114 BRITISH LEPIDOPTERA. 



(teste Bartel). Scandinavia: Scania — Helsingland, and southern Norway (Zetter- 

 stedt), not rare, northern limit Helsingland (Aurivillius), Gudbrandsdal (Schoyen), 

 Norway— Christiania, Vestre Aker, Odalen, Hedemarkia (Siebke), Hunneberg 

 (Lampa). Spain: Galicia — Santiago (Macho- Velado). Switzerland: almost 

 everywhere, but in some places not common, goes up 5000-5500^. (Frey), 

 Weissenburg (Pluguenin), Grisons (Killias), Bechburg (Riggenbach-Stehlin), Bern 

 (Hiltbold), Lausanne (Chaumette), near Zurich (Dietrich), Upper Engadine 

 (Pfaffenzeller), Aigle (Lowe), near Nyon, on Genier See, common (Ratzer), 

 Interlaken, near Thun, near Vevey, Aargau (Wullschlegel), Bremgarten, frequent 

 (Boll), Schiipfen, not common (Rothenbachj, St. Gallen, not common (Taschler), 

 canton Glarus (Heer), Bergiin (Zeller). 



Tribe : Hippotionidi. 



There is considerable doubt whether this should be considered 

 a tribe apart from the Eunwrphidi, or whether its constituent 

 genera should not be considered as specialised items in the latter 

 tribe. Pupally and imaginally, the Hippotionids (celerio, osiris, 

 &c.) are much more specialised than the Eumorphids (sens, strict.) 

 (elpenor, porcellus, &c). On the other hand, the larvae are very 

 similar and appear to have specialised along the same lines, but 

 our knowledge of the early larval stages of the Hippotionids is 

 so entirely lacking, and the accuracy of some of that relating to 

 the adult larvae (see anted, p. 57) so open to doubt, that one feels on 

 very unsafe and uncertain ground. We may, however, point out 

 that Weismann, on larval characters, combines (Studies in the 

 Theory of Descent, transl. p. 193) our Hippotionids and Eumorphids 

 (sens, rest.) in his groups 2 and 3, which he diagnoses as : 



The subdorsal line more or less retained, eyespots on the 1st and 2nd 

 abdominal segments (these markings absent or at most only present in traces on 

 the remaining segments) — elpenor, porcellus ; celerio, nessus, lucasi, &c. 



The subdorsal line retained in very variable degree ; the eye-spots repeated 

 on all the abdominal segments — bisecta (silhete?isis), oldenlandiae, alecto, acteus, tersa. 



Bacot, who has given considerable attention to the larvae of 

 this group, suggests {in litt.) to us the following grouping: 



I. Adult larva with subdorsal lines; a single central mediodorsal ocellated 

 spot on the 1st abdominal segment ; subspiracular obliques with a tail to head 

 slope ; young larva with the pro- and mesothorax greatly lengthened — Elibia 

 dolichus. 



2a. Adult larva with ocellated spots on 1st abdominal only, very dorsal in 

 position, no (or only exceedingly weak, and of different character) repetition of 

 spots on 2nd abdominal — Panacra vigil, nessus, lucasi. 



2b. Adult larva with a large and greatly developed ocellated spot on the 

 1st abdominal ; a series of smaller spots of different shape and character on 

 abdominal segments 2-7 on line of subdorsal band — Pergesa acteus, llathia clotho. 



3. Adult larva with ocellated spots on the 1st and 2nd abdominal seg- 

 ments in subdorsal line, not repeated on following segments — Eumorpha elpcno>\ 

 Theretra porcellus. 



4a. Adult larva with ocellated spots on the 1st and 2nd abdominals, more 

 or less weakly repeated on the following segments — Isoples alecto, Xylophones 

 oldenlandiae, Fiorina japonica. 



tip. Adult larva with ocellated spots on the 1st and 2nd abdominals, the 

 character of the ocellated spots much nearer those of Isoples (alecto) than Eumorpha 

 {elpenor, porcellus), the subdorsal line very strongly marked — Hippotion celerio. 



Bacot has here advanced very considerably beyond Weismann's 

 grouping and one is able to see in this arrangement the ground 

 for at least two (possibly three) tribes on larval structure alone. 

 Inverting 1, 2a and 2b we have undoubtedly three evolutionary 

 Steps in the development of a larva of which Elibia is the highest 

 exponent known to us, and which might be termed Elibiidi. 

 Again ^a and 4/; appear to form two natural sections of our Hippotionidi. 

 His group 3 still remains the exponent of our Eutnorphidi 



