2 SOUTH- AFEIC AN BUTTERFLIES. 



the Linncan Society, vol. xxiii. (1862).^ Mr. Bates's later arrangement, 

 the details of whicli are given at p. 176 of the Journal of Entomology 

 for 1864, will be followed in this work, and the linear order stands 

 thus, viz. : — 



Family I. — Ntmphalid^. 



Sub-Family i. — Danaince. 

 „ 2. — Satyrince. 



„ 3. — BrassolincB. 



,, 4. — AcrcBince. 



„ 5. — Heliconinoe. 



„ 6. — Nymphalince. 



Family II. — Erycinid^. 



Sub- Family i. — Lihythceince. 

 „ 2. — Sfalachtinte. 



,, 3. — Erycinince. 



Family III. — Ltc^nid^, 



Family IV. — Papilionid^, 



Sub-FamUy i . — Pierince. 



,, 2. — Pajnlionince. 



Family V. — Hesperid^. 



In this classification the Family characters employed by Mr. Bates 

 are those above mentioned, viz., the structure of the fore-tarsi in both 

 sexes, and the mode of suspension of the pupa. For the grouping of 

 the Sub-Families of the Nymphalidcc he relies firstly on the development 

 or atrophy of the lower disco-cellular nervule at the extremity of the 

 discoidal cell ; secondly, on the shape and clothing of the larvae ; thirdly, 

 on the clothing of the palpi ; and fourthly, on the presence or absence 

 of a pre-discoidal cell in the hind-wings. He divides the three Sub- 

 Families of the Erycinidoi in accordance with the mode of suspension of 

 the pupa, either freely by the tail only, rigidly in an inclined position 

 by the tail only, or by the tail and a girdle.^ The two Sub-Families 



^ Mr. A. R. Wallace, whose opinion is entitled to the most careful consideration, opposed 

 the removal of the Papilionidce from the head of the Butterflies in his most excellent paper 

 on the Malayan members of the Family in vol. xxv. of the same Transactions ; and after I 

 had, in vol. xxvi. (1S69), adduced various structures in which the Papilionidce showed their 

 affinity to the Moths, he argued at length, as late as 1871 (see his Contributions to the Theory 

 of Natural Selection, 2d edit. ), in favour of their being retained at the summit of the Rhopa- 

 locera. But in his Geographical Distribution of Animals (vol. ii. 1876) I was glad to notice 

 that he had virtually abandoned his contention, by placing (p. 479) the Family at the end of 

 the series, next above the Hesperidce. 



^ In a subsequent paper of great value on the entire Family Eri/cinidce {Journ. Linn. 

 Soc, Zool., ix. p. 367, 1868), Mr. Bates gave up this character of the position of the pupa as 

 distinctive of the Sub-Families, having found that in a species of Emesis, one of the Ery- 

 cinince, the pupa was suspended as in the typical Stalachtince. He omits the Libythoiince, 

 and arranges the Family into Nemcobiina', Eurygonincc, and Erycinince, in accordance with 

 the number of branches of the subcostal nervure of the fore-wings, and (in Euryrjonince) the 

 position of the lower radial nervule in the hind-wings. As it is preferable, for purposes of 

 classification, to depend upon the characters of the imago, it will be well to accept this 

 amendment, but at the same time not to exclude the Libythccincv. 



