A quantitative Study of Variation, Natural and Induced, etc. 197 
Pearson!) which states that the offspring inherits one half of its 
characteristics from its parents, one fourth from its grand parents, 
one eighth from its great-grand parents, etc., has had great influence 
in shaping the opinions of breeders and. investigators. Allowing this 
law to hold good, it would be possible ofcourse to shift the mean of 
a race from generation to generation by selection. The question is, 
were the experiments upon which this law was built up and is being 
supported of the kind from which general conclusions should be 
drawn? As we will see many of the men who formerly worked with 
this so called ,,statistical school“ are now among its strongest critics. 
With the publication of the ‚„Mutationstheorie“ by de Vries?) 
about the year Igoo new light began to be thrown upon the subject. 
De Vries believes that the shifting of a mean by selection can proceed 
only to a limited degree, and that when the selection is discontinued 
the mean will soon find its former position. He believes that selection 
alone cannot account for all of the facts of evolution, that experiments 
fail to show real accumulative effects, and he has gone further by 
showing experimentally that species may arise by sudden jumps or 
mutations. 
At the Svalof Experiment Station?) striking results have been 
obtained by isolating types and multiplying the strains thus produced. 
The work of this station is in direct confirmation of the mutation 
theory. 
Johannsen‘) published in 1903 the results of some careful and 
extensive experiments with beans and barley which brought out the 
fact that especial attention should be given to the analysis of the 
experimental data before attempting conclusions. In his selection 
experiments he found the greatest difference in the behavior of the 
material, whether “pure lines” or “populations’” were used. A pure 
line by definition is a self fertilized race which has come direct from 
a single seed. A population is a mixture of lines. His experiments 
duplicated the results of Galton where populations were used, but 
gave no results from selection in pure lines. Which is to say that 
the attempts to shift the mean of a pure line were unsuccessful. 
1) Pearson, Karl. Biometrica. 
2) Vries, Hugo de. Die Mutationstheorie. 1903. 
3) Vries, Hugo de. Plant Breeding. pp. 29—106. (1907. Open Court Chicago.) 
4) Johannsen, W. Uber Erblichkeit in Populationen und in reinen Linien. 
(1903. Verlag von Gustav Fischer in Jena.) 
