FAMILY TABANIDAE: BIONOMICS 295 



the flap of skin near the navel. T. bicallosns seldom settles on the 

 lower parts of the body but nearly always on the sides of the neck or 

 abdomen. T. ditaeniatus is most often seen biting on the udder, some- 

 times however, it will settle on the side of the neck. Chrysops dispar 

 generally chooses the inner sides of the fore legs, and Haematopota the 

 hump or the neck. 



If one of these flies is watched closely as it feeds, it will be noted that 

 as it becomes replete with blood it passes out a dark fluid if it has fed 

 recently which drops on to the hair of the host. Next a clear fluid 

 is voided, and last of all apparently unaltered blood. All this fluid 

 collects in droplets on the hair below the fly. The moment it withdraws 

 its proboscis one or two large drops of blood trickle out of the wound and 

 mix with the excreta. It is not unusual to see two or even three flies 

 feeding at the same time close together. These habits probably explain 

 how tabanids become infected with their natural flagellates (Crithidia 

 and Herpetomonas). A fly which sucks blood at the spot where 

 the excreta of another has been deposited must necessarily contam- 

 inate its labella when they are pressed against the skin just before 

 the wound is made. Immense numbers of flagellates are passed out in 

 the fluid excreta of an infected fly, and it is not difficult to understand 

 how r another fly may suck them up. For the description of the -method 

 of feeding, see page 27. 



On leaving the host the flies usually settle on the under surfaces of 

 leaves, the barks of trees, stones, or the walls of an adjacent building ; 

 they remain here for several hours, and later collect in the vicinity of 

 water. Judging from experiments carried out with tabanids kept in cap- 

 tivity, they appear to feed every three days. It is not definitely known 

 whether copulation takes place before the first feed of blood, but it 

 w r ould appear that fertilization is in some way connected with the subse- 

 quent food of the female. No satisfactory explanation can be offered as 

 to why bred females can hardly ever be induced to suck blood, even when 

 they are placed in a large enclosure along with cattle. It is interesting 

 to note Neave's observations on the habits of these flies when the males 

 predominate. He says, ' I think it very probable that female Tabanidae 

 ' will be found to feed on blood only during a certain period of their imago 

 ' stage. This period seems to be subsequent to pairing and to the death 

 ' of the male individuals, and most probably (though I have no actual 

 ' evidence of this) before oviposition '. Although numbers of both sexes 

 of T. albimediits and T. striatus have been kept in large cages by the 

 writers, under as near as possible natural conditions, the act of copulation 



