204 GENETICS [Bot. Absts., Vol. V, 



influences depend upon the kind and proportion of the varieties in the mixture, it is possible 

 to get mixtures that produce a higher average yield than any of the varieties of which the 

 mixture is composed. [See also next following Entry 1551.] — W. H. Eyster. 



1551. Koch, L. Onderzoekingen betreffende de praktijkwaarde van de iijnenselectie- 

 methode voor verschillende eenjarige landbouwgewassen. [Researches concerning the prac- 

 tical value of the line selection method for various annual tropical crops.] Teysmannia 29: 

 1-36, 96-127, 156-191, 389-423. 1918.— The line-breeding method was first practised in 1907 

 by van der Stok, then assistant at the botanical section of the Experiment Station for Rice 

 and other Annual Crops at Buitenzorg, Java. A great deal of line breeding had been per- 

 formed before 1915, the selected crops being specially rice, ground-nuts and soy beans. Dur- 

 ing the years when most breeding took place (1911-1915) some peculiarities were noticed, which 

 gave birth to the idea that line breeding was by no means a method for securing high-pro- 

 ducing rice strains, etc. In the trials (almost all of them with 8 or more control plots) it was 

 observed that the population (mixture of all strains, high- and low-producing) gave in most 

 cases an unexpectedly high yield, higher than most selected pure strains. Breeding did meet 

 with success where immunity for certain diseases or qualitative peculiarities were aimed at. 

 As most breeding was for increasing the yield, a series of trials was undertaken to determine 

 whether line breeding should be continued or not, and to investigate the reason why there 

 was so little success. — In the years 1914-1916 selection took place for 6 rice varieties. In only 

 2 of 16 trials did the selected rice strains give a fairly good yield in comparison to the unse- 

 lected mixture. As a rule, a strain that gave one year the highest yield, failed to do so in the 

 next. More than once such a strain yielded much less than some others had that been much 

 inferior the previous year. — As the climate at Buitenzorg is somewhat peculiar, and results 

 might perhaps be influenced by the great rainfall or the moist atmosphere, trials were made 

 at the same time at the experimental farms at Ngandjoek and at Sidoardjo, these places being 

 situated respectively in the central and the eastern part of Java. Out of six trials at Ngand- 

 joek, the pure strains and the unselected mixture were alike; at Sidoardjo, in 2 out of 3 cases, 

 the strains failed to give a higher yield than the population. — The supposition arose that the 

 high yield of the population might be caused by the fact that the mixture is, generally speak- 

 ing, more suited for uneven circumstances than is a pure variety. — In order to investigate 

 this matter author began, in 1915, a series of trials wherein mixed-up pure strains were com- 

 pared with the same races unmixed. The same was done by mixing up pure varieties. In 

 most trials the varieties or strains were compared in this way: (1) variety A, 100 per cent; 

 (2) variety B, 100 per cent; (3) A, 75 per cent + B 25 per cent; (4) A 50 per cent + B 50 per 

 cent; (5) A 25 per cent«+ B 75 per cent. — Not only the yielding but also the stooling power 

 was examined. When the paddy was ripe the ears were cut by hand and afterwards all the 

 product in the trials where pure varieties had been mixed up was separated by hand so that 

 one could know exactly which part of the yield had been provided by variety A, and what 

 part by B. All heads were counted, so that the average weight was determined. The result 

 of 4 trials with 8 controls showed that the pure strains and varieties did, on the whole, not 

 so well as the mixtures. The stooling power shown by weekly counts, was in most cases 

 higher than the pure strains; in one of the four cases, however, all the counts were remarkably 

 lower with the mixtures than with the pure strains. Of two varieties, the highest producer 

 (singly planted) did not always give the greatest proportion of the product of the mixture. 

 1 n most cases the heads of the varieties that suppressed the other one became heavier and the 

 heads of the suppressed one became lighter.— Trials .of the same order w r ere made with maize, 

 soy beans and peanuts. With maize, yellow Menado corn and Saipan corn, singly planted, 

 were compared with mixtures of these varieties. The mixtures yielded as much as 12 per 

 cent more than the highest-producing variety separately planted. With soy beans the same 

 was to be observed: 70 per cent of black mixed with 30 per cent of white soy beans yielded 12 

 per cent more than black alone, and 28 per cent more than white alone. With peanuts, 9 

 out of 10 mixtures gave a higher yield than might have been expected from the yield for the 

 pure strains. — In the year 1916-1917, out of 4 trials comparing pure strains with mixtures of 

 the same strains, no conclusions could be reached as to which should be preferred, strains or 



