PHRYXIDI. 141 



separated. This view becomes almost a certainty when we find 

 that really the only argument of any weight in support of the 

 separate origin of the Phryxid markings is based on the view that 

 Turneria hippophaes is the lowest of the Phryxids, and that the 

 eye-spots are there appearing, whilst there can be little doubt 

 that it is really one of the last evolved, and is eminently specialised 

 to suit its surroundings. The Eumorphid larvae, which Bacot 

 asserts are the primitive Sphingids, like those of Hemarids and 

 Sesiids, are feeders on low-growing plants, and this is especially 

 true of the Phryxids, of which, however, T. hippophaes is an ex- 

 ception, and may be supposed to have taken to a shrub comparatively 

 recently ; its imaginal colouring is also quite specialised (possibly 

 for hiding on sand or among the leaves of its foodplant), so that 

 its colouring is less like that of Eumorphids, and of Sphingids 

 generally, than almost any other species of the group. It is very 

 safe to assume, then (a suggestion that did not escape Weismann, 

 but which he rejected), that hippophaes is derived from a form like 

 euphorbiae, and not vice versa. This conclusion is supported by 

 the pupal evidence, which shows hippophaes to be, if anything, a 

 shade in advance of euphorbiae instead of very decidedly below it, 

 as, if Weismann's theory were correct, it ought to be. This brings 

 us to what appears to have been the real line of evolution of these 

 eye-spots, viz., that they originated in a tendency of the pale line 

 to break up into eye-spots all along the abdominal segments, and 

 possibly more pronounced at one place than another, but certainly 

 not reaching any such development as we find in Theretra (porcellus), 

 highly evolved on one segment before appearing on another. 

 Their first development was probably encouraged by a gain 

 being made towards inconspicuousness by breaking up the large uni- 

 form areas ; then the division into two groups occurred, as 

 these spots, so soon as they reached any definite development, 

 presented two new but very different kinds of usefulness. Of these, 

 the Phryxids are in tolerably uniform series, both larvally and 

 pupally, and present few difficulties, except such as that of 

 hippophaes, and, perhaps, a few others, if we had a fuller knowledge 

 of typical forms. They did not subdivide into several distinct 

 branches of any length. They do, indeed, rather trouble one 

 to decide precisely which are the higher and which the lower, 

 rather, however, from their being so close together than on account 

 of their presenting puzzling and antagonistic characters, and, 

 zygophylli, hippophaes, vespertilio and nicaea would probably be 

 terminal species of different divisions, whilst pupally, livor- 

 nica is certainly the highest and most elaborated. Except 

 for the position assigned to hippophaes this is substantially the 

 same phylogenesis as is given by Weismann (loc. cit., p. 358), 

 though this one change makes it look very different, as perhaps 

 essentially it is. 



We return now to the Eumorphidi, which contains of British 

 species — elpenor and porcellus (practically identical for broader 

 phylogenetic questions) and celerio. This is a tribe really much 

 more numerous in tropical regions than the Phryxidi ; unfortunately, 

 for this reason, we know less about it. Weismann recognised 

 two divisions beyond the basal Darapsa : (1) With eye-spots on 



