December 15, 1905.] 



SCIENCE. 



801 



bers 3,000 parts of atmosphere; the third 

 change, one part of liquid CS, to 6,000 parts 

 of atmosphere; the fourth change, one part of 

 liquid CS, to 12,000 parts of atmosphere, and' 

 so on. 



The greatest pains were taken to secure 

 volatilization of tne liquid. Nearly half an 

 hour was allowea to elapse after the drop was 

 enclosed before changing bottles, and no 

 change was made as long as the slightest evi- 

 dence of the liquid GS^ was observed. 



confusum) were treated, of which fifty-eight 

 were entered in tables as offering authentic 

 results. Experiments with fifty-two diptera 

 wer6 tabulated, as were also experimerits with 

 fifty-six hemiptera (mostly aphids). Forty 

 lepidopteran larvae were used and results 

 tabled, and thirty-one hymenoptera. A few 

 entries in the tables are here given as an 

 example of the method employed, and at the 

 same time offering an illustration of the some- 

 what remarkable vitality of the Aphidse. ' 



Now then, it is evident that if our calcula- 

 tions are correct, we could by placing it in an 

 empty vial, and then applying it as above 

 described to one of the vials containing a 

 mixture of gas and atmosphere, expose an 

 insect to any desired proportion of gas and 

 atmosphere. 



One hundred and fifty-six specimens of 

 coleoptera (largely composed of TriboUum 



Conclusions. — The experiments demonstrate 

 that it is very much easier to kill diptera than 

 coleoptera. It is further to be noted that 

 insects are apparently dead long before death 

 actually occurs. I have no doubt but that 

 there is an immediate effect upon the indi- 

 vidual insect coming in contact with the gas, 

 although it does not always show this. By 

 ' effect ' as noted in the tables, we mean the 



