1898.] BREEDING OF THE DEAGONET. 311 



feeder, taking even such unpalatable organisms as Alci/oniam 

 and Actinice ^, so that it is not surprising to find that the various 

 protective devices of Callionymus are frequently inefficacious. 

 Whether or not the dorsal filament of LopMus is attractive to 

 Calliom/mus I cannot say, but the male Dragonet, when courting, 

 rushes heedlessly against anything that may be in the way, even 

 against fully-expanded anemones, Adamsia rondeletii. LopTiius 

 appears to snap at precisely the spot where anything touches the 

 erect filament ", and, as a matter of fact, the Dragonets among the 

 stomach-refuse on the fish-quay are mostly large males. Judging 

 from the very varied assortment of things that have been found 

 in the stomachs of Lopliius, it may be presumed that its sense 

 of taste is not very discriminating. I have not found Dragonets 

 in the stomachs of John Dories {Zeus faber), but have seen a small 

 specimen of the former taken and instantly rejected by a young 

 dory. This fish does not appear willing to take anything from 

 the bottom, though it will sometimes do so. 



It is possible that prawns (^Palcemon serratus) find something 

 distasteful in the skin of a large male Dragonet. On two occasions 

 I have noticed that a dead specimen placed in the table-tank was 

 unmolested, though the prawns in the same tank will seize small 

 individuals even before they are dead. A large male, which 

 died in the aquarium tank during the breeding-season, was not 

 injured by the crabs {Cancer and Carcinus) and hermits (Eujpagurus 

 bernhardus) for some time. The viscera and part of the muscles 

 of a mature female were eaten, while her skin remained practically 

 untouched ; but I have seen a fully-coloured male chased by a 

 Carcinus. 



If the yellow pigment of the male is really obnoxious to any 

 predaceous fish, it is evident that the female must also profit 

 thereby at the moment when she is most exposed to danger, viz. 

 when preoccupied in the matrimonial ascent. 



VII. General Considerations. 



The observations which I have collected above are certainly not 

 so complete as they might be, but I do not think that further 

 investigations will reveal many new facts in such part ol: the 

 bionomics of the Dragonet as are intimately connected with the 

 interpretation of the sexual dimorphism, i'urther study of the 

 palatability of this fish, from the point of taste of the predaceous 

 forms which it runs the risk of encountering under natural 

 conditions, is certainly desirable, and will be carried out whenever 

 opportunity permits. 



In the meanwhile we know that the Dragonet is a species in 

 which the male assumes, at a period roughly cori-esponding to the 

 inception of sexual maturity, a differentiation of structure which 



1 



Thomas Edward, ' Naturalist,' 1855. 



2 Cf. Holt, Sci. Proc. K. D. S., n. s., vii. 1892, p. 456. 



