The next experiments were with _Entamoeba coli_. In the 20 cases fed with material containing _Entamoeba coli_ there was a uniform failure to recover them culturally and in no instance was dysentery produced. Seventeen became parasitized as the result of a single feeding in from one to eleven days, the entamoebae being found in the stools and persisting in their appearance in the stools for extended periods. They concluded that _Entamoeba coli_ is an obligate parasite, nonpathogenic, and cannot be cultured. The third series of 20 feedings, carried on by Walker alone, was with _Entamoeba histolytica_. The material was mixed with powdered starch or magnesium oxide and given in gelatin capsules. In these experiments they obtained tetragena cysts in the stools of men fed only motile _Entamoeba histolytica_, and motile _Entamoeba histolytica_ in the stools of men who were fed only tetragena cysts and, finally, an alternation of motile _E. histolytica_ and tetragena cysts in the stools of a man having a recurrent attack of amoebic dysentery. =Results.=—Seventeen of the men became parasitized after the first feeding; 1 required three feedings, and 2, who did not become parasitized at the first feeding, were held as controls. The average time for parasitization was nine days. Only 4 of the 18 parasitized men developed dysentery, which came on after twenty, fifty-seven, eighty-seven, and ninety-five days, respectively, after the ingestion of the infecting material. In 4 cases fed with material from acute dysenteric stools or from amoebae-containing pus from liver abscess, and containing motile amoebae, there was no resulting dysentery, the 4 cases of experimental dysentery resulting from feeding of material from normal stools of carriers. As regards the cases which became parasitized, but did not develop dysentery, it is suggested that the amoebae live as commensals in the intestine of the host and only penetrate the intestinal mucosa and become tissue parasites when there occurs depression of the natural resistance of the host or as the result of some lesion of the intestine. That the pathogenic amoebae are more than harmless commensals, however, is shown by the fact that they alone, and not the nonpathogenic _Entamoeba coli_, are capable of penetrating a possibly damaged intestinal mucosa. =Epidemiology.=—The chief factor in the spread of amoebic dysentery would seem to be the encysted amoebae in the stools of convalescents, or symptomless carriers, rather than the motile amoebae in dysenteric stools. When such carrier has to do with the preparation of food, he becomes a particular source of danger.