EUPLECTINI 115 



and metasternum have seven foveae developed (about parallel with the sternal 

 development of Bibloplectm) , foveae II, IV, and V are paired and III is 

 median and fused in the genotype and closely related forms. The metasternum 

 is usually simply convex in the female, and often sulcate or concave in the 

 male sex. 



The legs are usually secondarily modified in the males. This is diverse in 

 the genus. The femora of the males are generally inflated. In dentipes the 

 mesothoracic tibiae are medially spined; in others these tibiae are obtusely 

 prominent and pubescent medially, angulate, swollen and bearing a copulatory 

 pad on the mesial face at apical third {thoracica, sulcatula). This pad, when 

 examined on slide mounts at high magnification, is seen to consist of minute 

 semispherical transparent padules, the number of padules varying in some 

 cases with the species. The swollen male middle femora may bear large punc- 

 tures, or sensory pores, and these pores may be present on the anterior femora 

 as well. The tarsi are euplectine, with a minute first and a relatively large 

 second and third tarsomere ; second slightly longer, and much thicker than the 

 third; third bearing a large curv^ed claw. 



Although this present study is primarily concerned with the neotropical 

 Pselaphidae, I should like to take this opportunity to make up a brief for 

 Dalmosella of North America. Dalmosella (PI. IV, XII) was erected by Casey 

 (1897, p. 570) for a new species, tenuis, which I designate as the genotype, and 

 two doubtful inclusions, Trimium simplex LeConte and Trimium americanum 

 LeConte. Leng (1920) placed Dalmosella in Melba, and this course was fol- 

 lowed by Bowman (1934). I do not know simplex and americanum, but tenuis 

 is very distant from our present conception of Melba. Casey's original de- 

 scription of tenuis was based on a female, and the remarkable male characters 

 (PI. XII, 15, 16) were unknown to him. I have taken tenuis in copulation in 

 Illinois, and the male features have not been reported to my knowledge. The 

 following tabular comparison (Table II) between two genotypes will set forth 

 the striking differences between the two genera in question. 



The genus Melba shows numerous radiating trends into other melbaform 

 and some brachyglutine genera. Thus the general habitus and contiguous pos- 

 terior coxae are similar to the glabrous Eupsenius with subcontiguous posterior 

 coxae and an even larger distal antennal segment. The subglabrous Eupsenina 

 with very distant posterior coxae seems much more remote, but the capitate 

 setae of the ventral surface of the head are melboid. Melba frontalis suggests 

 certain outlines of Dalmosella and Ramelbida. Melba clavata approaches Mel- 

 bamima. Melba parmata suggests Dalmoplectus. It is evident that this genus 

 has been an active evolutionary center in the past, and there are probably 

 many new species awaiting discovery. 



Raffray (1904, p. 535) divided the genus Melba into four groups of species. 

 This grouping can no longer be maintained, since Group I of Raffray is Dal- 

 mosella, and the addition of many new species since this time, with increased 

 structural and zoogeographic information, makes a revision necessary. The 

 following key to subgenera is offered to bring our knowledge of this difficult 

 group up to date. 



