40 BRITISH LEPIDOPTERA. 



with darker transverse stripes and bands. Hindwings 

 almost unicolorous brown — syriaca, Led. 



In considering the general appearance of the imagines there can be 

 no doubt that Hippotion celei'io is especially Sphingid (sens, restr.), and 

 it is also nearer Sphinx in the structure of the large proboscis and eyes ; 

 Theretra porcellus and Eumorpha elpenor are the least so ; Phryxus 

 livornica is more Sphingid (sens: restr.) than the two last-named. 

 Separating Daphnis nerii (typical of a large tribe) as having markings 

 very unlike the rest of the Eumorphids that we find in Britain, 

 we find that the markings that cross the forewings show some 

 method in their evolution, and this is no doubt to be correlated 

 with different resting-habits. Taking T. porcellus with the oblique 

 line from apex to anal angle and two transverse lines, one before 

 and one beyond the middle of the wing, we find in E. elpenor a 

 coalescence of the two outer lines into one reaching from the apex 

 to the inner margin, and a second, almost median, line ; then 

 follow Hyles euphorbiae and Celerio gallii, then Phryxus livornica, and 

 lastly Hippotion celerio. The advance is by the greater straightness 

 and definiteness of the marking starting from the apex, and by 

 its reaching the inner margin nearer and nearer the base. This 

 is somewhat parallel to (i) Greater length of proboscis. (2) Larger 

 eyes. (3) More dorsal position of pupal labrum.* There is the further 

 stage beyond H. celerio, where the pupa has a horn. The loss of the 

 larval caudal horn in T. porcellus, &c, must be looked upon as a 

 specialisation!. The habits leading to this appear to have absolved T. 

 porcellus from competing with the others as to proboscis, &c, but 

 one cannot decide which is the higher of the two forms, specialised 

 in different directions; probably one would suggest that an imaginal is 

 higher than a larval specialisation. Without discussing the matter further, 

 we have no doubt that we have, represented among the species 

 taken in Britain, at least four very distinct tribes, which may be 

 tabulated thus : 



I. Tribe: EUMORPHID1 — Theretra porcellus, Eumorpha elpenor. 



II. Tribe : Phryxidt — Hyles euphorbiae, Celerio gallii, Phryxus livornica. 



III. Tribe : HipporiONiDi — Hippotion celerio. 



IV. Tribe : Daphnidi — Daphnis nerii. 



In suggesting the above groupings we know we shall be met with 

 at least one strong objection, viz., that the larva of Hippotion celerio is 

 nearer to those of the Eumorphids (sens, strict.) than to those of the 

 Phryxids, but, judging from figures of the larva of H. celerio, one 

 is inclined to doubt whether the resemblances are really fundamental. 

 It is true that the larva of Hippotion. has ocellated spots of a character 

 similar to those on the larva of the Eumorphids (sens, strict.), and 

 so also, in a less degree, has the larva of Daphnis nerii, but there 

 seems no real connection between the ocellated spots of Hippotion 



* These are, as pupal characters, to be correlated with the imaginal feeding-habits, 

 i.e., search for flowers at twilight and dusk requires large eyes ; the long trunk is 

 connected with feeding whilst hovering over may-be large and conspicuous flowers 

 without alighting; the position of the pupal labium is due to the modification of the 

 tongue, and cannot well be considered as a separate character apart from greater 

 length of proboscis (Bacot). 



f The loss of the horn in T. porcellus has probably been due to the habit of 

 hiding in crevices in the ground, where it would be very liable to injury (Bacot). 



