24 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 134 
exposure. Wedberg et al. (1949) failed to recover S. typhosa from 
Llaberus cranufer, even though billions of the bacteria were fed 
repeatedly to these insects. Similar results were obtained by Janssen 
and Wedberg (1952) with Blattella germanica which failed to produce 
a single positive stool 24 hours after feeding on billions of S. typhosa; 
only 2 of 45 stools were found to be positive, and these were passed 
within the first 18 hours after the insects had fed on the pathogens. 
Jung and Shaffer (1952) concluded that, although P. americana ap- 
pears to have some mechanism for eliminating many hundreds of 
ingested Salmonella, it may harbor certain strains for at least a week 
if it eats feces containing at least several thousand of these micro- 
organisms. 
Not all Salmonella decreased rapidly in the cockroach, however. 
Various species listed in Appendix B survived in the guts or feces 
of cockroaches for periods of 18 to 42 days, and on the pronotum 
78 days (Olson and Rueger, 1950). Fecal pellets have remained 
infective for more than 4 years (Olson, p.c.; see p. 66). 
The results of experiments with the vibrios of Asiatic cholera have 
been variable possibly because of differences in techniques of feeding 
the pathogens to the test insects (Akkerman, 1933). Some workers 
have reported a relatively rapid decrease in the number of micro- 
organisms excreted (Toda, 1923). In Periplaneta americana most of 
the vibrios died in the foregut and Akkerman (1933) suggested that 
the degree of acidity that prevails in the gut may be harmful to the 
pathogens and have something to do with their depletion. Morischita 
and Tsuchimochi (1926) tested the sterilizing properties of a solution 
of feces from P. americana. The pH of the solution was about 6.7; 
it did not kill Vibrio comma within 3 hours. Jettmar (1935) claimed 
to have extracted a bacteriophage active against V. comma from the 
intestinal tract of Blattella germanica. 
Periplaneta americana and Leucophaea maderae are relatively little 
susceptible to plague bacilli inoculated directly into the body cavity of 
these insects (Barber, 1912). Plague bacilli rapidly died or lost their 
virulence in the intestine of Blattella germanica (Jettmar, 1927). 
Macfie (1922) failed to recover Neisseria gonorrhoeae Trevisan 
(cause of gonorrhea) from the feces of P. americana after the in- 
sects had fed on the pathogen. Brucella abortus (cause of undulant 
fever in man) does not remain alive in the intestinal tract of the 
American cockroach for more than 24 hours (Ruhland and Huddle- 
son, 1941). Other negative results in feeding experiments are listed 
in Appendix B. The reader is referred to Steinhaus (1946, 1949) for 
a general discussion of immunity in insects. 
