44 Al^IMAL LIFE AND HUMAN PROGRESS 



by the Danish botanist Johannsen, undertaken for scientific 

 purposes, and with a definite object in view. 



Johannsen worked with the common bean, which is self- 

 fertilising, and therefore the difficulties introduced by the 

 mingling of characters in biparental reproduction are elimin- 

 ated. Beans taken at random from a number of unselected 

 plants were weighed and the results gave a typical curve of 

 probability. Johannsen kept separate the seeds from each 

 plant and sowed them separately. The result was that in 

 the new generation groups were sorted out which had different 

 modes from that of the original population. Some of these 

 groups were similar, some different, and the differences con- 

 sisted in the average or " median " being different. In some 

 the " median " weight was higher, in some lower than that of 

 the original population, but in each group the seeds varied in 

 weight according to the law of probabihty. Seeds from each 

 group were sown again and there was no further breaking up 

 into groups. In each group, called by Johannsen a " hne," 

 the indiAaduals ranged themselves about the median proper 

 to the line, and this median was the same as in the first filial 

 generation. But in every line there was variation among the 

 individual seeds ; some were small, many medium, some large. 

 In further breeding it made no difference whether a small bean 

 or a large bean was chosen for seed, the progeny conformed 

 to the modal curve of the " line " and gave large, medium and 

 small beans, within the modal Umits. All of which means that 

 the original population was mixed, and consisted of a number 

 of pure lines, each having its own modality. When these pure 

 lines were sorted out, each showed a limited range of individual 

 variability and bred true to it. In other words, the germinal 

 constitution of each pxae line is fixed, and the individual 

 variations occurring within it are due to the varying incidence 

 of external conditions upon identical constitutions. There 

 can be no progression, and also no regression, in a pure line, 

 and selection is powerless to alter it. 



But in plants that are not self - fertihsing, and in 

 animals in general (for animals are mostly biparental), every 

 union between male and female means the mingling of two 



