DISEASES TRANSMITTED BY THE COCKROACH 385 



pathogenic results from inoculations in this series. In another series of 

 iilatta which had fasted for 45 days and whose feces contained no 

 fluorescent bacilli and only a mildly pathogenic strain of Bacillus coli, he 

 fed the roaches with this organism. One strain derived from the feces of 

 a pigeon which had proven absolutely innocuous was fed to roaches for 

 five days. The feces of the Blatta collected aseptically by squeezing the 

 abdomen, killed a guinea pig in five days, with production of an abscess 

 at the site of inoculation. FVom the pus was isolated a fluorescent bacillus 

 which, when inoculated in pure culture subcutaneously, appeared patho- 

 genic and killed a guinea pig and a cony in four and five days respec- 

 tively, with production of a large purulence. Another strain isolated 

 from an infusion of putrifying flesh was fed to three roaches and after 

 eight days their feces were inoculated and killed a guinea pig in seven 

 da3's, with production of an abscess at the site of inoculation. The 

 fluorescent bacilli isolated from tlie feces were equally pathogenic. A 

 third strain obtained from the air did not acquire perceptible virulence in 

 passing through the roaches. A fourth strain isolated from earth in 

 wliich were living many Lumbricus, acquired in the Blatta a notable patho- 

 genicity. The feces contained germs which when isolated and inoculated 

 killed a guinea pig in 54 hours with subcutaneous edema and slight enlarge- 

 ment of the spleen, and exhibited its presence in tlie blood. He also con- 

 ducted a considerable series of experiments in feeding this organism with 

 other foods to the roaches, and demonstrated increased pathogenicity in 

 many cases after recovering it from the feces. 



Bacillus fluorescens nonliquef asciens Eisenberg and Krueger was iso- 

 lated by Cao in two series of experiments from the feces of Blatta 

 orientalis and when inoculated into a guinea pig caused its death in 48 

 hours without striking pathological symptoms, although the organism 

 may be recovered from the blood. It was not so virulent in the cony, caus- 

 ing death in eight days without purulence at the site of inoculation. When 

 cultures of tliis organism were fed to starved Blatta one strain recovered 

 from the feces of the guinea pig passed through the intestines of the 

 roach remaining in'ocuous, but a strain isolated from the earth had a 

 moderate patliogcnicity in the Blatta, producing abscess and death of 

 a guinea pig ; the inoculation of pure culture killing a guinea pig in four 

 days, with production of a large subcutaneous abscess. The germs were 

 recovered from the spleen and from the pus. Quite a series of experiments 

 were conducted with three strains of this bacillus, feeding them in con- 

 nection with other foods, and in a number of these experiments two of 

 the strains became moderately pathogenic. 



Bacillus megatherium Ravenel, a chromogenic organism found in soil, 

 was isolated by Cao from the feces of a Blatta orientalis in a single series 

 of experiments. In all of his experiments with this organism he did not 



