334 
loped, its setæ being rather long and niore or loss stiongly cuived 
in the vertical plane. 
In the median line above the hind margin of the sternum is 
always found an organ which I name the anal plate (p); it pos- 
sesses two, four or six branches directed essentially backwards, and 
it varies exceedingly in shape, but it shows slight variation in spe- 
cimens of the same species. A species can often be recognized or 
separated only by a correct observation of its shape; its basal part 
is sometimes withdrawn above the sternum. Not only the anal 
plate but also the setæ of the anal segment must be examined with 
care in all species of the order. 
Legs. It is known that in the family Pauropodidæ all the 
legs with exception of the first and the last pairs consist of six 
joints. I name these joints: coxa, trochanter, femur, tibia, metatarsus 
and tarsus, without entering into morphological comparisons with 
the legs in other orders or classes. In the first and the last pairs 
a metatarsus is not marked off from the tarsus. Latzel writes on 
the whole order: „Am ersten Beinpaare zahlt man je fiinf Glieder, 
an den (ibrigen je sechs“, but this is erroneous; Kenyon observed 
the real number in the legs of the animals examined by himself 
(a species of Pauropus and Eurypauropus spinosus), but when he 
embodies these statements in the diagnosis of the order Pauropoda, 
it does not agree with facts seen by me. In Brachypauropus su¬ 
perhus (pi. VI, fig. 3 h) and in Eurypauropus Latzelii (?) (pi. VI, 
fig. 4 1) I have only found five joints in all pairs of legs, a meta¬ 
tarsus being absent. The result is that in two species hitherto re- 
ferred to Eurypauropus (and even considered as identical by Latzel 
and Kenyon), viz. E. spinosus Byder and E. Latzelii Cook (?), the 
legs show a sharp difference: in E. Latzelii (?) all pairs have no 
metatarsus, while this joint is present in all legs of E. spinosus 
with exception of the first and the ninth pairs. This difference 
must be of generic value in the family Eurypauropodidæ'. I am 
inclined to believe that the number of joints in the Pauropodidæ 
and in Brachy pauropodidæ is a character for the families, but 
