264 GLOSSINA PALPALIS [CH. 



ingest more than twice its own weight of blood. The excess 

 of fluid is got rid of very quickly, yellow liquid being excreted 

 from the anus. 



There is no evidence that G. palpalis can exist without 

 vertebrate blood, but it will readily feed on both warm- and 

 cold-blooded animals. It is not dependent upon any particular 

 species for its food, but the experiments of Kleine shew that 

 the flies thrive much better on mammals and birds than on 

 reptiles. In one case three batches of 232 female palpalis 

 were fed respectively on wethers, fowls, and crocodiles, between 

 July 8th and August 6th. During this period there died 

 no less than 190 of the flies that fed on crocodiles, whereas 

 the numbers of flies that died in the other two batches were 

 respectively 25 and 35. Moreover, those nourished on the 

 reptilian blood never laid any larvae, whilst the flies fed on 

 wethers gave birth to 82, and those on fowls to 89 larvae. 



Fig. 68. Glossina palpalis in the act of feeding. (After Roubaud.) 



From these results it is evident that crocodile blood is ;' 

 very unfavourable diet for G. palpalis, in spite of the large 

 numbers of flies that may be seen feeding on these reptiles. 

 Koch stated that on Lake Victoria the blood of the crocodile 

 formed the staple diet of the tsetse-flies, but the experiments 

 of Kleine, confirmed by Roubaud, shew that the fly must also 

 derive its food from some other source in order to be capabk 

 of breeding!. 



^ This is a very good instance of the danger of introducing prophylactic 

 measures without a careful examination of every possible factor that migli' 

 influence the results. After Koch's observation, numerous persons advocatd 

 the destruction of the crocodiles as a means of diminishing the number c 

 palpalis, whereas it is probable that the only effect of this host instead oi 

 increasing, is to reduce the numbers of flies, by adversely affecting their repro- 

 ductive powers. 



