Annals of the Transvaal Museum 
13 
further. These characters are: the colouration, which does not vary 
much within a species, being similar in the two sexes, and fairly constant 
throughout large groups of species : the shape of the hind borders of the 
genital sternite, whether broadly rounded on each side or produced into 
a pair of mesial lobes, the former condition being correlated on the whole 
with plain colouration, and the latter condition with strongly contrasting 
colours such as occur in the hostilis group: also, but of somewhat less 
value is the degree of hairiness of the hind legs, long silky hairs being 
most strikingly developed in 5. sericea, whilst such hairs are absent in 
Venator and allies but more or less strongly developed in the species of 
the hostilis group: finally, the proportions in length of the segments of 
the palps, and legs, compared together and with the breadth of the 
head-plate, and the degree of development of cylinder bristles on the 
lower surfaces of the palp are characters which may serve to distinguish 
the females of certain species from each other. But, having thus rele- 
gated a specimen to its natural group, it is usually impossible to proceed 
further. Dr Purcell has commented on the difficulty of distinguishing 
between the adult females of 5. Venator and S. lethalis, and I may add 
that the three species ferox, schonlandi and globicornis, though very 
markedly distinct in the male sex, are practically identical in females. 
In this paper I have therefore largely ignored the females when distin- 
guishing between species, and those species which have been founded 
only on female specimens by previous authors are now placed aside as 
incertae sedis : most of them can never be identified, except through the 
locality data. 
The systematics of the genus Solpuga is largely a study of variation 
in the male flagellum. This organ presents quite a bewildering series of 
forms throughout the genus, but usually shews a high degree of constancy 
in the same species. Many of its forms shew no obvious relationship to 
each other: indeed, to a large extent the variation appears to be quite 
erratic and discontinuous. Species which in structure are practically 
identical as females, and which are no doubt closely related, may never- 
theless have profoundly different flagella, as is the case in ferox , schon- 
landi and globicornis, the flagella of which so far as is known are not con- 
nected together, even indirectly through other species. It is obvious 
therefore that as a guide to genetic affinity the characters of this organ 
may be of very subordinate value. Discontinuous variation is noticeable 
in every section of the genus, but is accompanied by considerable varia- 
tion of a continuous type. Dr Purcell has remarked on the variation in 
length exhibited by the flagellum of 5. Venator in specimens from various 
localities ( 9 ). In this case, a fairly wide range is observed, and it might 
be possible to distinguish several races or local forms on differences in 
the length of the shaft, the northern varieties having a longer flagellum 
than southern forms. In such a continuous series, the elimination of the 
intermediate forms would result in the formation of quite distinct groups. 
However, I do not know of any pair of species which differ from each 
other only in the length of the flagellum, and as a matter of fact, the 
observed variation in length of the flagellum in most species is limited 
to a comparatively narrow range. Nevertheless, in some sections of the 
