514 MITZMAIN. 



flies assemble when not feeding, and the light reactions of the 

 fly are taken advantage of in withdrawing and introducing them. 

 If desired, ether can be used to advantage in the transfer of 

 animals. It should be applied at the screened end of the jar, 

 lightly enough to prevent flight, but not sufficient to stupefy 

 the insects. 



Tubes containing single flies were also applied to guinea pigs 

 strapped to a stock. The fly was usually placed on some con- 

 venient part of the body, preferably on the side of the abdomen. 

 (Plate V.) 



Horses in sheds. -r— Horses for experimentation were kept in a 

 shed screened from flies, a method commonly employed by in- 

 vestigators. Here it is not possible to make close and accurate 

 examinations. Despite the fact that many thousands of flies 

 could be applied at once, they did not, in my experience, live 

 longer than eight days, usually dying in five days even when 

 food was constantly available. 



Dutton, Todd, and Hanington(34) made attempts to keep tsetse flies 

 alive longer by more nearly reproducing their natural habitat in their 

 cages. In a cubical gauze cage, 18 inches along the side, containing water 

 and growing grass, guinea pigs and rats were used. Flies were found 

 to feed much better when animals were immobilized in cages. Smaller 

 cages than the above were found more advantageous for purposes of closer 

 observations. 



Horses on operating table. — In this method, the obvious ad- 

 vantage is in obtaining accurate data in feeding operations, and 

 timing can be done when desirable. This method supplanted 

 the cruder one of throwing the horse to the ground and feeding 

 flies from inverted bottles. The violent struggling of the un- 

 willing host is not favorable for obtaining accurate results. 



In all of the methods attempted with the various animals, 

 the hair was closely cropped with scissors. To avoid abrasions 

 a razor was never employed. If the skin were broken in this 

 manner and a positive result obtained, obviously it might in- 

 validate the conclusions to be drawn from the experiment. Con- 

 tamination might be produced under these conditions by the 

 pulvilli of the feet, and possibly, though remotely, by the body 

 hairs and wings of the insects. It was found advantageous to 

 slightly dampen the skin of the host to make the animal odor more 

 attractive to the fly and arousing its blood-dravdng instincts. 



In conclusion, it must constantly be borne in mind that all 

 of these artificial accessories in methods may possibly jeopardize 

 valid results by increasing the opportunities for contaminative 

 infections. 



