R. I. Pocock—Carboniferous Arachnida, 507 
without order, amongst other foreign bodies, over the exposed surface 
of the ventral side of the opisthosoma. Lach piece consists of three 
or four segments, but they differ from each other so greatly in length 
and thickness that no successful attempt can be made to pair or 
homologize them. Moreover, no clue is afforded as to their original 
situation, and it is permissible to suggest that perhaps after all they 
are not appendicular in nature but fragments of the sternal skeleton ; 
or indeed they may not belong to the organism at all like the so-called 
parasitic Spiroglyphs with which the fossil is strewn. And if appeal 
be made to fig. 4 on the same plate, representing the ventral side 
of another specimen, referred to the same species, and showing what 
might be interpreted as a pair of stout two-jointed limbs diverging 
from each other and from the middle line a little in front of the centre 
of the lower surface of the opisthosoma, it may be replied that, apart 
altogether from being in contact in the middle line, these segments 
bear no sort of resemblance to the restored opisthosomatic limbs 
depicted by Fritsch. I venture to suggest that, if appendicular in 
nature, the segments in question belong to one of the prosomatic 
appendages misplaced. But the angle they form coincides suspiciously 
with the angular curvature of the sterna of the mid region of. the 
opisthosoma in Anthracomartus. In the specimens of the latter genus 
that I have seen, as well as in Ammon’s drawing of A. palatinus, the 
borders of the sterna show up as angular ridges, the anterior and 
usually the strongest lying near the middle of the opisthosoma. 
With respect to the eyes, two pairs are represented in the restoration 
of Promygale bohemica; but the drawing of the original specimen 
shows only a single pair placed differently from either of the pairs 
outlined in the restoration. 
Since, then, critical examination of Fritsch’s drawings of Promygale 
fails to produce any trustworthy evidence that this genus can be 
separated from Anthracomartus, there is no choice but to regard the 
two as identical. 
As regards the Hemiphrynide, I have no doubt that these Arachnida 
must also be assigned to the Anthracomarti, since, so far as can be 
judged, they have the characters of that order. Most emphatically 
they do not belong to the order Aranee. The genus Hemiphrynus 
was based upon two species—JZ. longipes and H. hofmanni; but since 
these cannot be congeneric, if the restorations approach reality, 
I propose to select H. hofmanni as the type. 
The following features in Fritsch’s restoration of Anthracomartus 
call for comment (text-fig. 41, a dorsal, B ventral side). In the 
figure representing the dorsal side, the palpi (appendages of the 
second pair) show six segments with a terminal knob, projecting 
beneath the fore border of the carapace. The subequality in the 
length of the segments is quite unusual in the Arachnida, and if 
the terminal knob be a segment there is one segment in excess 
of the normal found in the class. The following four pairs of 
appendages consist of eight segments, including the coxe. This 
number also exceeds by one the normal found in Arachnida, the 
additional segment arising from the division of the femur. But 
in the first two pairs of these appendages the proximal segment 
& 
