112 The Control of Mosquitoes 



near Corozal, the previous northward flight al- 

 ready referred to (i.e. against the wind) was 

 established beyond a doubt. One of the many dif- 

 ficulties in this work at first was due to seasonal 

 changes. 



From time to time the supply of mosquitoes 

 gave out and we had to transport larvas and pupas 

 across the Isthmus ; many died of the rough han- 

 dling and shaking unavoidable with inexperienced 

 collectors, and when many larvae must be collected 

 quickly. 



At Gatun, where staining methods were again 

 used to check the flight already observed, we 

 collected mosquitoes in tents near the breeding 

 places, and also tried small paper houses. The 

 work was tedious, and it was difficult to move 

 about in such small places, or to keep quiet while 

 being bitten by hundreds of mosquitoes and still 

 more sand flies. This method of collection was 

 replaced by using mosquito bars. They were hung 

 up in the woods and the lower edges pinned up and 

 kept a few inches from the ground. It was soon 

 found that by occasionally brushing the mos- 

 quitoes from the hands, face, and clothing, they 

 would fly upwards and eventually come to rest 

 in the upper part of the net. Sometimes they 

 were but half an inch apart. This method was so 



