[THOMPSON] EARLINESS AND LATENESS IN WHEAT 153 



It should be pointed out that this difference in variability would 

 make the order of the families arranged according to the date on which 

 their first members ripened, differ from the order of the F2 parent, 

 quite apart from the environmental influences mentioned in 4, above. 



6. In certain families there is a heaping up of individuals at a 

 particular point in the range of variation (e.g. Nos. 4, 5, 16). Follow- 

 ing the usual interpretation this would be taken to indicate domin- 

 ance along with segregation. In view of the fact, however, that there 

 was no indication of dominance in F2 and in many families of F3, 

 this interpretation is open to question. 



Table IV presents in a similar way the data on the F3 of the cross 

 A X D (first and fourth families with respect to earliness). Many of 

 the blanks in this table are due to the physical impossibility of taking 

 records on all the families every day. On the 140th day, for example, 

 no records at all were taken on this cross. In certain cases when a 

 day was missed, an attempt was made the next day to date the over- 

 ripe plants back, but this proved unsatisfactory and was abandoned 

 after a time. 



The points to which special attention was directed in connection 

 with the previous table are all well illustrated here, and this consti- 

 tutes a special point to be emphasized, namely, that though the differ- 

 ence between the two parents is so much greater the results are very 

 similar. Certain of the points are more clearly evident in this table. 



The chief difference is that which is to be expected — the greater 

 variability of the individual families. Compare Nos. 4, 10, 31, 34, 39. 

 The disturbance to the natural genetic order of the families through 

 disregarding this variability and using only the date of the first ripen- 

 ing plant is seen well by comparing Nos. 48 and 49. The F2 plants 

 which ripen on a given day particularly the intermediate ones, there- 

 fore have considerable genetic differences. In part, this is due to the 

 pushing of plants out of their proper class by environmental influences, 

 and in part represents the production of phenotypically similar results 

 by genotypically different constitutions. The differences in the con- 

 stitutions reveal themselves in the next generation by differences in 

 variability. 



Though the gap between the means of the parents is much greater, 

 all possible intermediate conditions are represented and only a small 

 proportion of the Fo intermediates furnished seed for F3. 



It is unfortunate that no reliable data were obtained for the third 

 generation of a cross involving greater parental difference. Such a 

 cross was carried to the second generation with results similar to 

 those in the other cases, but so large a proportion of the F2 plants 



