254 Mr. R. Shelford's Shcdies of the Blattidae. 



Avith accuracy the great majority of genera and species 

 described by Mr. Tepper ; in addition the types of Walker 

 in the British Museum have been examined. Dr. Th. 

 Kuhlgatz, late of the Berlin Museum, has sent me draw- 

 ings of Erichson's types and has compared specimens sent 

 to him. Dr. K. Holdhaus of Vienna has lent me some of 

 Brunner's types and Dr. Y. Sjostedt of Stockholm all the 

 types of Sttil. I trust that the net result of the consider- 

 able correspondence and hard work expended on this 

 memoir is a satisfactory claissification which will bear the 

 test of time and enable .other workers to identify the 

 species of the group without great difficulty. 



The Blattinae can be divided into two chief sections 

 according to the structure of the posterior tarsi ; the first 

 section, wdiich may be styled the Pol i/zoderia-group, has the 

 posterior metatarsus usually shorter than the succeeding 

 joints and usually unarmed beneath,* the succeeding 

 joints are entirely unarmed and with large pulvilli; the 

 second section or Blatta-gYoup has the posterior metatarsus 

 longer than the succeeding joints and armed beneath, the 

 second and third joints are also armed and their pulvilli 

 are apical (PI. Ill, f 40). A revision of the second section 

 I hope to undertake on some future occasion. The 

 Folyzostcria-grou-p can be subdivided into an Old- World 

 group and a New- World group, but the character on 

 which this subdivision is based — the relative distances 

 apart of the eyes and the antennal sockets — is so subtle 

 that I doubt its practical importance and I temporarily 

 regard the New-World group as distinct from the Old- 

 Worid group for convenience sake only. In the Old World, 

 Australia is undoubtedly the head-quarters of the Polyzo- 

 steria-gxonp, nearly all the apterous species are confined to 

 that continent and the islands immediately surrounding 

 it; but one or two, and notably Outilia nitida, Br., have a 

 much wider range, extending into the Oriental region and 

 Polynesian islands. The occurrence of an allied species at 

 Bombay is very singular and an error in locality is 

 suggested. As might be expected the winged species 

 (genus Methana) have a wider range than is characteristic 

 of the wingless forms, but even these do not extend to the 

 Asiatic mainland. The group in Africa is represented by 

 the single monotypic genus Paramethana. The geogra- 

 phical distribution of these cockroaches would appear to 

 * Gntilia is an exception. 



