Mav, 1921. 



SCIENTIFIC A G R I C U L T I' R E. 



19.] 



The consistently early ones show a great range of 

 characteristics. Not all of them by any means are 

 wholly inferior. Several are now grown commercially 

 in India and Australia. For different reasons none of 

 them could compete with those which we already have 

 except perhaps in districts where the season is very 

 short. 



Effects of Environment. 

 As is well known environmental conditions cause con- 

 siderable variation in the length of the growing season 

 in different years. In 1918 for example it was for most 

 varieties between 35 and 40 days longer than in 1920. 

 The variations during 5 years are shown in Table 2 for 

 several well known varieties. The difference between 

 any two varieties is, however, remarkably constant, no 

 matter how long the actual growing season maj- be. 

 There are of course variations in these differences but 

 not as much as one might expect. 



The comparative earliness may vary in different loca- 

 lities. Thus Professor Pye of Dookie, Australia, informs 

 me that Minister, which he originated, is there an early 

 mid-season variety while here it scarcely ripens at all. 

 Many other varieties which are later than it in Austra- 

 lia, are much earlier here. The same is true of Firbank 

 wheat. The reports of the Experimental Farms show 

 that in different localities in this country the differences 

 between varieties may vary. 



The figures given in Tables 1 and 2 show the mean 

 ripening periods of many indi^-iduals in each variety. 

 The time at which the different plants in a variety ripen 

 covers several days; usually it is about a week when soil 

 conditions are uniform (see Table 3i. The distribution 

 of the individuals in each variety is in the form of a 

 tjTjical curve of variation. 



Scope of Crossing Work. 

 From Table 3 it will be seen that different varieties 

 ripen at many different times between the extremes of 

 earliness and lateness. When varieties not included in 

 this table are considered it is found that there are all 

 imaginable conditions between extreme earliness and 

 extreme lateness. Crosses have been made between 

 parents differing by only a few days and located at 

 various places on the whole range of variation. Crosses 

 have also been made between pairs of varieties show- 

 ing all degrees of difference. It should therefore be 

 possible to analyse the whole genetic situation. 'Several 

 of these crosses have been carried to the fifth genera- 

 tion and as many families of the later generations raised 

 as the great amount of labor involved would permit. 

 Data on a few of the t^-pical crosses are given in the ac- 

 compam-ing tables. These particular crosses have been 

 chosen for discussion because they represent various 

 degrees of difference between tlie parents. 

 First Hybrid Generation. 

 In all crosses the first hybrid generation plants were 

 at least as late as the later parent. This cannot be in- 

 terpreted as due to dominance in view of the results in 

 the second and later generations. It appears to be due 

 to the method of planting. Owing to the necessity of 

 raising as many second generation plants as possible in 

 order to get all possible segregates, the individual first 

 generation i)lants were grown under conditions which 

 would yield the maximum number of seeds. They were 

 spaced much furtlier apart than under ordinary condi- 

 tions. Tills naturally encourages greater vegetative 

 growth and delays ripening, so that the results are not 

 comparable with those for the pareats or later hybrid 

 generations. 



In this connection it is interesting to note that Tscher- 

 mak (.5) working with peas found that m a cross be- 

 tween an early and a late variety the Fi were uniformlv 

 intermediate. Leake (2) reports similar results fo'r 

 cotton. 



Another point to be noted is that the Fj plants fre- 

 quently showed increased vigor due to crossing. This 

 leads to greater vegetative growth and may delay ripen- 

 ing. On the other hand Ea.st and Jones "(1) state that 

 one general effect of hybrid \-igor is to put fonvard the 

 time of maturity. The rapidity of growth of the hybrid 

 more than compensates for the increased size. Probably 

 different results in this respect may be expected with 

 different plants. 



Second Hybrid Generation. 



In every cross the great majority of second genera- 

 tion hybrids were intermediate between the parents. 

 The range of variation in this generation extended prac- 

 tically from the early extreme of the early parent to the 

 late extreme of the later parent. This is shown in 

 Table 3. The distribution of the plants was in every 

 case in the form of a typical variation curve covering 

 practically the whole range of the parental tvpes and 

 the space between them. There was no heaping of in- 

 dividuals at either of the parental positions (which 

 would indicate dominance with simple segregation) but 

 only in the intermediate position. If the extreme varia- 

 tion with the recovery of the parental types is taken to 

 indicate Mendelian segi-egation it is oljviouslv not of 

 the simple type involving only one or two hereditary 

 factors. In all eases these results were obtained for at 

 most a few hundred plants. 



Perhaps the most striking feature of these results as 

 shoi,vn in Table 3 is their uniformity. Xo matter how 

 small or how great the difference between the parents 

 may be. nor at what region of the general range of varia- 

 tion two slightly different parents may be located, the 

 F, curve is strikingly uniform and covers practieallv the 

 whole range of variation of the parents concerned." The 

 table shows that the more nearly the parents are alike 

 the greater the probability of recovering the extreme 

 parental conditions. This may be due not onlv to gen- 

 etic conditions but also to the fact that the necessarily 

 limited number of second generation plants have to be 

 spread over a wider range in the case of wider crosses 



The F, of crosses between other pairs of the parental 

 varieties represented in Table 3 were grown in subse- 

 quent years. Thus the F, of B x C and C x E were 

 grown in 1919. and results were similar to those just des- 

 cribed. The F„ of A X E, the cross involvin? the widest 

 parental difference, were raised in 1918 and extended to 

 the means of both parents. In that year the parents 

 were not as different as usual owing to the effects of a 

 rust attack. Consequently one cannot be sure that the 

 parental conditions were recovered in the F,. 

 Third and Later Generations. 



Many third generation families of each of these 

 cros.ses were grown in 1918 and later generations in sub- 

 sequent years. Some representative results in the cross 

 A X B (the earliest twol are given in Table 4. The 

 great increase in the number of plants with each genera- 

 tion made it impossible to record individuallv the^plants 

 of the fourth and later generations. But careful obser- 

 vations were made to determine tlie dav of ripenino- of 

 the fir.st plant in each family and the day when aU were 

 ripe. The fourth ?eneration families are therefore re 

 presented merely by lines which show tlie range of 

 variation. Really the fourth generation plants were 

 grown a year later than the third so that the dates given 



