66 CROSSING. 



felt cis a very uncomfortable one by the first "Mendelians", who 

 almost to a man conceived of their inherited factors as of direct 

 determinants for the characters studied. The first attempts to 

 bring these results into line with the hypothetical purity of the 

 factors consisted of trying to analyze the differences between 

 the types crossed into a great many separate pure unit-char- 

 acters. This could be done in a great many instances, but in 

 many other ^ases the attempt had to fail. A number of cases 

 remained in which, the difference between two individuals 

 was essentially a difference in just one character, but in which 

 nevertheless the inheritance of this character was not alter- 

 native. 



This difficulty was necessarily felt as a very serious one by 

 all those authors who held on to the determinant-conception 

 of genes. If there should be a direct and reciprocal connection 

 between a determinant for some character and this character, 

 it would be obvious that the production in generations follow- 

 ing a cross of individuals with a character intermediate be- 

 tween that same one in their parents would tend to prove that 

 the correspbnding determinant for the character could vary in 

 quality, and could exist in different conditions. Of late years 

 we have read a good deal about the difference between so- 

 called qualitative and quantitative characters. It was generally 

 conceded that the inheritance of quaUtative characters such as 

 colour, or presence and absence of organs, followed Mendel's 

 law, but that the case of quantitative characters constituted a 

 difficulty. 



What, really, does this difference between qualitative and 

 quantitative characters consist of? If we examine the differ- 

 ence, we find that it is only one of degree after aU Qualitative 

 differences are such as result when in the material we are work- 

 ing with, one gene happens to have a striking influence on 

 some character, and when this gene is not present in the geno- 

 type of all the individuals. In such cases, the distribution of 

 such a gene over the germ-cells of a heterozygote can easily be 

 observed, as its presence in the zygote can always be detected 



