Since Koch first announced his discovery many facts having a bearing on the subject have been brought to light, some in favour of his views, some apparently militating against them.
1. It has been found that there are several bacilli with morphological and cultural characters closely resembling those of the cholera comma, notably the Finkler-Prior bacillus of cholera nostras, Lewis's saliva comma bacillus, many of the commashaped bacilli discovered by Cunningham, and certain species found in river-water. Koch and others maintain that, though morphologically similar, as these various bacilli behave so differently from that of cholera in culture media, they must be considered as biologically specifically distinct from the latter.
2. Cultures of pure bacilli have many times been swallowed by way of experiment; yet, although in some instances diarrhœa with comma bacilli in the stools has resulted, in only a very few instances has true cholera been produced. On this account it is held by some that the comma bacillus cannot be regarded as the germ of cholera. Against this it is advanced that other factors must be present to ensure the induction of cholera by such experiments; for example, a suitable and peculiar condition of the body, possibly, as Buchner suggests, some second and as yet unknown micro-organism. Buchner accordingly regards cholera as the result of a mixed infection. It is to be presumed, therefore, that in the two or three instances in which cholera followed on the intentional or accidental ingestion of cultures of the cholera vibrio, these secondary but essential conditions were present. It must be remembered also, in assessing the value of negative feeding experiments, that the cholera vibrio, like other pathogenic bacteria, may lose through cultivation, or otherwise, its virulence while retaining its morphological and cultural qualities.
3. A few cases of what, from a clinical point of view, appears to be true cholera have been observed in which the most careful and prolonged bacterio-logical examinations failed to detect the comma bacil-
