446 MEDICAL ENTOMOLOGY 



is thickest. There are two rows on the epimerum of the metathorax, 

 and two on the first abdominal tergite. The small eminences of the 

 ninth tergite in front of and below the clasper bear tufts of hairs. 

 As the identification of fleas depends to a large extent upon the 

 arrangement of the hairs every care should be taken to avoid detaching 

 them during the preliminary manipulation of the specimen. If this 

 does happen, the position of the hairs can be judged from the clear spaces 

 which remain at the points of insertion. 



The bristle (or bristles) placed at the dorsal angle of the seventh 

 segment is called the antipygidial bristle. The number of these is 

 important. There is one pair in the cat flea. The number of hairs 

 on the joints of the legs, and particularly the presence of tufts on the 

 femora and coxae, are also important. 



CLASSIFICATION 



The Siphonaptera fall naturally into three families, the Sarcopsyllidae, 

 Pulicidae, and Ceratopsyllidae, distinguished from one another both by 

 structure and habits. The first of these families includes the well 

 known 'Jigger' flea, and contains species with a tick-like habit of 

 fixing themselves to their host, a habit which is accompanied by many 

 interesting modifications of their external structure. The majority of 

 known fleas belong to the Pulicidae, of which Pulex irritdns, the human 

 flea, is the type. In the third family, Ceratopsyllidae, there are two 

 genera, Nycteridopsylla and Ischnopsyllus, which are exclusively parasitic 

 on bats ; in these fleas the head is of a peculiar shape, the face sloping 

 strongly forwards and being recurved just above the mouth, with two 

 flaps situated at the frontal oral corner. 



The number of new genera and species which have been described 

 within recent years is very great, and it is not possible, within the 

 space available, to give more than a brief account of the more im- 

 portant and interesting forms. In the following the arrangement of 

 the genera is taken mainly from the works of Jordan and Rothschild, 

 and from Baker's papers on the American Siphonaptera. 



FAMILY SARCOPSYLLIDAE 



Thoracic segments shortened, together shorter than the first abdom- 

 inal segments in the dorsal line. Genal edge of the head always 

 produced into a triangular process at the ventral oral angle. Rostrum 

 (labiutri) rather long but weak, consisting of two or three segments, 

 inclusive of the unpaired basal one.. . . ". . 



