If the microscopical examination indicates a bacillary infection we should take a small mass of the stool, wash it in sterile water and then drop it in a tube of sterile bouillon or salt solution. After emulsifying in this tube of bouillon we take up 2 or 3 loopfuls of the emulsion and deposit them on a poured plate, later smearing out with a glass rod, either by successive parallel strokes or by revolving the plate while smearing the surface with the glass rod. It is in the first two or three days of an attack of acute dysentery that we obtain the best cultural results, often noting a pure culture of dysentery bacilli from proper material taken at the onset. Manson-Bahr states that he has never recovered true dysentery bacilli from a purely faecal stool. Even faecal contamination of the mucoid mass makes it difficult to recover the organism. Dysentery bacilli rapidly die out if the stool is acid so that it has been recommended to make the stool strongly alkaline where it has to be sent to a laboratory from a distance. It has seemed to me that litmus lactose agar gives results more surely than the more restraining faeces-plating media. Still I generally use Endo’s fuchsin agar because it is always at hand for typhoid or paratyphoid culturing and gives good results. The dysentery bacillus colonies on this medium are like those of typhoid—grayish white. In England they prefer MacConkey’s neutral red bile salt agar while others use the Conradi-Drigalski medium. We are now using the Teague medium, which is taken up in the chapter on Faeces. On all these media the colonies resemble those of typhoid and the differentiation is most easily made by examining for motility. At the same time one not infrequently finds lack of motility in bacilli from colonies just isolated on Endo’s medium which later on in subculture show motility and are found to belong to the typhoid or paratyphoid group. For the sure determination of dysentery bacilli or for differentiating the Flexner and Shiga strains one should carry out agglutination tests. The isolation of dysentery bacilli from chronic cases or from convalescents is more difficult as a rule and agglutination tests may be more practical. A trouble is that an agglutinating effect may be connected with a prior infection. Although some observers have noted the appearance of agglutinins in the serum of cases of acute bacillary dysentery within three or four days from the onset of the disease, yet it is usual not to obtain agglutination with the patient’s serum before the tenth day. With the Shiga strains agglutinating power in 1 to 50 is usually accepted as evidence of specificity but for Flexner strains we generally have a higher titre so that a dilution of 1 to 150 should be required for the test.