272 BRITISH LEPIDOPTERA. 



Phryxids. In the Amorphid imago, they are especially weak, and, at first 

 sight, appear to be wanting, but, in reality, they are scattered over 

 the whole dorsum of the segment and are but little more numerous 

 along the hind margin, and they have scale-cups for attachment, about 

 twice the size of those of the ordinary scales (Anwrpha popnli). In Mimas 

 tiliae they also exist over the whole segment but are here very little 

 differentiated trom scales, differing chiefly in being narrow and over- 

 loaded with pigment and often with bifid extremities, and it would almost 

 be more correct to say that they are wanting, but that the .scales are 

 somewhat modified. They are frequent also in Noctuids. In the Geo- 

 metrid, Nyssia hispidaria, they are developed in two forms, one of which 

 occurs all over the dorsum. Treating this as a structure characteristic 

 of the Sphingides, as being found in the whole superfamily, and highly 

 developed in many genera and families, it would seem to add another 

 to the many characters otherwise noted, that show the Amorphid 

 imago (and pupa) to be the most primitive in the group, although 

 the larval specialisation has made more advancement than in the 

 other families, perhaps actually, but certainly in having the earliest 

 stages more modified by the effects of later features passed back 

 to the first instar than in Eumorphids or even Sphingids (sens. rest?-?). 



We know of no real attempt since Hiibner's time (antea, vol. 

 hi., pp. 351-352) to classify the species included in the Sphinginae, and 

 most authors seem to have surmounted the difficulty by placing 

 as many heterogeneous species as possible in the same genus, 

 whilst our British authors, e.g., Stainton, Meyrick, &c, have dealt 

 with the matter in the simplest possible manner, viz., by placing 

 them all in the same genus and neglecting altogether any detailed 

 consideration of the structure of the species. The most recent 

 German work (Bartel, Pal. Gross-Schmett., ii., p. 37) deals with 

 the group in the same manner, and Bartel writes : "In 

 Europe, Sphinx is only represented by 3 species, of which .S. 

 convolvuli is very widely distributed, S. ligustri occurs also in 

 Siberia and japan, and only .S. pinastri is confined to Europe 

 proper." He then adds : "The genus is very numerous in the 

 tropics. Herrich-Schaffer united the genera Sphinx and DeilepJiila, 

 but Deilepliila is distinguished, not only by its different larvae, 

 and more especially by its pupae, but also by the neuration, 

 nervure 8 of the forewings running into the apex, while this nervine 

 in Sphinx runs into the outer margin just below the apex (!). Later 

 systematists have rightly separated the genera* again." He then 

 gives the-following "table of the Palaearctic species" that he includes 

 in Sphinx : 



I. Abdomen with red spots on the sides. Hindwings with black transverse bands. 



1. Fringes chequered; hindwings blue-grey- comiolvuli, L. 



2. Fringes nnicolorous ; hindwings light rose-colour —ligustri, L. 



II. Abdomen with black spots on the sides. Hindwings without dark bands. 



1. Hindwings nnicolorous— pinastri, L. 



2. Hindwings before the inner angle with a white longitudinal ray, followed 



bv a short, uninterrupted light transverse stripe — increta, Walk. 



We would now refer our readers to Hiibner's grouping of the 

 true Sphingids (antca, hi., pp. 35i~35 2 )> which he included in his 



* Bartel treats the whole of our Sphinginae as one genus and the Eumorphinae 

 as another. 



