360 



The Journal of Heredity 



tables. While it cannot be claimed that 

 the constants here obtained would be 

 found in other samples of seeds that 

 might be obtained elsewhere or grown 

 under different circumstances, or that 

 they are apijroximately true for other 

 varieties of radish, or other root crops, 

 nevertheless, most growers of vege- 

 table's would admit, 1 believe, that con- 

 ditions found in this study are fairly 

 representative of general conditions in 

 this group of vegetables. It is well 

 known by growers that much of the 

 seed planted is not viable, that a large 

 portion of the crop is not true to type 

 and must, therefore, be culled out as 

 unmarketable, and that there is great 

 variability as to rate of growth and 

 time of maturity. If data were obtain- 

 able on quality and color of flesh, it is 

 not to be doubted that the variability 

 in those characters would be any less 

 than in these here studied. It is real- 

 ized that much is to be desired in con- 

 stancy of all those characters upon 

 which the commercial value of kinds 

 and varieties of vegetables depend. 



'J'his excessive variability is easily to 

 be explained by the methods of selec- 

 tion generally practiced by seed grow- 

 ers. .Sometimes seed is saved from 

 those plants which are most vigorous 

 or first come into bloom (as was ap- 

 parently the case in White Box), and 

 seldom is careful attention given that 

 the root shape be very true to ideal ; 

 moreover, mass selection is used, 

 whereby the less desirable characters, 

 often lying hidden and latent, are kept 

 along with the most desirable characters 

 of the best selected plants. Tliere are 

 two chief reasons why it is difficult to 

 practice careful selection in such crops. 

 JMrst, that i)art of the plant where con- 

 stancy to type is desirable lies out of 

 sight in the ground. To dig and select 

 these roots at the time they are market- 

 able would be fatal and no seed could 

 be obtained from them. In the second 

 place, the i)lants. as grown for com- 

 mercial seed production, are left to open 

 pollination, with plenty of chance for 

 cross pollination with poorer individuals 

 or even with different varieties. This 



is a fault hard to obviate in commercial 

 practice, because the radish does not 

 readily set seed when the flowers are 

 covered, and the flowers usually need 

 manipulation to insure self-fertilization. 



SUGGESTIONS FOR SELECTION OFFERED 



The following suggestions are offered 

 as possible and practicable methods 

 of selection for securing better true- 

 ness to type in such crops. First, 

 determine what is the ideal type as to 

 shape, size, length of season, etc., for 

 the particular kind and variety. Then 

 carefully select a few plants which 

 closely approximate this ideal at the 

 time they should be ready for harvest- 

 ing or for market, 'i'his could be done 

 by removing the earth carefully from 

 about the roots long enough to observe 

 them and then quickly replacing it 

 around those plants which are found 

 to be close to standard. These only 

 should be retained and allowed to set 

 seed. Varieties likely to cross should 

 not be planted near together. The seed 

 from each of these selected plants 

 should be sown in separate plats or 

 rows, and no seed should be saved from 

 this generation except from those plats 

 or rows which show greatest constancy 

 to the type for which their parent was 

 selected. 



Another method of breeding that 

 might be suggested would be to grow 

 the plants as pscudo-bieimials ; that is, 

 to let the roots attain marketable size, 

 then dry them off and dig them. They 

 could then easily be measured, and 

 could be stored until the beginning of 

 the next growing season. It would be 

 possible to do this under controlled 

 greenhouse or irrigation conditions, or 

 possibly, by planting so late in the sea- 

 son that growth would be checked at 

 the right stage by the advent of cold 

 weather. The seed obtained from these 

 plants during their second growing 

 season should be planted in separate 

 rows or plats from each plant. The 

 seed from all tiie plants in each row 

 or plat might be harvested all together, 

 but unless each row or plat subse- 



