290 



J. A\-. tiCOTT MACFIK 



were growing. No green growth appeared during my experiments. The larvae 

 Uved in this vessel a long time, often over twenty days, but in none of my experiments 

 did a larva survive so long as 53 days, as one of Lima's is said to have done (Experi- 

 ment XVI), although I do not think there was any reason why this should not have 

 happened. The larvae with which H experimented almost always outlived the 



Fig. 1. Apparatus lor koepiug mosquito hirvae entirely submeriicd 



in flowing water. 



normal period for this state, and this being so, it seemed only a matter of chance how 

 long they survived. In the case of larvae kept in jars and allowed free access to the 

 air, the majority pass through this stage of development in a given period, but some 

 having reached the fourth phase do not proceed to pupate. Such laivae live a very 

 variable time, mai\y die soon, but others survive a long time, and, as 1 have recorded 



