348 PA THOGENIC BA CTERIA. media, and cease to be pathogenic after a few days. In his experiments with antipnetimococcic serum Washbourn found, however, that a pneumococcus isolated from pneu- monia sputum and passed through one mouse and nine rabbits developed a permanent virulence when kept on agar-agar made carefully, so that it was not heated beyond 100° C., and alkalinized 4 c.cm. of normal caustic soda solution beyond the neutral point determined with rosalic acid, to each liter. The agar-agar is first streaked with sterile rabbit's blood, then inoculated. The cultures are FIG. 99.—Diplococcus pneumonise: colony twenty-four hours old upon gelatin; x 100 (Frankel and Pfeiffer). kept at 37.5° C. Not only is this true, but ordinarily they seem to be unable to accommodate themselves to a purely saprophytic life, and unless continually trans- planted to new media die in a week or two, sometimes sooner. Kinyoun recommended to the writer that virulence could be retained for a considerable time by keeping blood from an infected rabbit, in a hermetically sealed glass tube, on ice. This plan seems to work admirably if the blood is not kept too long.