10 BULLETIN 707, U. S. DEPARTMENT OF AGRICULTURE. 



flour did not decrease the effectiveness of sodium fluorid, and it is 

 apparent also that the addition of a bait for the purpose of tempting 

 the roach to eat the poison is not necessary. The insects are killed 

 just as quickly through taking the fluorid into the system while 

 cleansing the appendages as though eating poisoned bait- 

 In addition to the foregoing room tests, many cage tests were 

 made in which sodium fluorid was mixed with lime or some other 

 inert substance in varying proportions, ranging from 1 to 100 per 

 cent. Under these conditions it was found that a mixture contain- 

 ing as low as 18 per cent of sodium fluorid killed all of the insects 

 within 24 hours. Such low percentages as 5, 3, and 1 per cent were 

 effective, but failed to kill 100 per cent, even when the tests were 

 continued for 7 days. 



Sodium fluorid proved the most effective of all the materials 

 tested. While it acts primarily as a stomach poison, it may act 

 also as a contact poison to a slight degree. 



PYRETHRUM POWDER. 



In tests of 22 different samples of pyrethrum powder (powdered 

 flowerheads of P. cinerariae folium and P. roseum) all were found 

 to be very effective against roaches treated in cages, and no prac- 

 tical difference in value was found between the powders made from 

 open, half-open, and closed flowers, or between the wild and culti- 

 vated flowers. Tests made during the winter of 1912 with pyrethrum 

 powder from the crops of 1908, 1909, and 1910 showed no practical 

 difference in their efficiency, thus indicating that this powder does 

 not deteriorate materially in four years. 



Pyrethrum stems, ground, were found to have no insecticidal value, 

 and their addition to pyrethrum powder is certain to reduce its 

 effectiveness. 



The general effect of pyrethrum powder on roaches is to paralyze 

 them and render them inactive almost at once. and. although slight 

 movements of the legs and antennae may continue for as long as 18 

 hours, a roach that has been thoroughly dusted will not recover. 



Pyrethrum powder is not nearly as active against roaches as is 

 sodium fluorid. Sodium fluorid may be diluted greatly and yet kill 

 quickly, whereas pyrethrum powder when mixed with only a small 

 percentage of inert matter kills very slowly. Therefore, in practice, it 

 appears necessary to use unadulterated pyrethrum powder for effec- 

 tive results. 



BORAX. 



Borax, when used under average room conditions, was found to 

 be partially effective, but can not be relied upon in cases of gross 

 infestation. 



