342 PLANT HYBRIDIZATION BEFORE MENDEL 



"if the pollen grains differ externally in the parent races, then one may 

 expect that the hybrid may form two kinds of pollen cells, externally 

 distinguishable among themselves, in case Mendel's rule holds good. This 

 is in fact true, as Focke first observed." 



Correns appears to have been the Hrst to comprehend so thor- 

 oughly the fundamental nature of the Mendel experiments, as to 

 give the general result the designation of "Mendel's Law." 



"The hybrid forms sexual nuclei which unite with the primordia [An- 

 lagen] for the individual characters of the parents in all possible com- 

 binations, but not those of the same pair of characters. Every combination 

 appears about equally often. 



"if the parent races differ only in one pair of characters (two charac- 

 ters, Aa), then the hybrid forms two kinds of sexual nuclei (A,a), which 

 are like those of the parents ; of each sort, 50% of the total number, if 

 they differ in two pairs of characters (four characters, Aa, Bb), then 

 there are four kinds of sexual nuclei (AB, Ab, aB, ab) ; of each sort 

 25% of the total number, if they differ in three pairs of characters (six 

 characters, Aa, Bb, Cc), then there exist eight kinds of sexual nuclei 

 (ABC, ABc, Abe, Abe, aBC, aBc, abC, abc) ; of each sort 12.5% of the 

 total number; etc." (p. 166.) 



Immediately following this illustration, the statement is made : 



"I call this the Mendelian Rule : it includes also De Vries' 'Loi de dis- 

 jonction.' Everything further may be derived from it." (p. 167.) 



We thus have here the first completely definite analysis of the 

 Mendelian paper itself, of Mendel's own results, and the first 

 use of the term, "The Mendelian Rule" or "Law." 



However, Correns was not at the time at all of the opinion 

 that Mendel's Law was a universal one in its character. He held 

 that it obtained for a certain number of cases, and presumably 

 for those where a single member of the character-pair dominates, 

 and for the most part only in the case of race- or variety-hybrids. 



Correns holds that the view : 



"That all parts of hybrids follow it, is quite out of the question." 



To this point he cites his cases of crosses between two races of 

 peas with colorless and with orange-red seed-coats becoming 

 brown on ripening, in which, in the first generation, different 

 stages of intermediates are found. In the second generation, the 

 extremes of yellow and colorless gave aga(n the same extremes, 

 united by a series of transition-stages. The same is stated to hold 

 for the character of the seeds, and their form and size. (p. 167.) 



We thus have in this paper an analysis of the work of Mendel, 



