( 439 ) 



granules of prniiotnm cnnspicnons ; bora also strongly granulose ; a pale dorso- 

 lateral line from head to horn, an inconspicuous line at each side of dorsal mesiM,] 

 line. — Food-plants : Eubiaceae : Lonicera ; Viburnum ; Prunus ; Scabiosa ; etc. 



Pu]ia without gloss, nearly black, rough ; two small frontal tubercles ; tongue- 

 ease not carinate ; cremaster flattened, triangular, rough, ending in two points, sides 

 rough with setiferous tubercles. 



Hi(b. Nearctic, Palaearctic, and Oriental Region, most species in the Palaearctic 

 Region. 



Grote has recognised from the beginning of Lis studies in Lepidopterology 

 that the sjiecies of Ihiemorrhagia { = Hemaris auct.) are generically distinct from 

 Macroglossum stellatarum and allies. We emphasise this fact, as most Palaearctic 

 writers — from the time of Scopoli, who invented the term Macroglnasum for 

 "stellatarum, etc.," and Dalman, who erected his genus Hemaris for stellatarum, 

 fucij'ormis, and titijus, down to the (Catalogue of Staudinger & Rebel — did not 

 recognise the wide dift'erence between the two gronps of species. 



Hnemnrrhugia is a northern development extending with two of its species into 

 the Oriental Region, being entirely absent from the Aethiopian and Neotropical 

 Regions. 



The species, so far as a series of specimens is known of them, are nearly all 

 obviously variable. The variation refers to size and colour, and to the dentition of 

 the marginal bands. Very fortunately, it has been i)roved by breeding that two of 

 the American species are variable seasonally and individually, so that the reader is 

 amply prepared for seeing the number of so-called species greatly reduced in this 

 Revision. Instead of the twenty American species of Kirby's Catalogue we 

 recognise as distinct only four, one of which consists of three subspecies, and there 

 are only eleven Old World species. 



Though there is strongly marked seasonal dimorphism in several species, the 

 differences do not appear in all individuals of the same brood. The most e.xtreme 

 summer forms do not seem to occur in springtime {thi/she f. loc. fuscicaudis, 

 (Uffmis f aest. axillaris), while sjjecimens of the character of the sj)ring form ajipear 

 among the summer brood. Further observations are a desideratum, especially in 

 Amurliiiid and Jajian, where the individuals with heavily dentate wing-borders will 

 jjrolialily turn out to belong to the summer broods. 



It is worthy of note that the American " species " were separated by Grote into 

 three suVigenera, which correspond to three species of this Revision. The fourth 

 siiecics was not known to Grote. 



In the recently published Catalogue of Palaearctic Lepidoptera by Staudinger 

 & Rebel, Staudinger erroneously put croatica under Macroglossum, where it does 

 not lielong. His scabiosue var. brunnrobasalis is the same as maiidarina, a form of 

 rinliaiis ; tfiiuis jiut with : under scabiosae has nothing to do with it ; margiiialis 

 treated as a synonym of " var." confinis is, like temns, a form of diffinis ; allinis is, as 

 correctly suggested, the Pacific representative of fucij'ormis ; var. robusta is not a 

 " var.," i.e. geographical form. 



Tlie asymmetrical development of the male se.xual armature is not so pro- 

 nounced as in some species of (Jephonodcs (see this), but mostly very obvious. We 

 find the tenth tergite divided in two processes in all the species, the left (right in 

 figures) process never aborting, as is the case in some species of Vephonode.s. The 

 tenth stc-rriite is nearly symmetrical in fucij'ormis and tilijus (PI. XLIll. f. 'I'S), and 

 iiiived Inwaiils llic h'I't sidct in the (;ther sjiecics (I'l. .\hill. I'. :V-\- ~^'<). l''r<)ni 



