MISCELLANEOUS 439 



number of samples larger spacings would give more information about the 

 area. 



Another example of sampling technique is discussed in a paper ^- on 

 the determination of arsenic spray residues on apples which concludes 

 that the sampling variation of the apples was in excess of the analytical 

 errors of the Gutzeit analytical method. 



The carrj^ing over of the identical experimental designs used in field 

 studies of crops to experiments in pathology is reported in a series of 

 papers. ^'*' ^^' ^^' '"*' A method for measuring the concentration of mosaic 

 virus makes use of the appearance of local lesions on Nicotiana glutinosa L. 

 leaves. In a given batch of plants the top leaves form a class which has an 

 average susceptibility which is different from the average for the leaves in 

 the second (or some other) position on the plant. Cutting across these 

 classes, the leaves on a given plant form a family such that all the leaves on 

 it are more (or less) susceptible than leaves in corresponding positions on 

 some other plant. Thus the whole assembly of leaves can be divided into 

 regions or parts in which a certain degree of homogeneity of response 

 exists; by planning the experiments with this in mind a three-fold increase 

 in the precision of comparisons w^as attained. ^^ 



A statistical study ^°° of the existing literature on the relationship 

 between \'irus concentration and the number of lesions produced on the 

 leaves led to the formulation of a mathematical relationship between these 

 quantities. Further experiments ^^ undertaken with a view to getting 

 improved data to test this equation led to a novel experimental design — 

 a Latin Square ^\'ith missing rows.^'* This arrangement (also called Youden 

 Square) was found especially adapted to greenhouse trials/^ since fewer 

 replications are required than in the Latin Square design. 



Two other papers illustrate applications of statistics. A statistical 

 analysis ^^ of a considerable body of data obtained in seed germination 

 tests showed how the Chi-square test can be used to check the sources of 

 variation in seed-testing methods. In connection with, studies of the effect 

 of SO2 on plants, a record of the SO2 concentration in the atmosphere was 

 available at half-hour intervals for a year. The concentration of the SO2 

 showed wide fluctuations,^^ with pronounced weekly and annual cycles. 

 By means of a linear regression technique the SO2 concentration was found 

 to be related to the wind direction and velocity, the temperature, and the 

 rainfall. 



Literature Cited 



1. Allen, R. C, "Influence of aluminum on the flower color of Hydrangea macrophylla 



DC," C. B. T. I., 13 : 221-242 (1943). 



2. Appleman, C. O., "Special growth-promoting substances and correlation," Science, 



48 : 31&-320 (1918). 



3. Backus, M. P., "Initiation of the ascocarp and associated phenomena in Coccomyces 



hiemalis," C. B. T. /., 6 : 33^379 (1934). 



