October 19, 1917] 



SCIENCE 



393 



As indicated in the table these varieties 

 were grown under 14 different conditions, 

 covering a period of eight years. In tests 

 one to eight inclusive the varieties were 

 grown individually by the hill method, while 

 in the remainder of the tests the grain was 

 grown in drill row. It will be noted from 

 Table I. that the bearded varieties have pro- 

 duced more tillers per plant in every case. 



In Table II. is given the summary of the 

 results with all varieties included in the dif- 

 ferent tests on tillering. The number of va- 

 rieties together with their repetition in the 

 13 different tests amounts to 973 cases. For 

 the sake of convenience the ten highest and 

 the ten lowest tillering varieties in each test 

 are separated into a smooth and a bearded 

 group. 



TABLE n 



Summary of Variety Test on Tillering showing the 



Proportion of Bearded and Smooth Varieties 



in the Highest and the Ten in 



Each Test 



The number of tillers per plant refers to 

 the average for all bearded and all smooth in- 

 cluded in each test. It will be seen that the 

 larger proportion of the high tillering vari- 

 eties in every test is bearded while the greater 

 number of the low tillering varieties is 

 smooth. Of the 130 cases of the ten highest 

 tillering varieties, 111, or 85 per cent., are 



bearded; of the 130 cases of the ten lowest 

 tillering varieties, 106, or 80 per cent, are 

 smooth. The results of these tests indicate 

 that the bearded varieties have a greater ca- 

 pacity for tillering than the smooth. 



In the study of environmental factors it 

 was found that the rate of seeding a space 

 per plant has a marked influence on the num- 

 ber of tillers produced per plant. Close seed- 

 ing resulted in a smaller number of tillers 

 per plant, earlier maturity and a better qual- 

 ity of kernel than wide seeding. The time 

 of seeding determines to a large extent the 

 rate of tillering. Early seeding is accom- 

 panied by a larger number of tillers per plant 

 than late seeding. The time of seeding, the 

 number of tillers per plant, the yield per plant 

 and the quality of grain are closely correlated. 

 The competition between plants induced by 

 heavy seeding is more marked among smooth 

 wheats than among the bearded. It appears 

 that heavy seeding has a greater effect in les- 

 sening the number of tillers, the length of 

 culm, spike, and yield of grain in smooth 

 wheats than in bearded. The fertility of the 

 soil is a factor that directly affects the rate 

 of tillering. Nitrogen and phosphoric acid 

 seem to stimulate the production of tillers; 

 potash has little or no effect. The relation 

 of tillering to yield is shown by the increase 

 in the yield per spike as the number of tillers 

 per plant increases to 4 or 5, beyond this the 

 yield per spike is more or less ujiiform. The 

 low tillering plants of a variety produce a 

 smaller yield per spike, and the grain is of 

 poorer quality. These experiments have 

 shown quantitatively the effects on the rate 

 of tillering of such factors as time and rate 

 of seeding, the kinds of fertilizer and the re- 

 lation of the number of tillers per plant to 

 other characters. The tendency of bearded 

 wheats to tiller more freely than the smooth 

 has not been brought before to the attention 

 of the wheat grower. A close analysis of 

 these results indicates that there is a physio- 

 logical difference between the two types of 

 wheat which may mean that the bearded 

 sorts are able to make better use of the plant 

 food supplied or are able to extract it from 



