132 



The Journal of Heredity 



When llic individual i)lanl or ear is 

 made the startinj^ i)oinL, wc have what 

 is known as individual seleetion. In 

 this case the selection commences with 

 a number of superior plants of a ^iven 

 variety, and the seeds from each plant 

 or ear are separately planted, and ke]3t 

 under continual observation. This en- 

 ables a strict com])arison to be made of 

 the jjroj^eny of each selection, so that in 

 a few years the best strain in the ori^^inal 

 selections may be determined and niulli- 

 plicd for future use. 



Nilsson, at Svalof, after subjecting 

 the older methods of mass selection to a 

 critical examination, decided to adopt 

 the method of single-plant selection 

 used by Shireff and Lc Coutcur, and has 

 achieved a considerable amount of 

 success. The method of ])rocedure has 

 already been described. x\s a ]mnciple, 

 it is based on De Vrics' conception of 

 the origin of species, and it assumes that 

 repeated selection is unnecessary. 



Another example of individual selec- 

 tion is afforded by the method intro- 

 duced by Willct M. Hays, founder of the 

 American Genetic Association. His 

 method consists in isolating the most 

 promising types of ])lants in a crop, 

 and of testing the efficiency of the selec- 

 tion by comjjaring the prolificacy of the 

 100 ]jlants derived from each of the 

 strains so isolated. 



For this purpose the produce from 

 individual plants are sown in "cent- 

 gener" plots. One hundred and forty- 

 four seeds of each selection are sown in 

 a square, with 12 seeds along each side. 

 At harvest-time the outside lx>rder row 

 is removed, and the remaining 100 

 plants are harvested, and the total 

 produce obtained is taken as a measure 

 of the prolificacy of a given strain. 



Hays made a close study of variation 

 in wheat, and found that those char- 

 acters such as yield, which can be 

 expressed in numbers, follow what is 

 known as Quetelet's Law of Varial)ility. 



A simple illustration of Quetelet's 

 law may be obtained by com])aring the 

 measurement of the height of 1 ,000 men 

 of the same nationality. If 1. ()()() men 

 be selected at random and arrangefl in a 

 row in order ai height, it will lie found^ 



(a) the man in the middle of the line 



represents the average height 

 of all men; 



(b) a line drawn over their heads will 



diverge only very slightly from 

 the horizontal throughout its 

 entire length, falling gradually 

 towards the end where the 

 smaller men are placed ; 



(c) the line will rapidly curve u]> 



ward near the Uj^i^er end of 



the line where the tall men are 



placed, and will curve rapidly 



downward at the lower end 



where the shorter men are 



standing. 



Hays" points out that if the individual 



yields of a large number of wheat plants 



of any given variety are arranged in 



order, a jirecisch^ similar curve may be 



oljtained to that illustrated above. The 



great majority of the plants give only 



an average yield, a few give a very poor 



yield, and a few give an exceedingly 



high yield. 



These latter arc the jjlants which he 

 uses for his future selections. 



Hays states that in "each 1,000 plants 

 of wheat there arc a few phenomenal 

 yielders, and the method of single seed 

 planting makes it jiracticable to secure 

 these exceptional plants, and from these 

 new varieties can be made." 



H.VYS' PRODUCTION'. 



Working on Fife and Bluestem varie- 

 ties of wheat, which were largely grown 

 in Minnesota, he succeeded in producing 

 im])rovcd strains which gave yields of 

 15-20% more than the original types, 

 and which have largely displaced them 

 from general cultivation. Thus Minne- 

 .sota 169 wheat was bred by a ])rocess of 

 selection from a Bluestem variety com- 

 monly grown in Minnesota. During 

 f(jur consecutive years it averaged in 

 field trials 4.9 bushels more than the 

 ])arent tyj^e. In 1902 it was distributed 

 in four-bushel lots to 375 fanners, and 

 re])orts showed that its average yield in 

 1903 was 21.5 bushels, as comi:)ared 

 with 18.2 bushels average for the com- 

 mon varieties, i. c, an increase of 3.3 

 bushels, or 18%. Hays judges the 

 eflicaey of a given selection, not by 



" Bulletin 62, Minnusola Rxpcrinu-nt Station, U. S. A. 



