292 Mr. Poulton's notes in 1886 



legs. Thus the larva often presents the appearance of 

 a twig passing obliquely from one branch to the other, 

 and, although such a position seems likely to attract 

 attention because it represents something unnatural, 

 yet in reality the concealment is very perfect, for the 

 twigs of the food-plant (birch) are so extremely 

 numerous, and present such a complicated network to 

 the observer, that one such oblique twig-like appearance 

 readily escapes detection, and may often fall into the 

 line along which a real twig is prolonged. Very com- 

 monly the larvae are supported anteriorly by holding 

 the leaves or stem of the branch to which the claspers 

 cling, and in this case the concealment among the 

 interlacing twigs of the food-plant is even more perfect. 

 It is probable that the same mode of support will be 

 found to hold good in other species of Geometrce. The 

 attitude is shown in Plate X., fig. 4, a green variety of 

 A. betvlaria being figured on a twig of birch. 



y. The softening of the contact between a Geometer 

 larva and the twig on which it is resting. — In a note 

 to the translation of Weismann's 'Essay on the 

 Markings of Caterpillars,' p. 292, Prof. Meldola states 

 as follows : — " The adaptive resemblance is considerably 

 enhanced in Catocala and in Lasiocampa quercifolia by 

 the row of fleshy protuberances along the sides of these 

 caterpillars, which enables them to rest on the tree- 

 trunks by day without casting a sharp shadow. The 

 hairs along the sides of the caterpillar of Pcecilocampa 

 populi doubtless serve the same purpose." This expla- 

 nation, which had been previously given by Professor 

 Meldola, is accepted by Sir John Lubbock (Trans. Ent. 

 Soc. Lond., 1878, p. 242) and by Mr. Peter Cameron 

 (ibid., 1880, p. 75), and the latter writer extends the 

 explanation to the hairy larvae of certain phytophagous 

 Hymenoptera. I can now bring forward a confirmatory 

 observation which supports the explanation offered by 

 Prof. Meldola in the strongest manner. The larvae of 

 Geometrce, in the typical attitude of protective re- 

 semblance to a twig, only touch their food-plant at and 

 between the two posterior pairs of claspers, and this 

 part of the larva, in relation to the food-plant, of course 

 represents the point at which a twig is united with the 

 branch immediately below the divergence. At such a 

 point the bark of twig and branch are continuous, and 



