358 



The Journal of Heredity 



for shape, an index was reckoned by 

 dividing the length by the width. This 

 index should be 1 for the two-round 

 varieties because the length and width 

 should be equal; the index for White 

 Icicle should be somewhere between 

 4 and 5, because the length of this va- 

 riety exceeds its width. The mean 

 index for Scarlet Glol^e was 1.35, the 

 mode falling at 1.02 which indicates 

 that this variety was too long for the 

 ideal shape. Only 21 out of 176 roots 

 were really true to type, having an 

 index of 1.00. White Box had a mean 

 of .97 — very close to perfect, but the 

 mode or class of greatest frequency was 

 at .86. Thus the greatest number of 

 uniform shape were too short or oblate. 

 Only 20 plants out of 236, or 8.47 fo, 

 were ideal. White Icicle had a mean 

 index of 4.25, the mode falling at 4.32. 



SCARLET GLOBE EXTREMELY VARIABLE 



Scarlet Globe was extremely varia- 

 ble in its shape, as shown not only by 

 its coefficients of variability of 25% 

 and 23%, respectively, for length and 

 thickness, but especially by that for 

 index of shape, which was 90.39%, as 

 against 21.96% for White Box and 

 26.55% for White Icicle. Scarlet 

 ( ilol)e also showed indications of com- 

 prising more than one genetic strain 

 for length, as indicated by its several 

 modes. The. larger of these modes 

 nearly coincided with the mean or ave- 

 rage of the population. 



It will be seen from inspection of the 

 accr)mpanying table of constants, which 

 will show, perhaps, better than a dis- 

 cussion, that there existed a great 

 amount of variability in all three varie- 

 ties in regard to weight, length, width, 

 and general shape. (Appendix, p. 384.) 



Weight was the most variable char- 

 acter in all three varieties, the coeffi- 

 cients of variability for this character 

 in Scarlet (ilobe, White Box, and White 

 Icicle being, respectively, 49.15%, 

 66.72%, and 55.34%. The ditTcrence 

 between the mean or average and the 

 mode or weight of greatest frequency 

 shows how far from uniform these 

 samples were. In Scarlet Globe the 

 greatest frequency was found to be at 



10.5 grams, 17% of the whole number 

 of roots occurring in this class, although 

 the mean or average of the population 

 was 12.58 grams. In White Box the 

 mean was 17.22 grams, but the greatest 

 number of roots of uniform weight fell 

 in the class 3.8 grams. This light- 

 weight class comprised 18.35% of all 

 the roots. In White Icicle with a mean 

 of 47.59 grams, the largest class of in- 

 dividuals was at 30.5 grams, and con- 

 stituted 12.31% of the whole popu- 

 lation. 



LENGTH AND IIIICKNESS VARY 



As to the length, we lind the first two 

 varieties both 25% variable, and the 

 third one 27% variable. While the 

 mean and mode run fairly close to- 

 gether in all cases, we find that only 

 about 13% of the whole population in 

 any case is approximately the same 

 length as the average. 



Essentially similar conditions exist 

 in regard to thickness, and variability 

 as to general shape runs from nearly 

 22% in White Box to as high as 90.39% 

 in Scarlet Globe. This is a very un- 

 desirable condition in a crop that is 

 grown to be harvested all at the same 

 time, and where uniformity is an im- 

 portant feature in the attrativeness of 

 a bunch of radishes in the market. 



In order to ascertain which factor 

 was more important in influencing 

 weight, whether length or thickness, 

 each of these dimensions was correlated 

 with weight. As can be seen by the 

 table (see appendix), width was far 

 more important than length in Scarlet 

 (jlobe and White Box. the coefficients 

 of correlation being 88% and 90%, re- 

 spectively, as against 65% and 82% 

 when length was correlated with weight. 

 Length was slightly the more influen- 

 tial factor in White Icicle, the long va- 

 riety, as shown by correlation coeffi- 

 cients 88% for length as against 87% 

 for width. 



COMMERCIAL APPLICATION OF RESULTS 



The results of this little study are 

 interesting in the light they throw on 

 the existing conditions of commercial 

 varieties of one of our common vege- 



