Summary: The use of plate glass sheets carrying twenty-four cylinders in place of Petri dishes for the biological assay of streptomycin by the cylinder-plate method initially led to large assay errors. The substitution of cavities for cylinders, of a single layer of agar for two layers, one sterile and one inoculated, of the four-point assay design for ‘reading off the curve’, and of an optical projection method of measurement diminished the fiducial limits of error of an assay to 80–125% ( = 0·95).

Over the range of concentrations 0·1–100·0 units streptomycin/ml. the deviations from linearity of the graph of zone diameter against the logarithm of the concentration were very small.

Variations in thickness of the agar layer and in the intensity of inoculation had a marked effect on zone diameter.

An 8 × 8 Latin square lay-out reduced the internal error of an assay to 91–110% ( = 0·95), and with this increased accuracy a bias in the results was revealed arising from the time interval between filling successive cavities on the plate. This time effect was eliminated by the adoption of an appropriate order of filling, with a quasi-Latin square lay-out. The form of the assay finally adopted had fiducial limits of error, estimated both internally and externally, of about 95–105% ( = 0·95).


Article metrics loading...

Loading full text...

Full text loading...



  1. Finney D.J. 1944; Mathematics of biological assay. Nature; Lond: 153284
    [Google Scholar]
  2. Fisher R.A. 1942 The Design of Experiments, 4th ed.. Chapter v. Edinburgh: Oliver and Boyd;
    [Google Scholar]
  3. Yates F. 1937 The Design and Analysis of Factorial Experiments p. 35 Harpenden Imperial Bureau of Soil Science
    [Google Scholar]

Data & Media loading...

Most cited this month Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error