DEPARTMENT OF EXPERIMENTAL EVOLUTION. 117 



turned in for publication a study of a somatic selection of the color 

 ])attern of rats, extending over seven years, in the course of which 

 over 20,000 rats were raised and studied. The results, they believe, 

 demonstrate the efficacy of mass selection; and they are led to con- 

 clude that selection is, in animal breeding, a more important agent 

 than mutation, partly because it is controllable and its results 

 therefore more certain, and parth^ because it may even determine 

 the occurrence of mutations of a particular sort. 



Quantitative Studies of Selective Elimination, by J. A. Harris. 



During the year. Dr. Harris's studies on natural selection have 

 been carried forward along the lines laid down in previous reports. 



The results already published for differential mortality with respect 

 to seed weight in field cultures of Phaseolus vulgaris have been fully 

 confirmed and somewhat extended by planting about 46,000 indi- 

 vidually weighed seeds in the green-house. There is unquestionably 

 a reduction in variabihty in weight due to differential mortality. 

 The data also seem to show^ that in some strains the higher mortality 

 falls upon the heavier and in other strains upon the lighter seeds. 



The results for Phaseolus have been confirmed on an extensive 

 though smaller scale by field experiments on Pisum. Some progress 

 has been made in the phj^siological phases of these investigations in 

 that a correlation between weight of seed and time required for 

 germination has been demonstrated in Phaseolus. 



The selective eUmination studies for seedhng variants in Phaseolus 

 have been continued on about double the scale and in greater detail 

 than last year. The results can not be announced until the data 

 are secured from the crops still in the field. 



The studies of fertility and fecundity instituted to check the find- 

 ings for the differential mortality of ovaries have been carried for- 

 ward and in part published. So far as completed they confirm the 

 results announced in Year Book No. 9. 



Thus it was demxonstrated in papers pubHshed from the Station 

 in 1910 that the chances of development to maturity of the ovaries 

 formed bj^ a plant are conditioned upon the structure of these ovaries. 

 It was proved, for example, that ovaries of Siaphylea have a death- 

 rate roughly proportioned to the number of carpels which are bilat- 

 erally asymmetrical with respect to the arrangement of their ovules. 

 During the year results have been published showing that in Phaseo- 

 lus such bilaterally asymmetrical ovaries are not only less capable 

 of maturing their seeds, but that the seeds which they do produce 

 are lighter than those borne in symmetrical ovaries. Thus the 

 selective ehmination and the phj-siological studies fully check and 

 supplement each other. The information gained from them is now 

 furnishing the basis for the study of asymmetry in the individual. 



