82 
Annals of the Transvaal Museum 
'boundary ridge is not present: this marginal surface rapidly narrows in passing 
ventralwards but is continuous throughout as a more or less definite though 
very narrow marginal strip free of coarse punctuations. There is also a much 
deeper secondary marginal surface, very coarsely pitted, which in the posterior 
half of the shield is delimited by a distinct ridge in the type : it is glossy and i 
hairy like the dorsal surface, of which indeed it is a part : in the Potchefstroom 
specimens, however, it is less noticeable owing to the absence of the ridge of 
separation between dorsal and secondary marginal surfaces, but, on the other . 
hand, the primary marginal surface is rather more sharply defined but not 
deeper than in the type. The shield of these Potchefstroom specimens is 
decidedly more hirsute than in the type, thus approaching hirsutum, where, 
however, the relationships of dorsal and lateral surfaces are quite different. 
Although the posterior ridge is absent in this variety yet, a deep secondary 
lateral surface may be recognised especially posteriorly, and in fact, as in the 
typical form, constitutes a complete but very indistinct girdle, the curvature of 
the superior surfaces being greatest along the subcircular line of junction: in 
the smaller example, these two surfaces in the mesial line posteriorly may be 
said to be angularly inclined to each other at about 120°, the secondary mar- 
ginal surface being not curved, but in the larger specimen the angle is greater, 
and the two surfaces merge to a greater extent. This variety I now designate 
Galeosoma coronatum var. spheroideum. Total length of the shield 10 -8, 
greatest breadth 8-2, anterior depth of true marginal surface 2-4, posterior 
depth of same -4. 
Galeosoma robertsi Hewitt (PL IV, fig. b). 
Four female examples have been taken on the town lands adjoining the 
experimental farm at Potchefstroom (G. van Dam and A. Roberts). The ridge 
separating dorsal and secondary marginal surfaces is quite well developed, 
extending into the anterior half of the shield. Otherwise, they do not differ 
appreciably from Pretoria specimens of this species. 
At Venterskroon (about twenty miles S.E. of Potchefstroom) Mr van Dam 
has found two specimens representing a distinct form of this species. These 
examples chiefly differ from the typical form of robertsi in the possession of 
long hairs on the upper surface of the shield: the hairs are rather sparsely 
distributed, being not quite so abundant as those on the shield of coronatum 
typicum. This variety may therefore be known as Galeosoma robertsi var. 
crinitum. The primary marginal surface in the typical form of robertsi is quite 
sharply differentiated from the secondary marginal surface, and, though 
greatly reduced in depth except anteriorly, is nevertheless continuous through- 
out uninterrupted by punctuations or furrows: in the Venterskroon specimens 
the two surfaces are not so sharply separated, and posteriorly the primary 
surface disappears altogether, the whole depth of the shield being coarsely 
pitted or furrowed posteriorly. The shape of the shield on the whole agrees 
with that of robertsi typicus. 
This variety does not differ greatly from typical coronatum, and may prove 
to be completely connected therewith by intermediates: at present, the two 
seem separable in the position of the posterior ridge delimiting the upper and 
secondary marginal surfaces: in the shield of robertsi, when viewed from the 
side, this ridge is approximately in a line with the anterior ridge separating 
the dorsal and marginal surfaces: in coronatum the posterior ridge is on a 
higher level at its anterior extremity, the depth of the marginal surfaces there 
being greater than one-third of the transverse distance between'the ridges of 
