492 



Col. Sir D, Bruce and others. 



[June 15, 



the contents pass posterior to the Malpighian tubes, that is, into the 

 hind-gut, they at once become faecal in character, and turn into a dirty, 

 yellowish-brown material, which is microscopically composed of fine granules. 

 It is passed from the fly in this form. 



The figures given in the subjoined table refer to flies which were caught 

 at various parts of the Lake-shore, and were generally kept for about 24 hours 

 before dissection. The majority of the dissections were done during October, 

 November, and December, 1908. 



Table I. — The Contents of the Alimentary Canal of 220 Glossince palpalis. 



Total flies 

 examined. 



Intestine 

 empty. 



Intestine 

 contained ' 

 blood. 



i 



Character of Blood. 



Mammalian. 



Non- 

 mammalian. 



Non- 

 recognisable. 



220 



160 



60 



20 



9 



31 



Percentages.. , 



72 -7 



27 -3 



9 1 



V 



4-1 



14 -1 



J 



27-3 



Semarlrs. — In two cases blood corpuscles were recognised as being derived from monkeys, as 

 the characteristic parasites of monkey malaria were found in them. 



The second experiment to ascertain the food supply of Glossina i^alpalis 

 was modified as follows : — 



A journey was made to a small peninsula, hereafter referred to as 

 " Crocodile Point." This place was distant about two and a-half to three 

 hours by canoe from tlie Kibanga landing-place, and lay in a sheltered bay 

 far out of the beaten track of the canoes whicli come backwards and forwards 

 between the islands of Kome, Damba, and Ikivuma, and the weekly market 

 at Kibanga clearing. It was chosen, therefore, partly because it was 

 isolated from liuman influences, and also because of the large number of 

 flies whicli lived tliere and the number of crocodiles and birds which 

 frequented it. 



When it was first visited, a large crocodile was disturbed from where she 

 was lying outstretched on a spit of sand. The canoe-men at once set 

 to work, and unearthed 58 crocodile eggs lying in layers about 18 inches 

 below the surface, over whicli the " form " of the parent was clearly defined 

 in the soil. 



The peninsula was pointed in .slia])e, and not more tiian GO yards long, 

 and was closed on tlie land side l)y the doise wall of forest whicli fringes 



