collectedly W. J. Burchell in Brazil. 333 



last few years the two species have never been distinguished 

 by any systematist, and they are still unseparated in every 

 collection that has failed to take into account the researches 

 of Riffarth and Stichel. The remarkable character of the 

 resemblance will be realized when it is remembered that 

 Westwood, having before him the whole series of 69 Burchell 

 specimens (including 6 //. nanna), wrote on a label affixed to 

 1395, — " 69 individuals in full of this species without any 

 variation of the least importance taken bet n . 21. 10. 25 & 

 24. 3. 29. See list in my Burchell Catalogue." 



[In attempting to understand Burchell's use of the name 

 " horta" it is important to ascertain the various forms to 

 which he applied it. These are as follows: — 



(1) True Heliconiinaj, such as //. nanna and H. erato 

 phyllis (st e 1315) ; 



(2) The Ithomiinse, Mechamtis lysimnia or Ceratinia dceta 



applies to H. nanna. The toothed bar of H. nanna is also carefully de- 

 scribed by Riffarth (iu Berliner ent. Zeit. 1901-2, vol. xlvi. p. 106), and 

 by Stichel and Riffarth in ' Tierreich,' vol. xxii. p. 134 (Berlin, 1905). 



I therefore suggest the name burchell i for the subspecies in which the 

 lower end of the red bar crossing the fore wing is not toothed or bears 

 only a trace of the lower tooth. This subspecies is found over an area of 

 unknown extent lying to the north and west of the Brazilian range of 

 H. nanna nanna and H. erato phyllis. 



cJ type from Porto Real (Nacionale) on the R. Tocantins : captured 

 Mch. 2, 1829, by W. J. Burchell : (1 31 5) in Hope Department, Oxford 

 University Museum. 



$ type from Villa Nivac, S.E. of Matto Grosso (Nov. 1901), in coll. 

 W. J. Kaye. 



H. nanna burchelli appears, as Mr. W. J. Kaye has suggested to me, to 

 be a form of H. melpomene amandus, Grose-Smith and Kirby, from Bolivia 

 and Peru. Mr. Kaye has kindly given me the opportunity of examining 

 two males of this species from the Mapiri R., E. Bolivia. Both of these 

 possess the yellow spot usually found over the end of the fore wing cell in 

 H. nanna burchelli and sometimes in nanna nanna. The only essential 

 difference between the two forms is due to the loss in burchelli of nearly 

 the whole of the red area within the fore wing cell of melpomene amandus 

 and the distal lengthening in the former of the yellow hind wing baud 

 of the latter. Another closely related form is, as Mr. Guy Marshall sug- 

 gested to me, H. amaryllis rosina, Boisd., from Colombia, Panama, and 

 Costa Rica. This subspecies is, however, without the longitudinal yellow 

 streak which is so conspicuous a feature in the basal half of the fore wing 

 of melpomene amandus and the forms of nanna. 



It is probable that all the forms mentioned in the preceding paragraph 

 will prove to be subspecifically related, melpomene amandus bearing this 

 relationship to amaryllis rosina in the north and to nanna burchelli in the 

 east. However this may be, there can be no doubt that nanna nanna is 

 a geographical race of burchelli, produced to the east and south of the 

 latter, under the influence of H. erato phyllis. — E. B. P.] 



