

 other Teleosteans by Cunningham and MacMunn, this inner 

 layer frequently adheres to the muscles when the skin is stripped 

 off. The epidermis, especially in mercuric chloride or old alcohol 

 preparations, can readily be isolated. It contains neither chromato- 

 phores nor reflective tissue in any part which I have examined. 

 In the blue bands of the side externally to the chorion are found 

 black chromatophores in variable number, but often abundant. 

 In the same plane and to some extent superficial to these occurs 

 a diffuse layer of prismatic bodies, similar to those of the second 

 dorsal, but showing less tendency to a retiform arrangement, 

 owing to the closerapproximation of the bundles. The internal layer 

 of connective tissue is very rich in strands of prismatic bodies, 

 overlying and passing between numerous black chromatophores. 

 In preparations I find many of the latter completely contracted. 

 It has been noted that these bands are subject to momentary 

 intensification of colour, and it may reasonably be supposed that 

 such intensification is effected by expansion of the chromatophores 

 in response to nerve stimuli, causing a greater surface of black 

 pigment to be interrupted by the overlying prismatic bodies. On the 

 pelvic fin (cf. Plate XXVI.) are certain streaks of blue, which, 

 during the breeding-season, remain of a constant deep hue. Here 

 it is found that prismatic bodies overlie layers of black chromato- 

 phores so closely set as to present a practically continuous surface, 

 a condition which sufficiently explains the constant character of the 

 colour K 



^ Heincke found that (as I can confirm) iu Gobii/s rufhcnsparri the metallic 

 lustre is brought about by the crowding together of the " chromatophores " 

 containing the glittering substance. CaUionymus is too large to be conveniently 

 studied under the microscope in life. Although the bundles of prismatic bodies 

 show a post-mortem tendency to contract, it appears to me that the background 

 is more influential in colour-production than the an-angemeut of the bodies. 



