South- African Land-MoUusca. 459 



narrower shell-lobes, " Approaching the forms timidly 

 known as H. natalensis and H. poep/n^i, but possibly dis- 

 tinct. '^ It is very close to No. 3245 (bicolor), but the shell 

 of that species has a narrower periplieral band. More 

 mature specimens of tliis species are required, and they should 

 be compared with K. natalensis from Durban or its neigh- 

 bourhood. 



From another specimen. 



Locality. ^Nlaritzburg (^Henry C. Burni/p), 21.iii.08. 



Animal. — Visceral sac. Wall of branchial cavity pale- 

 culuurid icith no marking, beyond a slight dark streak above 

 tlie kidney. Behind the heart dark grey sparsely spotted white, 

 merging into dark brown on ape.v, ivith the sutural margin 

 bordered ivhitish. Foot with a long horn over the mucous 

 gland, well-defined peripodial grooves, with others leading 

 from them to the keel. Right shell-lobe long, broadisli, 

 given oft' considerably below the rectum. The left is small, 

 triangular. Left dorsal lobe long and narrow, in two 

 sc])arate parts of about equal length. 



The generative organs may be compared with those of 

 K. pluedinius. The penis is bent on itself in S-shape, the 

 flagellum long. I give a drawing of its position as packed 

 within the animal and adjacent to the spcrmatheca (pi. xv. 

 (Hg. 1 b), shown again after extraction (pi. xv. fig. 1 c). 

 The free oviduct is not black as in K. phcedimus [ot.). 

 There was a single spermatophore (Part II., pi. xv. hg. 1 d), 

 quite perfect. This has six spines on one side next the 

 capsule, with twenty-one on the other side (^). Their 

 elongate form may be compared with those of No. 3379, 

 K. poeppigi, from Pinetown, near Durban. The branching 

 is something like those of phcedimus^ but far longer. 

 These three species arc evidently very close to each other, 

 the shell-lobes differing in breadth and length. Compare 

 figures of phiedimus on pi. xiii. figs. 1-2, No. 3379, pi. ii. 

 tigs. 3, 3 a, and No. 7, pi. xv. figs. 1, 1 a. 

 The radula shows a formula: — 



70 . 3 . 12 . 1 . 12 . 3 . 70, or 85 . 1 . 85. 



The centrals of usual type, the laterals all evenly bicuspid, 

 becoming very small on the outer margin. 



Jaw with a central projection. 



Zonantydrus, M. & P., and subcorneus, Preston, appear to 

 be identical, and are very near this species of the natalensis, 

 krauts, type of shell, quite smooth on the a])ical whorls. 



