Nöller stated that rats were not infected with _T. lewisi_ by infective fleas biting them, but by the rats licking up the fæces passed by the fleas while feeding. This is not in agreement with Minchin and Thomson’s earlier views of regurgitation, which, apparently, they have now abandoned.[95] Wenyon (1912) confirms Nöller’s experiments. He took a dog flea, containing infective trypanosomes in its fæces, and allowed it to feed on a clean rat. The fæces of the flea, passed while feeding, were carefully “collected on a cover glass and taken up in culture fluid with a fine glass pipette.” The contents of the pipette were discharged into the mouth of a second clean rat. Injury to the rat’s mouth was carefully avoided. The first rat, on which the infective flea was fed, did not become infected, while the second rat, in whose mouth infective flea fæces were placed, became infected in six days. [95] Report to Advis. Comm. Trop. Dis. Research Fund for 1913, p. 74. When infective forms of _T. lewisi_ have been developed within the gut of a rat flea, they may enter and infect the vertebrate host by[96] (_a_) being crushed and eaten by the rodent; (_b_) the rat may lick its fur on which an infected flea has just passed infective excrement; or (_c_) the rat may lick, and infect with flea excrement, the wound produced by the bite of the flea. [96] Nuttall, _Parasitology_, v, p. 275. The time taken for the full development of _T. lewisi_ in the flea is about six days. The intracellular phase is at its height about the end of the first day; the crithidial phase, in the flea’s rectum, begins during the second day; the stumpy, infective trypanosomes are developed in the rectum about the end of the fifth day. Wenyon[97] writes that, “the fleas, when once infected with _T. lewisi_, remain infected for long periods, for though many small infective trypanosomes are washed out of the gut at each feed, those that remain behind multiply to re-establish the infection of the hind gut. Further, the infection is still maintained even if the flea is nourished on a human being, so that fresh human blood does not appear to be destructive to the infective forms in the flea.” [97] Report to Advis. Comm. Trop. Dis. Research Fund, October, 1912, p. 91. See also _Journ. Lond. Sch. Trop. Med._, ii, p. 119. The best method of controlling fleas during experiments is that due to Nöller. He adopted the method of showmen who exhibit performing fleas, and secure them on very fine silver wire. Of fleas fed on an infected rat only about 20 per cent. become infective. About 80 per cent. are immune. If fleas are examined twenty-four hours after feeding, trypanosomes will be found in all, so that many of the parasites are destined to degenerate.