EXPERIMENTS AGAINST COCKROACHES. 15 



sprayed cages before the oil has dried and that they are obliged to 

 remain on, or near, surfaces that are well wet with the oil. Under 

 room conditions many of the insects are hiding in cracks and 

 crevices, and often escape being hit by the oil at the time of applica- 

 tion. If they remain in hiding until the oil has dried, which may 

 be from one to several hours, they escape death, but those venturing 

 out before the oil has dried are killed by contact with it. Owing 

 to the chance that many may escape contact with the oil, such sprays, 

 while effective, are less .satisfactory than sodium fluorid. 



The addition of nitrobenzene or phenols does not appear to in- 

 crease the effectiveness of the mineral oil. 



"^ COAL-TAR CREOSOTE EMULSIONS. 



Tests were made with many coal-tar creosote emulsion prepara- 

 tions, undiluted, and in dilutions ranging from 1 part of the emul- 

 sion in 10 parts of water to 1 part of the emulsion in 100 parts of 

 water. When applied undiluted the emulsion killed all of the 

 roaches within 24 hours, but as the dilution was increased the ef- 

 fectiveness decreased rapidly, so that when the dilution was greater 

 than 1 part emulsion to 20 parts of water the material proved only 

 slightly effective, or ineffective. 



Both cage and kitchen test-s indicate that such emulsions should 

 not be diluted more than 1 part in from 20 to 25 parts of water. 



FUMIGATION AGAINST ROACHES. 



SULPHUR. 



Five tests were made with sulphur as a fumigant, the dosage 

 ranging from 4 ounces to 55 ounces per 1,000 cubic feet. The num- 

 ber of tests was so small that no definite conclusions as to the ef- 

 fective strength are drawn. The tests showed, however, that roaches 

 can be killed by sulphur fumigation, and indicate that the minimum 

 amount of sulphur to be used for 1,000 cubic feet (10' by 10' by 10') is 

 about 9 ounces. 



NICOTINE. 



Nicotine, when used as a fumigant, was found to be of no value 

 against roaches. A dosage of as much as 8 ounces of tobacco extract 

 (40 per cent nicotine, as sulphate) to 1,000 cubic feet failed to kill 

 them. A much greater strength than this would probably be effec- 

 tive, but the cost would be prohibitive. 



SUMMARY. 



1. Sodium fluorid was found to be the most rapid killer of roaches 

 of all the substances tested. Only 24 hours were required to kill 100 

 per cent in cage tests, even when the material was diluted down to 



