On Some Carboniferous Arthropods. 145 
General Remarks.—To illustrate the close affinity of the remains above 
described to recent forms, Pl]. iv. fig. 3 shows a drawing of Spherothervum 
grossum, Koch, the type species, natural size, after Koch, a form that inhabits 
South Africa. Fig. 4 is drawn, also natural size, from a specimen of a large 
species of Sphwrotheriwm from Madagascar, now in the Royal Scottish 
Museum, to show the method of rolling in these onisciform Millipedes. The 
terminal tergite completely hides the head and the first tergite under its 
hollow arch, and overlaps the anterior margin of the second tergite to come 
to rest against the swollen body of that tergite. It may also be seen that 
the turgid part of the second tergite, where it is carried down to the side 
lappets, forms a central knob round which the following segments arrange 
themselves, somewhat like the spokes of a wheel round the nave. The use 
of the knob seen on this part of the fossil form (Fig. 2) is thus explainable. 
The relative arrangement of the tergites and the side lappets in the succeeding 
tergites is similar in the fossil to that of the recent animal, thus leaving no 
doubt that these ancient creatures were adapted for rolling. 
To mark the contrast between these Millipedes and an Isopod Crustacean, 
which has adopted the same tactics of rolling for the protection of its more 
delicate organs, Fig. 5 has been drawn (magnified about three diameters) 
from a specimen of Armadillidium vulgare, or what is commonly known in 
Scotland as “the pill slater.’ The crustacean nature of the rolled animal 
is at once apparent. The sclerite bearing the eyes is exposed partially 
enveloped within that of the first free trunk segment, the two together 
simulating the second or nuchal tergite in Sphwrotheriwm and playing very 
much the same role as it. These are followed by the sclerites of six other 
trunk segments. Then comes a sudden change between the trunk and tail 
sclerites; five of the latter only are exposed, the first being either entirely 
hidden or aborted, and the second partially overlapped by the pleura or 
epimera of the last trunk segment. The body ends in three plates which 
simulate the end tergite of the Millepedes, but which is in fact a tail fan 
made up of a central telson flanked by the uropods of the sixth tail segment. 
Figs. 6 and 7 are diagrammatic representations of these tails seen end on. 
Fig. 6, the single plate, is that of a rolling Millepede like the fossil one, 
while Fig. 7 is that of Armadillidium. 
Though the method of rolling of Armadillidium is somewhat similar to 
that of the Millepedes, there is a considerable difference. There is no overlap 
of the epimera of the tail, or of the margin of the tail fan, over that of the 
opposing margins of the two segments which simulate the nuchal plate of 
the Millepede, so that, although the delicate antennular and antennary parts 
are covered up, the eyes are left unprotected. The epimera of the trunk 
VOL, XIX. K 
