286 



SCIENCE. 



[N. S. Vol. XIV. No. 347. 



point, but one that is usually overlooked. 

 These pests are to be found in fewer or 

 greater numbers of both species and indi- 

 viduals in every dwelling, oflBce or store- 

 house, and no perfectly efficient means 

 either to prevent their gaining an entrance, 

 or to exterminate them when they are once 

 established, has as yet been devised. 



Eecent successful applications of hydro- 

 cyanic acid gas for the extermination of 

 insects infesting greenhouse plants have 

 suggested the use of the same remedy for 

 household pests. It is now no longer a 

 theory, but an established fact, that .10 

 gram of 98-per- cent. -pure cyanide of potas- 

 sium volatilized in a cubic foot of space 

 will, if allowed to remain for a period of 

 not less than three hours, kill all roaches 

 and similar insects. 



The experiments which led to this con- 

 clusion were made in a small building 

 which is used for laboratory purposes by 

 the Division of Botany. This structure 

 .has for some time been infested with sev- 

 eral insect pests, the more numerous and 

 troublesome being the common cockroach 

 (Periplaneta americand). The building con- 

 sists of one story and basement, the upper 

 part being rather loosely built, as it is 

 ceiled throughout with matched lumber. 

 This method of construction provides nu- 

 merous hiding places for the insects and 

 also renders fumigation difficult, by permit- 

 ting the gas to escape too quickly. Within 

 the building are several sources of mois- 

 ture, a rather high and constant tempera- 

 ture is maintained in some of the rooms, 

 and large quantities of seeds and sub- 

 stances that serve as food for insects are 

 stored, making conditions well adapted to 

 the development of cockroaches. 



During the early part of last year the 

 roaches became so numerous as to be a det- 

 riment to the work of the laboratory and it 

 was necessary to adopt some means of 

 checking them. On the evening of May 



10, 1900, the building was closed, and 

 after opening up the interior of the rooms 

 as much as possible the entire structure 

 was fumigated experimentally with about 

 .08 gram of 98-per-cent.-pure cyanide of 

 potassium per cubic foot of space. The 

 gas was allowed to remain during the 

 night, or until it gradually escaped. When 

 the rooms were entered the following morn- 

 ing there remained a perceptible odor of 

 the gas, but this soon disappeared after 

 opening the windows and doors. The 

 ledges and window sills were strewn with 

 dead houseflies and the floors bore abund- 

 ant evidence of the e£fect of the gas on 

 roaches. Not a single insect that showed 

 indications of remaining life was to be 

 found in the building. About a quart of 

 the flies and roaches was gathered up and 

 placed in a cage where they were allowed 

 to remain until the following day, when 

 two roaches showed signs of life by slow 

 movements ; these, however, could not 

 walk when placed upon their feet and sub- 

 sequently died. 



For some time after this fumigation no 

 roaches were to be found in the building, 

 but eventually the eggs that had been pre- 

 viously deposited hatched and developed, 

 adults were carried in from other buildings, 

 etc., until in March of the present year the 

 roaches had again become so numerous as 

 to be a nuisance and a detriment to the 

 work of the laboratory. The building was 

 again treated with cyanide gas, this time at 

 the rate of .10 gram per cubic foot of space, 

 but was allowed to remain only fifty min- 

 utes, when the windows were opened and 

 the gas permitted to escape. The roaches 

 were strewn over the floors and several mice 

 were found dead. A large number of the 

 roaches were again collected and kept in a 

 cage until the following day, when it was 

 found that fully ten per cent, of them had 

 not been killed and were as lively as before 

 treatment ; the mice, however, showed no 



