428 Mr. G. A. K. Marshall on 



character which is non-acra^iforni and purely aposematic 

 is far inore emphasized than in the latter species. It is 

 probable that scsarnus I'epresents a later development, and 

 that in it the synaposematic elements have been gradually 

 strengthened and the peculiar aposematic character cor- 

 respondingly reduced. 



We now pass to the con.sideration of a species in which 

 the conspicuous characters of the under-side of the wet 

 phase are probably entirely peculinr and aposematic. 



Mr. Marshall's suggestion in 1896 that Precis jje/asgis is 

 the wet phase of P. archesia has never been confirmed by 

 breeding the one form from the other. It is, however, 

 certain that his conclusion was sound. Tlie two forms 

 have often been captured in coitu. The female i^e/rts^w 

 represented in Plate XII, fig. 4, was captured by Mr. 

 Marshall in coitu with the male circhesia shown in Fig. 5 of 

 the same plate. Intermediate forms are much commoner 

 than in the case of sesamus and antilopr; and above all the 

 relationship of wet phase to dry is fir closer in cd'chcsia, so 

 that it is possible to see how the one was derived from 

 the other more fully than in any of the species with mark- 

 edly-different seasonal forms. The under-.side of one of the 

 commonest forms of the dry phase is represented in Fig. 6, 

 Plato XIII, and opposite to it that of the typical wet phase 

 in Fig. 5. At first sight they appear totally different, and 

 certainly the latter is as conspicuous as the former is well 

 concealed. An uncoloured illustration cannot do justice 

 to the varied shades of brown and grey on the under-side 

 of archesict {Fig. G), and a long scries of specimens would be 

 required to show the immense range of individual variation 

 by which all kinds of common appearances presented by dead 

 leaves are reproduced. Among such variations is one in 

 which the dark-brown ground-colour is almost uniform 

 and unmottled inside the mid-rib-like stripe (Fig. 7). From 

 this we pass to forms in which the stripe widens into a 

 light band (Fig. 8), clearly showing its homology with the 

 still more conspicuous band of ^^f/as^/is (compare Figs. 5 and 

 8). Such a variety as that shown in Fig. 8 is still a long way 

 on the arclicfiia side of a form intermediate between the 

 wet and dry phase, and would certainly be cryptic rather 

 than conspicuous in nature, although not so well concealed 

 as the form shown in Fig. 7, and still less so than that 

 shown in Fig. G. Truly intermediate varieties between 

 the wet and dry phases are not uncommon, in which 



