358 Journal of Agricultural Research Voi.xxn.No. 7 



In conclusion, it may be said with certainty tJbat besides tlie S factor 

 in series 13255a and 13260a and the Si and So factors in 13126a, modifiers 

 are present which tend to dilute or intensify the spelt character. 



In statistical studies of density in wheat, the junior author has found 

 two characters whose mode of inheritance is almost identical with that 

 of the spelt character. When a dense wheat (T. compactiim) is crossed 

 with a lax wheat (T. vulgare) a 3 to i segregation is found in the Fj gen- 

 eration. The F2 density curve consists of two distinct curves. One of 

 these is a skew curve in the dense classes which contains 75 per cent of 

 the individuals. After a gap, the other curve, which is composed of the 

 lax segregates containing the remaining 25 per cent of the F2 population, 

 begins. Although the F3 progeny of these heterozygous dense plants of 

 the F2 generation invariably produce bimodal curves similar to that of 

 the F2 just described, their modes or the means of the dense and lax 

 curves shift at times considerably toward the lax classes and sometimes 

 toward the denser classes, much in the same manner as does spelting. A 

 similar phenomenon has been observed by Nilsson-Ehle (8) among his 

 dense X lax wheat crosses. 



The other parallel case is density of the type just mentioned, but in 

 this case the modifier is known to be the spelt factor itself. The curves of 

 the progeny of the heterozygous dense individuals of dense wheat X spelt 

 crosses have the general bimodal form, but the populations which, in 

 addition to density, carry the S factor always have their density curves 

 shifted toward the lax classes. 



With some spelts, the S factor shifts the density curves so much toward 

 the lax classes that this S has to be regarded also as an inhibiting factor 

 for density. 



Hull-lessness in oats, according to Love and McRostie (7), is inherited 

 in a similar manner. While this character segregates in the simple 

 mendelian ratio of i hulled to 2 intermediates to i hull-less, the inter- 

 mediate forms vary appreciably as regards the percentage of hulled 

 kernels they produce. By correlating the percentage of hull-lessness of 

 parent and offspring, they have shown that these variations within the 

 I to 2 to I ratio are hereditary. 



The mode of inheritance of the spelt character as shown in Tables IX 

 and X closely resembles also Castle's (2) case of hooded rats, which had 

 for a time aroused considerable controversy for and against the question 

 of inconstancy of unit characters. In numerous crosses between rats 

 having the hooded pattern and the wild (totally pigmented) or the Irish 

 (white belly) types the hooded pattern behaved as a mendelian recessive. 

 The ratio of nonhooded to hooded F2 offspring was 3 to i , showing that 

 the hooded condition is dependent upon a single factor difference. 

 Among the hooded individuals a considerable degree of variation was 

 observed with respect to the degree of the extension of this pattern. By 

 making selections for many generations in plus and minus directions 



