946 SUMMAKY OF CURRENT RESEARCHES RELATING TO 



cylindrical cells. In the cells of these glands, but not in those of the 

 epithelium, mitotic division of the nucleus was observed. The rectum of 

 Eremobia consists of two divisions which are separated from one another 

 by a muscular valve ; in both, the epithelial layer is well developed, and 

 in that of the rectal glands there are mucous cells in addition to those of 

 the ordinary cylindrical form. The rectum of the larva of ^schna also 

 consists of two divisions, but these are not separated by any valve. Through 

 its whole extent the epithelial cells are of two kinds, some being large 

 with large nuclei, and the others small. The latter form compact folds, 

 and the former either lie close to the muscular wall or give rise to simple 

 folds widely separated from one another. In addition to its exterior gills, 

 the rectum of these larvae is provided with typical rectal glands. 



Protective Value of Colour and Markings in Insects.* — Mr, E. B. 

 Poulton has made a large series of experiments with the object of proving 

 the protective value of colour and markings in insects in reference to their 

 vertebrate enemies. He concludes that the extremely specialized defence 

 of the larval stage follows from its delicate anatomical construction and 

 the necessities which are imposed upon it as the great feeding stage. 

 Highly conspicuous insects nearly always possess some unpleasant attri- 

 bute, such as disagreeable taste or smell in the tissues and fluids of the 

 body, irritating hairs, or stings, but in a small number of cases a con- 

 spicuous appearance has not yet been shown to be attended by any 

 unpleasant attribute. In various species the same colours and patterns 

 are again and again repeated, so that the vertebrate enemies are only com- 

 pelled to learn a few types of appearance, and these types are of a kind 

 which such enemies most easily learn. Certain appearances are especially 

 impressed on them by highly aggressive insects, feared because of stings and 

 so on ; and hence, there is especial advantage in any approximation to such 

 types. In a relatively few cases aggressive forms among the Vertebrata 

 (serpents), are mimicked, though the insect itself is quite harmless. 



Insects which are protectively coloured not uncommonly assume, when 

 detected, a terrifying aspect, and in some cases take up offensive measures, 

 such as the discharge of irritating fluid. A few forms, which are probably 

 transitional, may be unconcealed, and yet not very conspicuous ; these may 

 possess unpleasant qualities, or may be eaten readily. As the likes and 

 dislikes of insect-eaters are purely relative, and as, if pressed with hunger, 

 they may eat the most disagreeable and highly conspicuous insects, we may 

 here find an explanation of the fact that only a relatively small number of 

 insects adopt such a means of defence. It seems probable that when one 

 vertebrate eats an unpleasant insect and another refuses it, the former has 

 conquered its prejudices, having originally disliked the insect. 



In the sexually mature forms warning colours can be distinguished 

 from sexual colours by their distribution on the surface of the body, by the 

 way in which they are displayed in flight, by their type of pattern, and by 

 the colours employed. The sexual colours or patterns are beautiful, the 

 others conspicuous. This conspicuous appearance has relation to the injury 

 which would be inflicted by the experimental " tasting " of certain enemies, 

 such as birds or lizards ; though enemies which, like frogs, inflict no injuries 

 in tasting, have, to a limited extent, taken advantage of the warning colours. 



Insects which evade their enemies by protective resemblance and attitude, 

 by rapid movements or habits of concealment, are generally palatable, but 

 they may possess an unpleasant taste or smell which may or may not protect 

 them from their enemies ; in a very small number of species the most perfect 



* Proc. Zool. Soc. Lond., 1887, pp. 191-274. 



