The Biological Assay of Penicillin by a Modified Plate Method

Summary: The general form of assay for streptomycin described by Brownlee, Delves, Dorman, Green, Grenfell, Johnson & Smith (1948), based on an 8 × 8 quasi-Latin square layout of cavities on a large agar-covered glass plate, in place of Petri dishes, can be adapted to penicillin. A new design, using the principle of double confounding, allows the estimation by the usual four-point assay, of seven (instead of three) unknown solutions per plate of 64 holes. In routine use this assay gives fiducial limits (P = 0·95) of about ± 9·6%, including dilution errors.

The assay procedure previously described in this Journal (Brownlee et al. 1948) for streptomycin, is immediately adaptable to penicillin. For dealing with large numbers of fermenter samples in particular, we have found modifications desirable, which we describe in the present paper.