S U M M A R YSeeds of white-seeded varieties of dwarf beans infected with Pseudomonas phaseolicola may show fluorescent areas in the testa under ultra-violet light. Fluorescence is not visible in seed with dark-coloured testas and is masked by coloured seed dressings. Bacterial infection is not the only cause of fluorescence in testas and the type and colour of the fluorescence cannot be used to diagnose infection. Of 542 fluorescing seeds examined I IZ were infected by the pathogen, and only six infected seeds were found in 495 non-fluorescing seeds from the same samples. It is estimated that in two seed stocks used for these experiments 64 and 68 yo of infected seeds were of the fluorescent type.A 34 Ib sample of seedstock has to be taken for examination to detect, with 95 yo confidence, ten infected seeds in I cwt. The selection of all fluorescing seeds from such large samples effectively concentrates the infected seed into a sampIe small enough to be tested by bacteriological methods.Test seeds are examined bacteriologically by enrichment-culture in nutrient broth. Ps. phaseolicola is identified in these cultures from the diagnostic symptoms of halo-blight disease induced in bean seedlings inoculated with aliquots of the cultures.
I N T R O D U C T I O NSerious outbreaks of halo-blight disease, originating from infected seedstocks, occurred in dwarf bean crops grown in southern England during 1964 (Wharton, 1966). From the field observations it was estimated that contamination of seedstocks might prove serious if it exceeded ten diseased seeds per cwt. One method of controlling the disease is to ensure that seed drilled is not heavily infected and for this a suitable test is required. The distribution of infected seed in a stock is assumed to follow a Poisson distribution. To detect infected seeds in I cwt with IOO ( I-P) % confidence it is necessary to examine a total sample of W = -IIZ;, log, p lb, where r = the mean number of infected seeds per cwt. T o detect ten infected seeds in I cwt of seedstock a sample of 34 Ib should, with 95 yo confidence, contain infected seeds. Katznelson & Sutton (1951) and Katznelson, Sutton & Bayley (1954) used a bacteriophage-multiplication technique to examine dwarf bean seed for seed infected by Pseudomonas phaseolicola (Burkholder) Dowson and Xanthomonas phaseoli (Smith) Dowson, the bacteria causing halo-blight and common blight diseases. Guthrie, 306 A. L. WHARTON Huber & Fenwick (1965) detected seed infected with Ps. phaseolicola by serological methods. Fenwick (1965) subsequently prepared a standard procedure for the serological detection of the halo-blight pathogen in bean seed by a slide agglutination method, for use by American seed firms.