1238 MR. C, S. ELTON ON THE 



on tlie vaiidity of the experiments with mites etc. described 

 l)ere; but the uniformity of the results makes it pretty certain 

 that the fish would reject Water-mites' in nature as well as in 

 captivity. 



The fact that the fish in Expts. 10 and 11 did not learn well 

 on other coloui'S than scarlet does not weaken the case in favour 

 of warning coloration, since the non-red mites used Avere not in 

 the presence of their natural enemies. Acercios lutescens came 

 from a pond in which there were no fish. Hygrohates longipalpis 

 and Limnesia undulata Avere from the Thames. Now, although 

 Sticklebacks are abundant in most rivers. Maxwell (2) says : 

 " It has been observed that the Stickleback is very scarce in the 

 Thames." 



The chief importance of these non-red mites here is that it has 

 been shown that they might be at a disadvantage in Stickleback- 

 haunted habitats (where scarlet ones are successful), and that 

 non-red mites actually tend not to come in such places. 



Some other evidence may be added to this, though very little 

 is known about the associatious of Water-mites. 



There is a flood-pond near Railway Ditch which was full of 

 3-spined Sticklebacks in May 1921 (see Expt. 8). The very large 

 and conspicuous scarlet Eylais extendens Miill. was taken here 

 swimming about among the fish. The only other mite was 

 Fiona conglohata Koch, which is veiy minute and is incon- 

 spicuously coloured. Hyphrydrus ovatus was unusually abundant 

 in Railway Ditch in May 1921. This beetle is coloured bright 

 orange below. It secretes a liquid when handled, which smells 

 like honey and may make it distasteful. It is possible that the 

 Mullerian mimicry included this beetle. Shelford (6) records 

 Limnochares aquatica, which is a large scarlet mite and is unable 

 to swim, as occurring abundantly in a pioneer association of 

 Ghara, in which were also many species of fish. Mr. Soar tells 

 me that all the crawling Water-mites are red, which supports the 

 idea that they are warningly coloured. 



5, Discussion. 



Scarlet is rather rare among animals, except in the tropics. 

 Higgins (1) has suggested that it is due to the fact that red rays 

 are more valuable (there being more energy per unit wave-length 

 than in other colours of light), and are therefore usually absorbed, 

 not reflected. It is more likely that the rareness of scarlet is due 

 to its conspicuousness when seen against natural browns and 

 greens (as in our own road-traffic warning signs). In any case, 

 we generally find that animals are scarlet only when thei-e is 

 some very important use for the colour. For the reason given 

 above, it makes an excellent warning colour. It is possible that 

 red can be seen from further away in water than other colours. 

 This is known to be the case in air. 



Any protection gained from wanaing coloration and Miillerian 



