70 THE ROYAL SOCIETY OF CANADA 



ost in some localities. The truth of these statements is emphasized 

 by a consideration of the meteorological data. At the very centre 

 of the prairie provinces (Saskatoon) the average number of days 

 between the last spring frost of 3 degrees and the first autumn frost 

 of 3 degrees is 121 days. At other points in Saskatchewan the period 

 varies from 88 to 132 days. And, of course, in many years the period 

 will be two or three weeks shorter than the average. 



In addition to lessening the danger from frost a short growing 

 period is desirable for other reasons. In the case of cereals, for ex- 

 ample, experience shows that the early ripening plant is much more 

 likely to escape the ravages of rust because it will be nearer maturity 

 before it is attacked. And the seasons in which there are the biggest 

 yields are just those in which environmental conditions favor both 

 rust development and the lengthening of the ripening period. Again 

 the early plant is likely to escape the effect of the all-too-frequent 

 droughts. By far the greatest precipitation occurs in June and early 

 July and thereafter prolonged droughts are liable to occur. The 

 plant which has neared maturity before the end of the rainy period 

 is less likely to suffer from drought. Then, too, the development 

 of earlier varieties would extend the area of crop production further 

 northward. 



For these reasons a thorough study of the genetics of early ripen- 

 ing appears to be a necessary preliminary to work of the most far- 

 reaching character in plant improvement in Western Canada. This 

 is particularly true in the case of wheat which is by all odds the most 

 important crop. 



2. History. 



Tschermak (1911)has carried on studies on the blooming time in 

 peas. An early blooming variety crossed with a later variety gave 

 offspring with an intermediate blooming time. In the next generation 

 there was great variation, some plants being earlier than the early 

 parent and some later than the late parent. The extremes bred true 

 but the intermediates gave various results. Tschermak explains 

 the results by assuming the existence of two factors, one of which 

 causes an intermediate blooming time and the second (in the presence 

 of the first) causes the plant to bloom much earlier. The results 

 expected in Fg and F4 on the basis of this hypothesis do not agree 

 very closely with those actually obtained. 



Hoshino (1915) also working with peas, obtained very different 

 results. The Fi plants were a little earlier than the late variety. 

 The plants of the next generation fell into two main groups, an early 

 and a late, separated by a group of minimal frequency. This result 

 shows clearly that segregation has occurred and that only one im- 



