378 BRITISH LEPIDOPTERA. 



are closely followed in those of Phalera bucephala and of other 

 species allied to it; the terminal segments are much like those of the 

 Amorphids ; it possesses the marginal batons, and the arrangement 

 of the hairs is very similar ; the keel is not so distinct, and does 

 not so definitely divide the hairs of the two sides as it does in 

 most Sphingids, and the ventral projection carrying the hairs, though 

 very similar to that of Sphinges, does not meet, its neighbour, but 

 is obviously distinct, and so does not look as in Sphingids, as if a part 

 of the true shaft of the antenna. The antenna of ^Egeriids (Sesiids) 

 is also very like that of Sphingids, much more like, in fact, 

 than is that of Phalera. Nevertheless it is essentially very differ- 

 ent, whereas that of Phalera is based on a real identity of structure. In 

 the JEgeriids the general form of the antenna is the same, though rather 

 more clubbed, and it ends in a very similar tassel, and the arrangement 

 of hairs is much the same. The antenna of the y^geriids is, how- 

 ever, solid ; it has no added ventral portion, the arrangement of hairs 

 is a little different, they also fail completely over the club, which in 

 both sexes carries only the fine appressed sensory hairs. A still 

 more fundamental difference, showing the ^geriid antenna not yet 

 to have escaped from a low Tineid form, is that in the females of 

 many, if not of all, species, one row of scales completely encircles the 

 antenna. The ^Egeriid antenna is, therefore, an early micro form in 

 its essential characters, although very specialised, and specialised 

 so largely on similar developments to those of Sphingids, and probably 

 in view of similar requirements, that the association of the two 

 families in classification is a result that very naturally followed. The 

 scaleless wings of some Sesiids (Macroglossids) appeared to be so 

 obviously a confirmatory character, that one understands how the 

 association of these two families is still a matter of faith in many 

 quarters. The differences between the antennae of the different 

 families of the Sphingids are not too well defined, that is, the forms 

 that are commonest in each group are not always very well marked 

 in all its members. The differences consist in the general straight- 

 ness of the antenna, in its being more or less clubbed, and in the 

 curvature and specialisation of its terminal segments. Thus — 



The antenna of the Amorphids differs from those of the remainder of the 

 Sphingids in having but little curvature at the tip, and a fairly round blunt un- 

 specialised last segment. 



That of the Ambulycids (Ambulyx) is a longer, more slender antenna, 

 with a gradually diminishing curved tip ending in a last segment carrying very 

 long scales, but not itself elongated. [In all the other groups the last segment 

 is specialised by considerable elongation and a clothing of bristles.] 



In Sesiids (Macroglossids) the antenna is straight, very thickly clubbed, and has 

 the terminal hook consisting of half-a-dozen very short compressed segments, the last 

 one elongated so as to be as long as two or three of these, but still very short, as 

 compared with the following groups. [This is probably not a stage intermediate 

 between Smerinthus and Spliin.x, but a specialisation for day-flying as seen in 

 Anthroccrids, Papilionids.] 



In the Eumorphid antenna the terminal hook is long and sweeping, and 

 involves a good main- gradually diminishing segments, with a terminal segment 

 equal in length to about three previous segments. 



In Sphingids and Manducids the hook is shorter, fewer segments being involved 

 in the dwindling tip, and the final segment is very long, equal to about six previous 

 segments. The Manducid antenna is straighter and thicker than that of Sphingids, 

 and the final segment is therefore proportionally more slender." 



The connection between antennal structure and habits is well- 



