COLLEMBOLA— CARPENTER. 261 



some species of Achorutes, though here more strongly developed ; it diifers from 

 the thick spines in a cuticular groove or furrow that form the corresponding organ 

 in the genus Xemjlla (see Linnaniemi, 1912, p. 41, PL V., figs. 3, 4, etc.). The 

 presence of a post-antennal organ also serves to distinguish Gompkiocepkalus from 

 Xenylla. 



Jaws. — In my previous account of this insect (1908, pp. 2-4, figs. 1,/, 3, 4, 5) 

 I commented on the supposed absence of the basal region of the normal coUembolan 

 mandible, on an apparently mandibuliform maxillula, and on a supposed tooth-like 

 sense-organ on the hinder reo-ion of the head. These turn out to be mistakes into 

 which I ought not to have fallen, even though the material at disposal for study was 

 unsatisfactory. I can now withdraw and explain these errors. 



Mandible. — The mandible is of the usual coUembolan type (see fig. 7). In the 

 insects from the " Discovery " collections, the middle and basal regions of the jaw 

 had disappeared, leaving the molar area and apex which I figured (1908, figs. 3, 4), 

 and also the very acute condyle (fig. 7, c), which, breaking, in some of the specimens, 

 through the cuticle of the head, seemed to be an external spine (1908, fig. 1,/), so 

 that I was misled into describing a non-existent cephalic sense-organ. 



Maxillula. — This appendage (fig. 8, M) is also of the normal coUembolan type 

 and has the usual relation to the tongue (fig. 8, T). In my previous paper (1908, 

 fig. 5) I represented it like a small and delicate mandible. From recent studies it 

 is now clear that I was examining an immature insect in which the cuticle of the 

 next instar had already been formed beneath the mandibular cuticle then in use, 

 giving the deceptive appearance of another appendage. 



Maxilla. — This jaw also agrees closely in structure with its homologue in allied 

 genera (figs. 8,9). The lacinial head (figs. 8, la, 9) ends in two prominent teeth and 

 bears internally an elongate, curved bifid process, and three delicate leaf-Uke lamellae, 

 between which projects a slender spine (fig. 9). The elongate process (fig. 8, g) of the 

 maxilla — probably to be regarded as a galea and not as a palp (see Borner, 1908) — 

 has a bluntly conical tip, below which is the short papilla (fig. 8, p), now regarded 

 as a vestigial palp with its long terminal bristle. The maxiUae, and indeed the jaws 

 altogether in Gomjjhiocephalus , resemble closely those oi Achorutes. 



Legs. — There is little to be added to my former description (1908) of the legs. 

 The feet, as then stated, with their simple claw and tenent hairs, resemble those of 

 a Xenylla. In many of the specimens, at least, there are two tenent hairs on each 

 fore foot (fig. 10) and three on each intermediate and hind foot (fig. 11). 



Abdominal Segments. — The abdominal terga have elongate stifi" spines arranged 

 for the most part in three transverse rows on each segment. Each abdominal tergum 

 carries a pair of sensory hairs, two pairs being visible on the fourth and also on the 

 fifth segment (fig. 2, s). 



Spring and Catch. — The specimens now avaUable for examination show that 

 these organs, though reduced, are not in the vestigial condition suggested in the 



P 2 



