76 MR. R. TRIMEN ON BUTTERFLIES [Jan. 20, 



This individual is smaller than the type, expanding only 2 in. 5| 

 I'm. The ocellus on the superior half of the disk in the hind wings 

 is relatively smaller, and there is a similar ocellus (only half the 

 diameter of other) on the lower part of the disk between the first 

 and second median nervules ; also in the fore wings there is a less 

 conspicuous but quite similar small discal ocellus between the first 

 and second median nervules. These additional ocelli occur also, 

 though less distinctly, in a larger example from Chaponga on 

 the Zambesi, and the fore-wing ocellus faintly appears in another 

 from Mashuna-land, both taken by Mr. F. C. Selous. 



The characters noted approximate to those of the intimately allied 

 P. nachtigalii, Dew'itz (Z. c. p. 194, t. xxv. f. 16), described from a 

 single specimen taken by Pogge in Angola (lat. 1U° S.) ; but the 

 underside agrees with that of the typical form, possessing a very 

 well-marked median streak in the hind wings but wanting the three 

 ocelli of the Angolan form. 



In the Hewitson Collection specimens of P. ai'taxia are also 

 recorded from Lake Nyassa. 



Genus Crenis, Boisd. 

 25. Crenis natalensis, Boisd., var. (Plate IX. fig. 12 S •) 



Crenis natalensis, Boisd. App. Voy. Deleg. dans I'Afr. Aust. 

 p. 592. n. 80(1847). 



? Crenis amazula, Mab., Grandid. Hist. Phys. etc. Madag. p. 153, 

 pi. xvii. ff. 9, 10 (1885-87). 



Orarora (November) and Okavaugo River (December). Four- 

 teen male examples. 



These specimens are all distinguished by a very much paler 

 ochreous-yellow upperside, and a very much paler lilacine-greyish 

 underside of the hind wings and apex of the fore wings, than are 

 found in the male G. natalensis ; but still more remarkable is the 

 fact that, althongh very faintly shown, the darker and paler marking 

 of the apical area of the upperside of the fore wings is that proper to 

 the female (not to the male) C. natalensis. Indeed, these un- 

 questionable males from tropical S.W. Africa look very much like 

 the female C. amazula figured by Mabille ' . They differ, however, 

 in having the basal half of the wings much yellower (almost free 

 from any darker clouding), and the costal-apical dark markings of 

 the fore wing much fainter and less developed ; on the underside 

 the latter distinction is also noticeable, but all the small black mark- 

 ings on this surface are more developed (especially the subbasal 

 ones in the hind wings), and the yellow stripes bordering the 

 ocelli of the hind wings are much deeper in colour and more strongly 

 marked. These differences are all of course more marked in com- 



^ M. Mabille himself remarlvs {I. c.) that C. amazida may perhaps be only a 

 form of C. natalensis. He adds that he had adopted the name {amazula) giyen 

 in Boisduval's collection to a specimen from the " Cote d'Afrique," that the 

 form is rare in Madagascar, and that he had seen only two examples in the 

 Paris Musemn. 



