ANTHRACOMARTI. 57 



The sternal elements of tlie opisthosoma appear to correspond in number with 

 the terga, except that, with the doubtful exception of Eopluijniis. there is no 

 sternal plate to represent the tenth tergum or anal operculum. Hence there 

 are nine sternal plates in all, the last l)eing the ventral element of the plate 

 surrounding the anal operculum. The anterior sterna are variously modified and 

 arranged accoi'ding to the genus, the arrangement in Antlirncomartus being very 

 different from that of Eophrynns, the first, whether large as in the former or small 

 as in the latter, being probably the genital operculum. 



The morphology of the anterior region of the opisthosoma in this Order opens 

 a wide field for speculation. I have already suggested two hypotheses to account 

 for the facts, either or neither of which may represent the actual truth (see Greol. 

 Mag. [4], vol. ix, pp. 443—448, 1902, and vol. x, pp. 250— 251 , 1903). But since 

 seeing examples of Anthracomartvs with its enlarged and posteriorly-projecting 

 first sternal plate, which recalls so forcibly the genital operculum of the Pedipalpi, 

 I am inclined to the opinion that my first-published vieAV is probably correct, 

 namely, that the genital orifice opened l)ehind the first sternal plate in Eoj^Jirynns 

 and Antliracomartus. This sternal plate, therefore, will be the sternum of the 

 genital segment, the tergum of which is apparently the short first tergal plate, 

 which is sometimes partially or wholl}^ overlapped by the hinder border of the 

 carapace. The tergal element of the pregenital segment seems to be present 

 both in Eojjlirynns and Kreiselieria ; but I have not been able to recognise its 

 sternal region in these genera, although in Antliracomartus there seems to he a 

 comparatively large unchitinised area between the first sternal plate of the 

 opisthosoma (the genital operculum) and the sternal plate of the prosoma. 



The British genera referred to this Order appear to fall into the following four 

 groups : 



(1) Brachypyrie ; Maiocerciis. 



(2) Antliracomartus. 



(8) Anthracosiro ; Trigonotarhv.s. 



(4) Eophrynus ; Kreischeria ; Apliantomartiis. 



In 1902 I proposed to refer the genera Brachypyge and Avtiiracoviartiis to the 

 family AnthracomartidjB, and Eophrynus and Kreischeria to the Eophrynidte. To 

 these I added in 1903 the family Anthracosironidre for Anthracosiro. At that 

 time, however, I had not seen representatives of the genera Anthracomartus, 

 Kreischeria, and the two described below as Trigonotarhns and Aphantomartus. 

 The structure of the carapace shows that Aphantomartus belongs to the same 

 category as Eophnjnus and Kreischeria, and that Trigonotarhns is apparently 

 related to Anthracosiro. On the other hand, the structure of the ventral surface of 

 the opisthosoma in Anthracomartus is very different from that of Eophrynus and 

 Trigonotarhns, the only other genera of the order in which this region has been 



8 



