REVISION OF THE SARCOPSYLLIDAE 31 
meral suture. This bristle is absent in Malacopsylla. The episternum also bears, as 
a rule, one or more bristles. In the genus Lycopsy/la, however, there are no bristles 
on this plate. Pulex irritans, lep oris, pallidas, etc., have a bristle on the sternum and 
one or more on the episternum. In the family Sarcopsyllidae there is one compara- 
tively small bristle on the sternum ot Hectopsylla and a minute one on the episternum. 
The species of the genus Dermatopbilus have only a minute bristle on the sternum, 
whilst there is no bristle at all on either the sternum or the episternum of the meta- 
thorax of Echidnophaga, the last genus being in this respect the most highly specialised 
ot the three. The bristles or the metathoracic epimerum resemble those of the genus 
Pulex in their arrangement. In the species with the largest number ot bristles on 
this epimerum there is a row of bristles extending from the stigma downwards to- 
wards the middle, a solitary bristle being sometimes placed further ventrad. There 
are apparently never more than six bristles placed in this position, which is the case 
in Hectopsylla Pulex. Occasionally one or two bristles are placed in front of the 
normal row, these being a vestige of the anterior row of bristles commonly found in 
the genus Pulex. In some species of the genus Echidnophaga the number of bristles 
is reduced to two. In contradistinction to the Sarcopsyllids the species ot the genera 
Vermipsylla and Chaetopsylla have the bristles of the thorax developed in great pro- 
fusion. 
The abdomen consists of the same number of segments as in other Sipbonaptera. 
In the and the non-pregnant it appears rather short, the segments being, as in 
Pulex irritans, strongly developed in a vertical direction. Apart from the extension 
ot the % abdomen, there are some other sexual differences. The $ is less convex 
dorsally than the though more so than is the case in the of other Sipbonaptera. 
The bristles are also not always alike in number and length in the two sexes, their 
number being usually less reduced in the £ than in the %. In the Sarcopsyllids there 
is a similarly great reduction in the bristles of the abdomen as exists in those of the 
thorax. Two rows of bristles to each tergite of segments one to seven of the 
abdomen may be considered normal in Sipbonaptera, though there are many species 
with more than two rows. These two rows are in the genus Pulex usually found at 
least on the first tergite, but the anterior row has disappeared in the %% of several 
allies of Pulex irritans, for instance in P. pallidus, as well as in both sexes ot the species 
of Malacopsylla ; the posterior row, however, is always well developed. Between 
each pair of long bristles of this row there is always a small hair in all Pulicidae. In 
the Sarcopsyllidae the anterior row of bristles is represented only on the first tergite, 
and only in the $$. The second row is most complete in the $ ot Hectopsylla psittaci, 
in which there are tour or even five fairly long bristles on each side of the tergites 
one to seven ; but this row differs from that of the Pulicidae in the small hairs in 
between the bristles being practically absent. From this row, which represents 
doubtless the ancestral type of abdominal vestiture of hairs in the Sarcopsyllids, to the 
