Appendix E 401 



will consist of quantitative measurements of characters to be 

 studied. These data are later analyzed, certain constants are 

 derived therefrom, and, lastly, the constants are interpreted. 

 The conclusions of the breeder or the investigator are based on 

 his interpretations of these constants. The meaning of the 

 various constants is explained in Chapter IV. 



In collecting data, it is important that as large a proportion 

 as possible of the entire population should be measured. Fail- 

 ing this, the sample should be fairly representative of the whole. 

 The time or season during which measurements are taken is 

 important where populations are to be compared. It would 

 obviously be unfair to collect data one year on fully matured 

 plants and another year on immatured plants. It is not always 

 easy to avoid a selection, conscious or unconscious, but the 

 collector should try to take his data with absolute impartiality. 

 He should collect at random until he has obtained a represent- 

 ative sample. Much time and labor will be saved if he can 

 conveniently limit the number of individuals measured to a num- 

 ber whose square root is an integer. 



The frequency distribution. Having measured a representa- 

 tive sample of the entire population, the next step is to sort the 

 data. All individuals of the same or nearly the same size are 

 grouped together in one order of magnitude. In order to give 

 a clear understanding of what follows, let us take, for example, 

 the data collected by a class of students on 500 bean plants. 

 The individual lengths range from 5 cm. to 95 cm. This is 

 known as the range of variability and the way in which the in- 

 dividuals are distributed along the successive equal intervals in 

 this range is spoken of as the frequency distribution of the vary- 

 ing character. For convenience, these lengths may be grouped 

 into classes, thus: 5-14; 15-24; 25-34 . . . 85-94. 



It is desirable that the number of classes be limited to not 

 more than about a dozen, and thus the size of the class will 

 depend upon the nature of the material. For example, bean 

 2o 



