NOVITATES ZOOLOGICAE XXVI. 1919. 23 



[Castnia (Jephrostola) gramivora Schaus. 



There are uniortunately no Sao Paulo specimens of gramivora at Tring 

 for comparison, but there are 8 from Castro, Parana, collected by Mr. E. Duckin- 

 field Jones, who took Mr. Schaus's type. I must at once say that I do not imder- 

 stand how the entirely erroneous figure in Seitz was produced ; Strand 

 described his gramivora parana from my smallest specimen, so either the figure 

 is an utter travesty or, what is so often the case. Dr. Seitz, without consulting 

 the author, used a totally different insect for his plate. All my 8 specimens dis- 

 tinctly show the large brown patch starting from the costa and alnxst dividing 

 the large diaphanous space in two; they also have the " secondaries " (hindwings) 

 and their large submarginal spots much more rufous than in Mr. Oberthiir's 

 figure of Houlbert's feneslrata. I therefore am convinced that all 3 forma 

 gramivora, -parana, and fenestrata are only aberrations of one. insect, gramivora 

 Schaus.] 



Castnia (Tephrostola) gramivora Schaus. 



Castnia gramivora Schaus, Journ. New York Entom, Soc. vol. iv. no. 4. p. 147 (189G) (Sao Paulo). 



I consider that the only difference between Strand's parana and Schaus's 

 gramivora is that the basal area of hindwings and their submarginal spots 

 appear more rufous. This I consider merely aberrational. As regards size my 

 8 ?? vary much as follows : 



Length of forewing : 25 mm. — 32 mm. Expanse : 54 mm. to 81 mm. 



The specimen measuring 54 mm. in expaiirse is the "type" of Strand's 

 parana and which he says expands 45 mm. This is due to faulty measurement. 

 As I have proved by repeating this method, Strand must simply have placed 

 the specimen against a rule and taken the breadth between the points of the 

 wings ; this is bound to be wrong, as no two specimens are ever set with their 

 wings at absolutely the same angle. I take my measurements with a pair of 

 compasses from the thorax to the end of wing, then again with the compasses 

 take the width of thorax ; this by adding the width of thorax to the length 

 of both forewings gives the exact expanse of the insect. A more rough-and- 

 ready way of arriving at an almost equally accurate result is to measure from 

 the pin to apex of one forewing and then double the resulting figures. This 

 is the method employed by Sir George Hampson and is fairly accurate if the 

 pin is properly in or about the centre of the thorax. It will be seen that my 

 largest $ is the same size as the Schaus " type." 



8 ?? Castro, Parana, February 1897 (E. D. Jones). 



Castnia (Xanthospila) mimica Feld. 



Castnia mimica Felder, Reise Novara, pi. 79. fig. 7 (1874) (Ega). 



The Tring Museum possesses two specimens of this remarkable species: 

 1 (J Ega, Amazons (Bates, ex coll. Felder) (type) ; 1 6 JeSe, Amazons, 

 November 1907 (M. de Mathan). 



Castnia (Enicospila) marcns Jord. 



Castnia marcus Jordan, Novit. Zool. vol. xv p. 2.53 (1908) (Pebas). 



This remarkable species has up to now remained imique. 

 1 $ Pebas, Amazons, November 1906 (M. de Mathan). 



