TYRINI 309 



an association. As to whether elongatus is Hamotus or Tyrus cannot be ascer- 

 tained from the original description, but a single glance at the palpi of the type 

 will settle the question as the two genera are wholly distinct, with no overlap. 

 Presumably this has been done by Bowman. The habitat is too arid for Hamotus 

 but I do not know this species. 



America north of the tropic of Cancer has but one ally of Hamotus. This 

 relative is the myrmecophilous, monotypic Cercoceru^ batrisoides Leconte 

 (1861, p. 57). This species is in entire agreement with the subgenus Hamotoides 

 of Hamotus. In fact it could be placed in Hamotus with no great difficulty as 

 the only differential between Cercocerus and Hamotus is the relative size of the 

 eleventh antennomere. In Cercocerus this distal segment forms the antennal club 

 and is about seven times longer than the tenth segment. In Hamotus the last 

 three antennomeres form the club but there is great variation in the relative size 

 of these segments, thus in some species the ninth and tenth segments are small 

 and the eleventh is longer than the preceding five united. 



Otherwise these two genera are congeneric, Cercocerus having the typical 

 hamotine palpi. ^ Cercocerus batrisoides is a remarkable pselaphid, with un- 

 usually strong secondary sexual characters as indicated in the plates. Its ap- 

 parent rarity in collections is probably a consequence of its being represented 

 by a stray specimen from log mold. It is an habitual synoekete of ants, being 

 perfectly unmolested by its host. I have usually taken it from nests of Aphaeno- 

 gaster tennesseensis Mayr. Leng (1920) lists it from Pennsylvania, Indiana, 

 and Louisiana. My series comes from Cincinnati, Ohio; Smith, Indiana; Topeka, 

 Kansas, all with the ant host noted (Park, 1935) . From these few data it appears 

 that the species does not occur south of New Orleans. It is a nearctic extension 

 of Hamotus, distributed through the Mississippi basin from the subtropics, east- 

 ward to Pennsylvania and westward to Kansas. Thus this species and Hamotus 

 opimus become of great zoogeographic importance. The distribution of batri- 

 soides is linked with its formicid hosts and we await more collecting records as 

 well as its detailed ecology. 



Within the neotropical region Hamotus has also many specialized relatives. 

 It is closely allied with Apharus through Hamotus alleei in a phylogenetic 

 shortening of the palpal sulcus and lengthening of the first tergite ; it is allied 

 to Pseudohamotus with its semi-circularly swollen distal palpomere but typical 

 palpal sulcus; it is allied to a long series of genera in which the distal palpomere 

 becomes elongate-cylindrical and the palpal sulcus very narrow {Cercoceropsis, 

 Cercocer aides) ; and more remotly allied to Pselaphocompsus and Cercocerulus. 



In other words, Hamotus is a stem genus which has one close ally in the 

 nearctic, and many allies in the neotropical region. Throughout its neotropical 

 range, the species seem to be a part of the rain forest fauna. They are especially 



'In Bowman's recent treatise (1934, p. 132) the genus Cercocerus will not key out. 

 In this key to tyrine genera Hamotus is given an apical palpal cone along with Tyrus, while 

 Cercocerus is given no palpal cone. 



This was probably a typographical error, as Tyrus is the only one of the three with a 

 terminal palpal cone while both Hamotus and Cercocerus have a distinct palpal cone set 

 obliquely within the apex of the longitudinal palpal sulcus. 



