IMPROVEMENT OF WHEAT BY SELECTION 89 
comparison in the laboratory to 100. Hach of these 100 
heads was threshed separately and their seed sown in 
small plots close together on land as even as possible. 
Among the 100 plots there were placed at frequent inter- 
vals plots of the ordinary Hunter’s of commerce to serve 
as a check upon the selections. As soon as the plants 
had started to grow many differences in habit of growth 
became apparent. Some were more vigorous, some taller, 
some more spreading, some darker green than the others. 
As growth progressed, further variations were seen and 
noted, and finally the 100 plots were cut and threshed 
separately. Of these rows the best 11 were selected and 
their seed sown next year again in long plots beside each 
other so as to avoid as much as possible the variations 
in soil. The growth characters in this year proved to 
be in the main the same as those noted in the preceding 
year, though in some cases the features were more marked 
and in some cases less marked. This was doubtless due 
to the difference in the season. The yields of these 11 
rows after threshing were as follows :— 
Average of Commercial Plots 100. 
Plot. named. bush.p.ac. Plot. named. bush. p. ac. 
Ts Al 118:1 7. AZT 106-7 
2. A2 82°6 8. A8 143-4 
3. A3 103-4 9. AQ 96°4 
4, A4 116-3 10. A10 112°6 
5. A5 77:9 11. All 121-2 
6. A6 114-3 
Plots 1, 4, 6, 8, 10, 11, were kept for further trial. 
Plot 8 had proved to be a natural hybrid—one of the 
three that have been observed among the thousands of 
single ear selections that have since been grown. Its 
hybrid nature explains its great productivity, but this 
gradually lessened in succeeding years, and since there 
were qualities of grain and straw that made it undesir- 
able it was finally discarded. It will not be referred to 
again. The seed from the remaining rows were sown in 
