( 11- ) 



HF.TEROCERA. 

 SFHINGIDAE. 



r\ Macroglossum stellatarum (Liun.). 



1 (?, South Oned Mya. 



This widespread species is comiuoii all over Algeria, and wc have sjieciiueus 

 taken in previous years from almost every place we visited. 



[We saw single specimens in various places from Biskra to the southernmost 

 part of the Oued Mya.— E. H.] 



21). Celerio euphorbiae deserticola (Bartels). 



1 c?, north of Jill-Golea, 



The most northern locality I have specimens of this race from is El Kantara. 

 When Dr. Jordan and I wrote our revision of the Sphingidae, the series in the 

 Tring Museum consisted of a few dealers' specimens labelled " Mnaretania" 

 This led us to erroneously unite the desert form of Celerio euphoibiae with the 

 more northern (_'. e. mauretanica. This is not the case, as we discovered, since 

 I Iiave collected and procured in Algeria a large series of caught and bred 

 specimens of the true <'. inauretanica. The most southern localities I havo 

 examples of the latter from are Khenchela and Batna, in the Aures Mountains. 



The larvae of (.'. e. deseiiicnln are easily distinguished from those of C. e. 

 mauretaiiica by their pale green ground colour and sparser dark markings. They 

 feed, as far as I have been able to ascertain, exclusively on Euphorbia guyoniatm, 

 while ( '. e. maui-etanica feeds on several species of Euphorbia. I have taken the 

 larvae of ('. c. deserticola at Biskra, Zaatcha, and along the road to Touggourt in 

 large quantities, and caught tlic moth at light freipiently at Biskra and once at 

 Ghardaia. 



[The i caught north of El-Golea, on May '^1, was the only imago seen, but 

 young and old larvae were found in the same district on Euphorbia (/ui/oniana. 

 — E. H.] 



21. Celerio lineata livornica (Esp.). 



1 cJ, Bordj (Jlicgga ; 1 t?, half-way between Onargla and El-Golea; 1 c?, ? ¥, 

 north of EMiolea ; 1 ?, El-Golea ; 1 ¥, South Oned Mya ; 3 c?c?, 7 ? ?, north of 

 Ain Guettara ; c?c?, a ? ¥ , Ain Guettara ; 1 cj, In-Salah, Tidikelt ; 1 c?, Oned Nga. 



This is a widely spread race, being recorded from Portugal on the west to 

 Japan ou the east, and from Scotland in the north to Cape ( V>lony in the south. 

 C. lineata is found all over the world, and has three sharply defined subspecies : 

 C.lincal'i. ^;«t'a/r<, confined to tlie Western Hemisphere; <\ lineata lifornicoides, 

 confined to Australia ; and ' '. lineata livornica, inhabiting the rest of the Eastern 

 Hemisphere. 



Whether the race under discussion is a permanent inhabitant of Cenlral and 

 Northern Asia and Japan 1 I'nr the moment cannot accurately ascertain, but in 

 Europe it is certainly only a migrant from Africa, and not always a regular 

 one. The last great livornica year was lOOti, when great numbers invaded 

 Enrol)!' ; one collector in England cajjtured eleven in two evenings, and Dr. Hartert 

 and I found tlieni in such masses at Lnclion that they were ipiite a nuisance 

 when working tin' lamps. On our |)receding expeditions to Algeria we have 

 caught them at Alger, Biskra, Ghardaia, Gnelt-cs-Stel, and Khenchela, 



