NUDIBKANCHIATA OF THE L.M.li.C. DISTKICT. 161 



branchs. Our nudibranchs were all alive, healthy, and 

 good-sized specimens ; and our fish were probably the 

 right kind, being nearly all shore fishes, found in the im- 

 mediate neighbourhood of where the imdibranchs live. 

 But still the conditions were, of course, to a certain extent 

 artificial, and that must be taken into account in drawing 

 conclusions. Dropping the nudibranchs into the tank 

 from above is unnatural, and may give rise to a mislead- 

 ing result, especially where the fish are accustomed to 

 have their food thrown in from above, and onhj receive 

 edible food. 



Then again, at least some of the fish — those that have 

 been some time in captivity, have been educated to com- 

 pete with one another for the food masses. When anytliiiKj 

 is throw^n in— a bit of white shell will do — there is at 

 once a rush made upon the falling object, and no time is 

 allow^ed for inspection or consideration. I would account 

 for the seizing of Eolis by the shannies (very active, vora- 

 cious, and apparently impulsive fishes), even when the 

 prey is evidently distasteful and has brilliant warning 

 colours, as a result of this acquired habit of competition, 

 and of pouncing upon anything throw^n into the tank ; 

 several times when a morsel was suddenly bolted, it 

 seemed to be because another fish was coming up to seize 

 it. Still there is a marked difference between the man- 

 ner in which they take a cockle and, say, an Ancula. The 

 cockle is taken right in and swallowed at once, while the 

 distasteful nudibranch, even if seized, is usually only 

 partly taken into the mouth, in some cases it is seen to 

 be held by the very front of the jaws, and is then ejected 

 with force. 



Ancula has been a particularly interesting case. Start- 

 ing with the general opinion that Ancula is a perfectly 

 defenceless soft-bodied animal, we were astonished to find 



