TAXONOMY OF MUSCOIDEAN FLIES — TOWNSEND 53 



other than by its mouthparts. It can not cling during the act* of 

 piercing by any structure except its mouth-hooklets. The acquire- 

 ment of such a habit has been possible through the fact that the na- 

 tives of the region inhabited by it have from time immemorial slept 

 on mats spread upon the earthen floors of their dwellings. The 

 larvae probably originally fed on fermenting juices and liquids, as 

 evidenced by the fact that they are especially common beneath the 

 urine-stained mats which have been occupied by sleeping children. 

 The flies are attracted by sour-smelling liquids, and doubtless ovi- 

 posit beneath the sleeping-mats of young children. 



The peculiar mode of reproduction of Glossina is carried even 

 farther by the Hippoboscid genera of mammal ectoparasites (Lipop- 

 tena, Melophagus, Hippobosca, Ortholfersia). The larva in these 

 forms is retained and nourished within the oviduct of the female 

 until full grown, but upon being extruded is incapable of movement. 

 The Glossina larva upon extrusion is capable of only sufficient move- 

 ment to find a suitable place for pupation, whereupon its integument 

 undergoes chitinization to form the pupal envelope. The Hippo- 

 boscid larva upon extrusion forthwith undergoes this process of ex- 

 ternal chitinization. The Hippoboscid female therefore extrudes 

 the larva in a situation and position suitable for it to remain during 

 its pupal period. It is thus evident that some relationship exists 

 between Glossina and the Hippoboscidse, doubtless to the extent of a 

 not very remote common origin. The Hippoboscidae are probably an 

 offshoot from the old muscid stock on the one hand, and the CEs- 

 tridae are likely an earlier offshoot in a quite opposite direction from 

 several stems of the same stock. 



The QEstrid habit of parasitism seems the oldest, the Glossina and 

 Hippoboscid habit next, while the Auchmeromyia mode is evidently 

 verv recent. 



