The Mendelian Inheritance oi Certain Chemical Characters in Maize. 23 



2.42 = expected percentage of crude fiber in F.y sweet 



kernels. 

 2.54 = observed mean percentage of crude fiber in all F^ 



sweet kernels. 

 Difference between observation and theory for sweet kernels - 0.12. 

 This result is distinctly better than that for the starchy kernels, 

 in fact the agreement may be said to be quite close. 

 For pentosans the figures are as follows: 

 6.47 = expected percentage of pentosans in all Fg starchy 



kernels. 

 5.955 = observed mean percentage of pentosans in all F» 



starchy kernels. 

 Difference = 0.515. 

 For sweet kernels we have 

 7.3g = expected percentage of pentosans in Fg sweet 



kernels. 

 8.09 = observed mean percentage of pentosans in all F.i 



sweet kernels. 

 Difference = 0.70. 



Here again the agreement between observation and theory is not 

 so close as is was found to be in the other (non-carbohydrate) characters. 

 Careful consideration of the facts, however, leads us to the opinion that 

 this is mostly due to unavoidable difficulties and errors in the analytical 

 work upon these characters. It does not necessarily mean either a failure 

 of the hypothesis on which the calculations are based, or that these 

 characters "crude fiber" and "pentosans" do not, as a matter of fact, 

 segregate independently. By the ordinary methods of analysis one 

 does not get absolute determinations of either crude fiber or pentosans. 

 These are not direct determinations, as are nitrogen and crude fat, 

 for example, but are comparative. 



Bj' way of summary i) it may be said that analysis of random 

 samples of the four visibly different classes of Fo kernels obtained 

 from a cross of a yellow starchy and a white sweet race of maize 

 indicate in general a Mendelian segregation of invisible chemical 



') Lest there should be any misunderstanding as to why sugar and starch 

 content are not discussed in this section of the paper, it may be said that there is 

 obviously no need for further discussion as to the separate independent segregation 

 of these characters because they are both directly connected with the external visible 

 appearance of the kernel. The evidence that they segregate as unit characters is 

 direct and unmistakable, and is to be found by direct observation of the kernels in 

 F.^, or in more refined form, from Tables I and II. 



