1898.] BBEEDING OF THE DEAGOTTET. 307 



were dropped into the large table-tank. Several wrasse, Labrus 

 and Crenilahriis, darted at these objects, but either retreated without 

 touching them or dropped them a.s soon as they were seized, and 

 departed in evident disgust. Gobies, chiefly G. paganellus, inves- 

 tigated the matter and seized the coloured balls, but mostly 

 dropped them very shortly. One carried off a ball to a shelter in 

 the middle of the tank, but then dropped it. Only one goby 

 made any attempt to masticate a ball, and that was soon abandoned. 

 The young Callionijmi, on the contrary, took the balls greedily 

 and chewed them. The same fish would take a ball, chew it for 

 some time, reject it and seize it once more. Certainly the yellow 

 matter was not distasteful, nor was the colour terrifying. Frag- 

 ments of the second dorsal were treated in the same way as 

 the balls by wrasse and gobies, but greedily attacked by young 

 Callionymi ; the fin-membrane was swallowed, when separated 

 by repeated chewing from the rays. A quantity of the solution 

 was poured over the assembled Callionymi, who took no apparent 

 notice of it. A wrasse saw the yellow colour and darted out of 

 its hiding-place, but rapidly retreated on reaching the foreign 

 matter. It seems, therefore, that terror is inspired in young 

 Dragonets by the menacing gestures of the courting male, and not 

 by the optical or olfactory properties of the yellow pigment. Of 

 course I am not contending that the yellow colour is actually 

 attractive to young Dragonets, since most of the fish, of whatever 

 species, in the table-tank are so far tame that they will come and 

 look at anything that is offered to them. It appears to me im- 

 possible to decide in what manner the elements of the coloration of 

 the male influence the female. If it appeals to her sense of scent, 

 the yellow element only can be concerned, since the blue results 

 from a combination of two elements which are not soluble. It 

 certainly appears most probable that the whole coloration-effect 

 merely renders the large dorsal fins more conspicuous and so 

 advertises the whereabouts of the male. As we have seen, the 

 dull-coloured female does not appear to be readily perceived by 

 the male even at the distance of a few yards. Her presence may 

 presumably be indicated by some odoriferous product of the genital 

 organs at the season of ripeness. 



Apart from the sexual question, we have seen that the yellow 

 colouring-matter is distasteful to gobies and wrasse. The latter, 

 however, are not to my knowledge fish -eaters. Grobies are more 

 or less indiscriminate in their appetite. Except dog-fish, rays, 

 and conger, which seldom feed in the daytime, the only large 

 fish-eating forms in the aquarium are pollack and turbot. Both 

 these species must be present on the Callionymus ground during 

 the breeding-season. The pollack feeds very largely on fish. 

 Those in our tanks appear hardly large enough to take a full- 

 grown Dragonet, so my experiments have been made with pieces 

 cut from the sides of large males in various stages of colour, large 

 females treated in the same way, and small living undifferentiated 

 specimens. The larger pollack often took the bits of Dragonet, 



20* 



