] 898.] BREEDING OP THE DftAGONEl. 295 



the breeding-season of the Dragonets, I endeavoured, with the 

 assistance of my friend Mr. F. Gover, to observe the behaviour 

 of large males when dropped into the presence of the turbot. 

 However, the former invariabl}^ managed to reach a dark corner 

 before the turbot became aware of their presence, and we never 

 saw them again. They may be there still, since there are many 

 small fish, wrasses of several species, in the recesses of the tank, 

 which do not show themselves at the front once in a month. On 

 another occasion a similar experiment, with more fortunate result, 

 was made by Mr. E. J. Allen, Director of the Laboratory, and 

 myself. "We succeeded in making a large male Dragonet, brilliantly 

 coloured, swim over the part of the tank frequented by the turbot. 

 One of the latter started in pursuit, and the Dragonet bolted at 

 full speed, with dorsal fins depressed, but was caught and engulfed. 

 No sort of effort was made to display the colours. Here the 

 Dragonet was on strange ground and the turbot at home. 



After pairing had ceased and the colours of the male Dragonets 

 had greatly faded I placed a large turbot in their tank. As the 

 big fish descended to the bottom the Dragonets, large and small, 

 darted wildly away, and some buried themselves in the gravel. 

 The turbot appeared only concerned to get out of the tank, which 

 is much better lighted than the one in which it has spent the last 

 few years. It paid no attention to the Dragonets, but kept 

 swimming backwards and forwards along the bottom, generally 

 close to the glass. Whenever it approached a large male Dragonet 

 the latter would put up his dorsals and dart out of the way, not 

 very rapidly. Once the turbot came to rest opposite a corner in 

 which was a large male Dragonet, who erected his dorsal fins and 

 slowly glided past the intruder. As long as the turbot was left in 

 the tank the Dragonets seemed uneasy and kept moving about, but 

 they did not display their fins unless the turbot was quite near 

 them. One got out of harm's way by clinging to the side of the 

 tank by the pelvic fins, which, as is well known, are capable of 

 acting as a sucker (as in Gobius), though not very often used in 

 this way. Here the Dragonets were at home, and were not 

 actually attacked by the turbot, whose attention was distracted by 

 his unwonted surroundings. I think it is clear that the male 

 Dragonet does display his fins to intimidate a possible enemy, but 

 it is impossible to say to what extent he relies on the efficacy of 

 the exhibition should the enemy actually attack him. As a matter 

 of fact, fully-developed male Dragonets are, under natural con- 

 ditions, frequent victims to predaceous fishes ; but this question 

 may be discussed more conveniently at a later stage. 



Small Dragonets, whether male or female, have also been 

 observed to erect one or both dorsal fins at the approach of a 

 large fish or a net, always provided that they have been in the 

 tank for some time. While catching specimens in the various 

 table-tanks in which they have been kept, I have often noticed 

 that small Dragonets wUl hoist the first dorsal fin at the approach 

 of the net ; sometimes they remain still for a time, waving the 



