438 n. FRAENKEL-CONRAT 



typical sterile virus preparation to a concentration of 0.1 mg./ml. with 

 chloroform-saturated water, and storing this in a refrigerator for use over a 

 two-week period. With each set of unknowns, a fresh secondary dilution of 

 this stock solution in 0.1 M phosphate (pH 6.8) is tested. The dilution is 1-4 

 /xA to 2 ml. for N. glutinosa (depending on the season), and an additional 

 10-fold dilution for the Xanthi variety. The most generally used level of 0.01 

 jLtg. virus/ml. (in Xanthi) contains about 10^ virus particles per milliliter and 

 produces in average about 30 lesions per half leaf. Over the range of 10-50 

 lesions the dose-response curve appears to be nearly linear. 



The standard, the unknown solutions (5-15) at appropriate concentrations 

 in the same buffer, and the buffer alone are then appHed to equivalent half 

 leaves of 6-18 plants (3-6 leaves per plant, depending on size). Generally, 6 

 half leaves are used for each solution, and such assays are repeated at least 

 once at levels selected to approximate the lesion number of the standard. 

 For final, more exact evaluation, 12 half leaves are used, and, at times, 12 

 opposite half leaves are inoculated with unknown and standard. 



Prior to inoculation, the plants are prepared for assay by cutting off all 

 but the leaves which are to be used, and powdering these with carborundum. 

 The solutions (0.02-0.05 ml. per leaf) are applied to the leaf and spread 

 evenly over the entire half-leaf surface by means of a glass spatula with a 

 rough-ground rubbing surface. The plants are then rinsed with water. All 

 glassware, including the glass rod, is sterile. Nevertheless, there are occasional 

 sets of assays in which contamination is present, as indicated by lesions 

 appearing on the phosphate blank-inoculated leaves; if contamination is 

 appreciable (more than 2 lesions), such sets are discarded. 



The assay of nucleic acid preparations is performed by the same procedure 

 with one important exception. Since the nucleic acid is more sensitive than 

 the intact virus, particularly to salts, it is diluted in an ice bath with water of 

 0°C. to 90 % of the final volume. To each of these solutions is added at 0°C. 

 one-tenth volume of M phosphate (pH 6.8) in the greenhouse just prior to 

 apphcation to the plants. Neither lower concentrations of phosphate nor 

 other buffers investigated were found to give as many lesions as were obtained 

 under these conditions. 



Lesions on each half leaf are counted twice on the third to fifth day, and 

 the highest number used in evaluation. Obviously, the accuracy of such 

 assays is primarily dependent on the number of half leaves used in each group. 

 Comparison of a series of solutions tested in one set of plants is most reliable. 

 The significance of the differences of the group averages can be evaluated 

 statistically if a sufficient number of leaves is used and some preliminary 

 mathematical transformation is performed (Kleczkowski, 1949). The sug- 

 gestion of Hart and Perez-Mendez (1957) of excluding the extreme individual 

 cases and basing evaluation on the median of the remauiing group may bear 



