290 PA THOGENIC BACTERIA. the surface, especially when the cultivation is made by the method of Fernbach with a passing current of air. This mycodertna, which may appear quite regular when the flask is undisturbed, is so brittle that it at once falls to pieces if the flask be moved. Spronck has recently determined that the characteris- tics of the growth of the diphtheria bacillus in bouillon, as well as the amount of toxin-production, vary accord- ing to the amount of glucose in the bouillon. He divides the cultures into three types : Type A. The reaction of the bouillon becomes acid and remains acid, the acidity increasing. The bacilli accumulate at the bottom of the clear liquid. The toxin-production is meagre. Type J3. There is no change from alkalinity to acidity, but the original alkalinity of the bouillon steadily in- creases. The culture is very rich, the bottom of the flask shows a considerable sediment, the liquid is cloudy, and a delicate growth occupies the surface. The toxicity .is very great. Type C. In a few days the reaction of the culture "becomes acid, and then later on changes to alkaline. During the acid period the liquid is clear, with a white surface-growth. When the alkalinity returns the bouillon clouds and the surface-growth increases in thickness. Sediment accumulates at the bottom of the flask. The toxicity of the culture is much less than in Type B. Spronck regards the varying reaction as due to the fermentation of the glucose, and asserts that the most luxuriant and toxic cultures are those in which no glucose is present. To exclude as much of the undesir- able sugar as possible, he makes the bouillon from the stalest meat obtainable, preferring it when just about to putrefy. Of the meats experimented with, beef was found to be the best. In large cities meat is ordinarily kept sufficiently long before being offered for sale to meet Spronck's require- ment..