99 
hours to the fumes of tobacco burned in a proportion of 2 pounds 
per 1,000 cubic feet. 
We found 1 pound of pyrethrum per 1,000 cubic feet, with an 
exposure of two hours, sufficient to kill Stegomyia fasciata , and, 
although many of them still showed signs of life after this exposure, 
none revived. Two pounds per 1,000 cubic feet, with an exposure 
of two hours, was sufficient to kill them outright. 
We made only one experiment with sulphur dioxid in order to 
kill some Stegomyia fasciata , which had escaped in the laboratory. 
We selected sulphur on account of our confidence in the insecticidal 
value of this substance. In this experiment we used about 3 per cent 
of the gas, generated by burning in open pots, with an exposure 
of two hours. The previous work of one of us° showed that 1 per 
cent, with an exposure of one hour, is quite sufficient in a small inclo¬ 
sure, but we used the excess on account of the peculiar construction 
of the house. 
When burning sulphur for its insecticidal effects care should be 
taken to keep the gas as dry as possible, while the contrary is neces¬ 
sary in order to obtain its germicidal action. 
The small jDercentages of sulphur dioxid kept dry are not sufficient 
during short exposures to bleach pigments or injure fabrics; and 
this, in our opinion, is the best of the poisonous gases to insure the 
destruction of such vermin in houses. Formaldehyde gas is utterly 
untrustworthy in this connection. It is not an insecticide. 
We made no experiments with hydrocyanic-acid gas. While rec¬ 
ognizing its great insecticidal power, we considered it entirely too 
poisonous a substance to use in the household or other inhabited 
places. In our opinion the use of this gas should be limited to green¬ 
houses, railroad cars, granaries, and other isolated and uninhabited 
places. 
SUMMARY AND CONCLUSIONS. 
The cause of yellow fever is not known. 
The Myxococcidium stegomyiae is not an animal parasite. Yeast 
cells sometimes simulate the coccidia in form and staining reaction. 
The infection of yellow fever is in the blood serum early in the dis¬ 
ease. Yo abnormal elements that bear a causal relation to the disease 
can be detected in the serum or in the corpuscles with the best lenses 
at our command. 
«Disinfection against mosquitoes with formaldehyde and sulphur dioxid, M. J. 
Rosenau. Bull. No. 6, Hyg. Lab., U. S. Pub. Health & Mar.-Hosp. Serv., Wash., 
Sept. 1901. 
