COLOURS OF WATER-MITES. 1235 



All Water-mites possess large skin-glands, provided with 

 so-called "sensory" hairs, whose function has been, so far, 

 uncertain. As Piersig (3) points out, it has long ago been 

 suggested that these glands secrete some unpleasant fluid ; but 

 he states that there is not enough evidence to decide whether 

 the mites are distasteful or not. That they are so to Stickle- 

 backs is certain, in view of the experiments here described. The 

 idea that it is the skin-glands which make the unpleasant taste, 

 is supported by the fact that the fish does not injure a mite 

 when it takes it into its mouth, and therefore cannot actually 

 taste its flesh ; also by the fact that the mite is spat out so 

 promptly. The anatomy of these glands and the nature of the 

 supposed secretion require to be worked out in detail. The 

 hardness of the body probably helps to render some mites 

 distasteful, as indicated by the fact that a Stickleback when not 

 very hungry will sometimes reject a young Aselhts, after taking 

 it into its mouth. But this could not apply to such soft-skinned 

 mites as Eylais and Diplodontus. 



These experiments make it clear that Sticklebacks recognize 

 the scarlet mites by their appearance. It is extremely likely 

 that it is the colour which is remembered, although no accurate 

 work has been done on G aster osteus, in order to prove this. 

 But White (8) has recently shown that the American Stickle- 

 back {Eucalia inconstans) and Mud Minnow (Umbra limbi) are 

 well able to distinguish difierences in colour as distinct from 

 diflferences in the intensity of light. She proved by expei'iment 

 that the fish were capable of associating different colours with 

 their food, though they were unable to distinguish between fine 

 shades of the same colour. The same may well be the case for 

 the 3-spined Stickleback. 



There is another point which is of some importance. It may 

 be asked what advantage is gained by mites, seeing that they 

 appear to be uninjured by fish, even after being snapped up and 

 spat out several times. The answer to this is that: (1) Mites 

 miust be actually desti'oyed by some animals before their dis- 

 tastefulness can be realized, e. g. Agahus ; (2) the fluid which 

 seems to be secreted by the skin-glands would ultimately become 

 exhausted ; (3) being continually attacked must interfere with 

 the life pi'ocesses of the mite. The degree of efl:ectiveness of 

 warning colours depends among other things on the state of 

 hunger of the enemies. 



4. The Evidence for Mullerian Mimicry. 



There is no evidence at present of Batesian mimicry, i. e. of 

 edible mimics, among Water-mites. It is unlikely that it exists, 

 in view of the facts that all Water-mites have the large skin- 

 glands and that there are very few cases on record of mites being 

 edible. However, the possibility should not be ignored. There 

 is, on the other hand, strong evidence of Miillerian mimicry 



