o24 MR. H. W. BATES ON THE LEPIDOPTERA 



long and straight, directod outwards, and the upper straight, nearly reaching the apex 

 of the wing. In the female, the lower disco-cellular forms a less obtuse angle with the 

 median ; the recurrent nervule is emitted either close to the junction of the lower radial 

 or fi'om the middle disco-cellular ; the latter is directed across the wing, joining the sub- 

 (tostal ; the up})er radial, is either emitted close to this junction, or is situated as a branch 

 of tlie subcostal, after the cell. The male fore tibise and tarsi are reduced to a small 

 knob ; the femur is not notably abbreviated. The female fore tarsi have the joints 

 slender and filiform. 



This genus is interesting as exhibiting the wing-neuration in a vacillating state. Not 

 only do the sexes show an imjiortant difference in the position of the angle and recur- 

 rent nervule of the hind wing disco-cellulars, but individuals of the same species 

 vary in the position of certain ner-^iires. Those parts of structure which form fixed 

 generic characters in other groups are here variable in the sexes and in individuals 

 of the same sex. Ceratmia is nearly allied to MechanlHs (as defined in this me- 

 moir), on the one hand, and to Ithomia, through such species as /. Ij^jhianasso, on 

 the other. 



1. Ceratinia Ninonia, Hlibner. 



Ceratinia Ninonia, Hiibn. Exot. Schmctt. 



Hlibncr's figures represent an insect with ratlier l)road fore and hind wings, and with 

 two large yellow spots across the middle of the fore Aving, besides a crooked yellow belt 

 across the black apical part. I found a species extremely common at different stations 

 on the Amazons, which Avas evidently the same as Ninonia, but very variable in shape 

 and colours, and presenting very few examples which agreed exactly with Hiibner's figures. 

 The species, however, evidently varies in different ways in different localities ; yet the 

 local varieties are not definite, the segregation of the races is not complete ; so that it 

 is embarrassing to decide whether to treat the form as one polymorphic species, in- 

 ciiuling the variations under one and the same definition, or to describe separately the 

 ty})e and the local varieties. Besides these incomiDlete local modifications, easily trace- 

 able to the type, there are, as often happens in the case of prolific, widely distributed, 

 and variable species, a numbc^r of other forms rather more strongly marked and better 

 defined, which inhabit regions rather more distant from the locality of the type than 

 those which the mere varieties inhabit. These are admitted on all hands to be distinct 

 species ; l)ut I think it would be diificult to |)rove that these were not also varieties of C. 

 Xiiioiiia, which have become more completely segregated from the parent fonu. Such are, 

 amongst others, C. T/'m (Hewits.), Eio Negro; C. Leprieuril (Feisthamel), Cayenne; 

 C. Fcncstella (Hewits.), Venezuela; C. Mclphis (lliibn.), S.E. Brazil; and C. Fiwhria 

 (Hewits.), New Granada. 



The ibllowing are the chief varieties of Ceralinia Ninouia occurring in the Amazon 

 region. 



\ ar. 1. C. Borri (Boisduval's Coll.). 



Expanse 2" 1"' to 2" 3'". Uing wings in l)oth sexes much narrower than in the type; 



