14 PROCEEDINGS OF THE NATIONAL MUSEUM. 



tions have been examined by either the writer or S. A. Rohwer, and 

 the United States National Museum collection contains t^^pes, homo- 

 types, or specimens compared with the types of most of the species 

 discussed. 



With a few exceptions the species of this tribe are rare. Most of 

 them have been described from uniques, and some are still repre- 

 sented in collections solely by the tj^pes, while a few are known only 

 from the original descriptions, the types having been lost or destroyed. 

 Very likely with more specimens available for study some of the 

 species would be found to be synonyms. Some synonymy has 

 already been published and more is indicated in the present paper. 



All of the species of which the habits are known are externally 

 parasitic on spiders, the larva lying curved around the anterior 

 margin of the abdomen of the host. These records include members 

 of all the genera and subgenera except the new genus described below. 

 As suggested by Marley ^ the European records of Gravenhorst and 

 others ascribing hosts other than spiders to members of the genus 

 Polyspltincta are the result of the incorrect determination either of 

 host or of parasite. The fullest and most interesting account of 

 the attack of one of these insects on its host and the subsequent 

 development is by Bignell ^ who watched a female of Acrodacfyla 

 madida Haliday attack its host, the spider Linyphia ohscura. The 

 polysphinctine cocoon is highly specialized, being usually loosely 

 woven and having at the caudal end an opening through which the 

 meconial discharge and exuvia are voided. 



Most authorities have treated the genera Acrodactyla Haliday and 

 Zatypota Foerster as subgenera of Polyspltincta Gravanhorst while 

 treating Colpomeria Holmgren as a distinct genus. But Foerster, 

 followed by Ashmead, treated them all as genera. All of the writers 

 who have considered these groups as subgenera also included Za- 

 glyptus Foerster because of its incomplete areolet, although on all 

 other characters it is much more closely allied to Tromatohia Foerster. 

 Moreover, certain species of PolyspJdncia have the areolet fully 

 defined, while the type of the new genus, Zahrachypus, has it penta- 

 gonal in position though open behind, almost exactly as in Zaglyptus. 

 The character employed for separating Colpomeria, the angularly 

 incrassate front femora, is a unisexual character, being found only 

 in the female; but the cristulae on the front of the prescutum are 

 common to both sexes and therefore of more value. This character 

 will not separate Colpomeria from Polemoplithorus Schulz (=Symphy- 

 lus Foerster), which Morley ^ tells us is synonymous with Acrodacfyla 

 Haliday. In this synonymy Morley is in error if a specimen in the 



1 Brit. Ichn., vol. 3, 190S, pp. 119-120. 



2 Trans. Devon. Assn., 1S98, pp. 471-472. 

 8 Brit. Ichn., vol. 3, 190S, p. 131. 



