95 



mally dark. They were always within sight of each other, 

 and generally in some sheltered place. On other occasions 

 what appeared to be a combat was witnessed between two fish, 

 which had assumed this same dark shape, but this always 

 happened out in the open on the floor of the tank. The two 

 iish circled round, finally settling down in a position 

 facing each other, about a foot or a foot and a half apart. 

 The affair seemed to consist more of threatened than actual 

 infliction of injury. Their action resembled that of two cocks 

 fighting, and consisted of a repeated dipping down of the head, 

 and the general assumption of a threatening attitude. They, 

 however, seldom came into actual conflict and, even when they 

 made a rush at each other, generally sheared off, apparently 

 before the contact. The end of such a contest, if contest it 

 was, was never witnessed, and a few days afterwards they were 

 seen swimming about amicably with the rest of the group as 

 before, having lost their dark hues. 



Another observation made on these fish is of interest as the 

 only recorded instance, so far as I am aware, of fish eating 

 Alc^^onarians. On one occasion an Alcyonarian {Cavernularia) 

 was introduced into the tank. It is a large fleshy form about 

 a foot long, with a basal part, with which it slowly dug into the 

 sand. After it had thus assumed an erect position, the polyps 

 expanded fully to about half an inch. One of the Steenjes, 

 after a careful survey, made a sudden dart at an extended 

 polyp, seized it and, with a lateral jerk, tore it off. The fish, 

 after a prolonged munching of its prey, swallowed it and came 

 back for another polyp. Curiously enough this rough usage 

 had no effect on the other polyps, which remained expanded 

 as before. The whole of the polyps were ultimately removed 

 in this way. This occurrence does not prove, however, that 

 this fish is in the habit of eating Alcyonarians, and it is to be 

 noted (i) that only one fish of the group acted in this manner, 

 and this particular fish was one which was less able to hold 

 its own, often swimming by itself, and generally did not get 

 a fair share of the usual food ; (2) the fish, after seizing the 

 polyp, apparently had some difficulty and hesitation in swal- 

 lowing it, probably owing to the presence of the spicules ; (3) 

 the colony of polyps made no attempt to protect themselves 

 by retracting though they readily did so when disturbed other- 

 wise ; (4) the Cavernularia was from deeper water than that 

 frequented by this species of fish, which did not attempt to 

 attack any of the other Alcyonarians in the tank, e.g., Gorgonia. 

 The observation is, perhaps, not of so much importance as 

 proving that fish will eat Alcyonarians, but as indicating the 

 reason why thev do not habitually do so, in spite of the ap- 



[C.P. 6— '14. 1 



