ORTHOPTERA 



FAM. BLATTID/E 



SUBFAM. EPILAMPRIN.^ 



by K. SHELFORD 



WITH 2 COLOIRED I'l.ATES 





^w~iV^ HK Epilamprina- form the fnurtli division in the classiticatioii of the Blattidae. 



Characters. — Antennae setaceous, never plumose, vciy occasional!}' incrassated 



slightly. Pionotum variable in form. Tegmina coriaceous or corneous, fully developed or 

 reduced. Wings fulh' developed, reduced or absent, mediastinal vein typically multira- 

 mose, costal veins irregular, ramose, ulnar vein with several incomplete rami. Supra-anal lamina : 

 {(^) more or less quadrate with obtuse angles, (9) sub-bilobate. produced. Femora armed beneath spar- 

 selv or strongly. Tarsi with distinct pulvilli, and in all hut one genus with arolia. Ovo-viviparous or 

 viviparous. 



This sub-familj' presents almost as many difficulties to the systematist as the Phyllodromiinae; 

 Saussure attempted a revision of it in i8q5 {Rev. Suisse Zool.), but in reality only mdicated the lines 

 along which revisif)n should proceed. The most important characters for tax(3nomic purposes are 

 found in the tarsal structure and as nearly all authors have omitted detailed notice of these characters 

 in their specific diagnoses, no realh' satisfactorx' revision can be made until the vast majority of type- 

 specimens are re-examined. The scheme of classification here adopted can only be tentative and 

 extended knowledge of the group will doubtless lead to a shuffling of many species. Further subdivision 

 of the genus Homalopteryx based on the form of the tegmina and tarsal structure and of the genera 

 Calolampra and Ai'uireia based on the tarsal structure is possible and. as I believe, desirable, but is post- 

 poned until more material for such sul:)division comes to hand. Epilampra as a genus is as unwieldy as 

 Phyllodroviia and is even more difficult to deal with: the species of the latter genus do present readily 

 recognisable characters useful for splitting the genus into well-marked secti<ins and the difficult}' of the 

 svstematist lies mainh' in the fact that manv authors fail to describe these characters. But in the case of 



