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



18 per cent content. Practically 100 per cent were killed in treated 

 kitchens b}' the use of a mixture containing 50 per cent of sodium 

 fluorid. 



2. Pyrethrum powder, pure, killed practically all of the roaches ' 

 within 48 hours in cage tests. Its effectiveness was greatly reduced 

 when slightly diluted. 



3. Borax, unadulterated, and in combination with inert matter 

 wherein there was less than 12 per cent borax, required from three 

 to seven days to kill all of the roaches in cage tests, proving it to 

 be very slow. Borax, used alone, was only partially effective in 

 kitchen tests. 



4. Phosphorus pastes were partially effective in cage tests and only 

 slightly effective in room tests. 



5. The various hydrocarbon-oil sprays, undiluted, killed from 80 

 to 100 per cent in treated cages. 



6. Coal-tar creosote emulsions, undiluted, killed all in cage tests, 

 but their effectiveness fell very rapidly when even slightly diluted 

 with water. 



7. Tobacco powders containing as high as 5.26 per cent of nicotine 

 were found to be ineffective in cage tests. 



8. The plaster of Paris and flour mixture was ineffective against 

 roaches. 



9. Alum, arsenious oxid, barium carbonate, barium fluorid, boracic 

 acid, calcium fluorid, copper sulphate, hellebore, mercury bichlorid, 

 paradichlorobenzene, sabadilla seeds, and sodium silicofluorid were 

 found to be more or less active against roaches. 



10. Thirty-eight miscellaneous materials were found to be ineffec- 

 tive. 



11. Sulphur fumigation, at the rate of 9 ounces of sulphur to 1,000 

 cubic feet, was effective. 



12. Nicotine fumigation, at the rate of 8 ounces of tobacco extract 

 (40 per cent nicotine, as sulphate) to 1,000 cubic feet, was ineffective. 



CLOTHES MOTHS. 



SPECIES USED IN EXPERIMENTS. 



There are two important species of clothes moths occurring in the 

 United States, the webbing or southern clothes moth {Tineola hisel- 

 liella Himi.) and the case-making clothes moth {Tinea pellionella 

 L.). The former species was used exclusively in the experimental 

 work on account of its abundance in Washington, D. C., its larger 

 size, and the fact that it readily leaves its case when disturbed. 



METHODS OP TESTING. 



Against the adults. — Many cage tests were made against the adults 

 in the laboratory, as follows: Twenty large, well-ventilated cages 



