Testing the Value of Fumigating Agents 



319 



of surface in the box as compared with that of a very large room. 

 The absorption and surface condensation are increased as the sur- 

 face increases, and a smaller room, of course, has relatively a larj^er 

 surface than a larger room of the same general shajn-. 



Hence, as a rule, the fact that mosquitoes are killed in this appa- 

 ratus with a certain amount of fumigant per unit space, under given 

 conditions, will hold for a larger room, under the same conditions. 



je-. 4'-o" ^ 



'o 



7" 



Sides 



Go5 



Fic. 9. 



Fig. 7. — Section of side showing one cage in position; the Cage projects into the lumen of the txii 

 and is closed on the outside by a small stopper. 



Fig. 8. — Sides of box. 



Fig. g. — One of the wire-gauze cages showing wooden stopper, with hole for introductma oi 

 mosquitoes closed by a wooden stopper. 



Conversely, and even more certainly, if a given concentration of 

 fumigant will not kill in the experimental box, in a large space the 

 fumigation, under the same conditions, will be unsuccessful. 



A very interesting and important point has come up in con- 

 nection with this work; namely, the rapidity with which the aciive 

 fumigating agent is evolved makes a great deal of dilTerence in the 

 efficiency of the fumigation. For example, with the less active 

 fumigants and culicides, as pyrethrum, the dilTerence Ixnwecn com- 

 plete success and complete failure may depend upon the ra|)idity 

 with which the culicidal substance is evolved as smoke. One j)ound 

 per 1,000 feet burned in three portions simultaneously is rather more 



