COLOURS OF WATER-MITES. 1237 



it ill its mouth and spat it out twice, and afterwards avoided it. 

 Hydraraclma paludosa (brilliant scarlet) was now introduced, 

 and the fish avoided it after one trial. There were also some 

 (hjpris virens present, coloui'ed a uniform brownish green. This 

 Ostracod was too hard for the fish to crush and too big to 

 swallow; yet the Stickleback kept rushing at tlae C?//5ri's when- 

 ever they moved, trying to swallow them. This it continued 

 to do throughout several days. 



This experiment suggests that scarlet was remembered by the 

 Stickleback better than yellow and black, although the hunger 

 of the fish has to be taken into account. But both were more 

 effective than the dull uniform colour of the Cyjyris. 



Expt. 11. The same fish was used as in Expt. 1. It was 

 experienced as regards scarlet mites. A Bajjhnia was eaten 

 eagerly. An Acercus lutescens Hei-m. was put in. This is a 

 small mite, coloured pale pink, dark brown, and yelloAvish white. 

 The fish snapped njj the mite, but immediately spat it out again. 

 This was repeated three times. After this the mite was avoided. 

 A short time afterwards the fish noticed it again, and the whole 

 procedure was repeated {the same number of times). A Daphnia 

 was eaten. Eylais hamata (scarlet) was introduced, but avoided 

 after inspection. Both mites were removed. Next morning the 

 experiment was repeated. The fish saw Acercus and snapped 

 it up, and then spat it out. Next time the mite was avoided. 

 The mite was then removed. Tiiirty-five minutes later Acercus 

 was put in again. The fish went right up to the mite, but did 

 not touch it. After this the mite was ignored. Baplinia was 

 put in and eaten. Eylais was put in and avoided. This experi- 

 ment suggests that the fish was able to remember the scarlet 

 much better than the other colours. It is unlikely that the 

 difference in impressions made on the fish was due to the scarlet 

 mite having a more unpleasant taste than the other, since both 

 were spat out equally promptly. It is probable, then, that 

 scarlet is the most efficient warning colour for use in Stickleback- 

 haunted habitats. Perhaps it is only the best for such habitats 

 at one time of year, i. e. in the breeding-season of the fish. 

 I observed yellow-and-black mites (unidentified) in a Hampshire 

 pond where 3-spined Sticklebacks occurred on August 4th, 1922. 

 This shows that such mites can exist in some Stickleback ponds. 

 Not much is known about the food reactions of these fish, except 

 that they are usually very fierce and voracious. It is necessary 

 to be very cautious in applying the knowledge about food habits 

 of Sticklebacks in one place to those of another. J. T. Saunders 

 (5) records a pond in which the adult 3-spined Sticklebacks ate 

 nothing but one kind of diatom, while the young ones in the same 

 pond and all the fish of neighbouring ponds were carnivorous. 

 He also states that the fish soon learn to distinguish by sight 

 between different animal foods, and that they are very sensitive 

 to changes in their environment — e.r/., the diatom-eaters became 

 carnivorous in captivity. The latter fact might throw some doubt 



Proc. Zool. Soc— 1922, No. LXXXII. 82 



