PATHOGENICITY 821 



efforts to step up the virulence of less pathogenic strains by serial passage 

 through kittens and dogs have always failed at the first transfer, but 

 Cleveland and Sanders (1930) have made some experiments bearing on 

 this point, except that they ascribe the effects observed to bacteria rather 

 than to amoebae. Using for the first passage in kittens a strain that had 

 been carried on in culture for from 460 to 540 days, they found that only 

 2 out of 26 animals became infected in the first passage, 5 out of 5 in 

 the second, 3 out of 7 in the third, and 2 out of 2 in the fourth. They 

 conclude: "An increase in the percentage of animals that became infected 

 with passage is demonstrated in these experiments, but this may be due, 

 as in the liver passages, to an increase in virulence by the bacteria rather 

 than the amoebae." The need for further work on the possibility of exalt- 

 ing the virulence of less pathogenic strains by animal passage is appar- 

 ent, but efforts along this line, in order to obtain results of significance, 

 will first have to eliminate the effects of bacteria accompanying the 

 Protozoa. 



The comments of Cleveland and Sanders regarding bacteria suggest 

 the next point, the effect of bacteria on pathogenicity of E. histolytica. 

 Their criterion for virulence was principally infectivity for the liver in 

 kittens, when inoculated with a hypodermic needle directly into this 

 organ. Pure cultures in liver-infusion, agar-horse serum saline medium 

 lost most of their ability to establish infection in the liver after a year 

 or more. Such a strain was reduced to an infectivity of 20 percent in the 

 first passage. The infectivity increased, however, with succeeding pas- 

 sages, until by the sixth passage it amounted to 73 percent. Was the ap- 

 parent increase in virulence to be attributed to the amoebae or to the 

 bacteria accompanying them? Which had lost virulence during the year 

 of life in the artificial media.-* 



An attempt to settle the issue was made in a crisscross experiment. 

 Bacteria from the fifth passage were inoculated with the culture amoeba 

 that had not been passed; and, conversely, the passed amoebae were 

 inoculated with bacteria that had not been passed, the latter being the 

 nonpathogenic Bacillus brevis. The experiments showed that fifth-pas- 

 sage bacteria increased the virulence of amoebae in culture for a year, 

 and that amoebae were not able to maintain themselves in the liver of the 

 cat unless accompanied by bacteria capable of damaging the liver. Thus 

 Cleveland and Sanders concluded that it was the bacteria accompanying 



