260 "TERRA NOVA" EXPEDITION. 



a secoud species myself," writes Mr. Nelson, " and should on cursory examination 

 refer them all to Gomphiocephalus hodgsoni, Carp." The springtails seem to be, 

 indeed, all referable to this species, and the | mm. specimens mentioned by Taylor 

 are probably young individuals. 



The fragmentary nature of the specimens described in 1908 necessarily made 

 the account then given incomplete, and I regret to find that in some particulars it 

 was inaccurate. Before proceeding to details, it may be well to state briefly that 

 the reference of Gomphiocephalus to the Poduridae, in association with the genera 

 Achorutes and Xenylla, proves to be fully justified, but that the structure of the 

 insect is less aberrant than I imagined in 1908, so that there is no occasion to establish 

 a new sub-family for its reception. Gomphioce^yhalus proves to be a close ally of 

 Xenylla, and falls naturally into the Achorutinae (Hypogastrurinae of Borner, 1906). 



Order COLLEMBOLA. 



Sub-Order ARTHROPLEONA. 



Family PODURIDAE. 

 Sub-Family ACHORUTINAE. 



(= Hypogastrurinae, Borner, 1906.) 



Gomphiocephalus, Carpenter, 1908. 



Revised diagnosis. — Cuticle finely granulate. Abdomen with two anal spines. 

 Catch and spring present but reduced, mucro fused with dens. Foot bearing tenent 

 hairs, but without empodial appendix. Seven ocelli and a simple post-antennal organ 

 on each side of head. Feeler with a protrusible apical sense-organ on the fourth 

 segment, and a row of sensory spines bordering a sinuous, chitinous ridge on the 

 third. Jaws of the normal collembolan type, the mandibular condyle (see fig. 7, c) 

 acute and prominent, the maxilla with typical elongate galea and vestigial palp. 



Type.' — Gomphiocephcdus hodgsoni, Carpenter (1908), Granite Harbour, South 

 Victoria Land. 



The two important features in which the original description (1908) of this 

 genus needs correction are emphasised by italics. 



Ocelli can be distinctly seen in some of the specimens and are probably present 

 in all. There are seven on each side of the head, the arrangement being shown in 

 fig. 3, which represents the position of the left ocelli in relation to the post-antennal 

 organ ; the third ocellus of the inner row is the one wanting from the normal 

 Collembolan eight. 



Antennal sense-organs. — As to these, there is nothing to add to the description 

 given in the 1908 paper. The organ on the front edge of the third antennal segment 

 (figs. 4, an, 6) appears to be characteristic of the genus, resembling that found in 



