IN THE TKOPICS : III. 171 



The statistical proof is not conclusive, because it is quite 

 possible for a normal curve to be made up of a number of 

 minor curves having independent modes different from that 

 of the main curve. Still a certain interest attaches to this 

 method of presenting the facts. 



A portion of the above result was described in a letter to 

 Nature (vol. 70, p. 601). The account was however expressed 

 in such an imperfect manner that so astute a critic as Prof. 

 Karl Pearson missed the main point of it, as appears from his 

 letter in the following number. For this reason I feel con- 

 strained to deal here with one or two of the statements which 

 Prof. Pearson's letter contains. 



Like myself, Prof. Pearson headed his letter " Mendel's 

 Law," but states : " The difficulty is to know what is understood 

 by this term." What I personally understand by the term 

 and what is in my opinion the only legitimate use of it since 

 it was so first defined by Correns, I have stated on the 7th page 

 of the paper from which Prof. Pearson quotes in his letter — 

 I mean the law that the gametes of a heterozygote contain in 

 equal numbers the pure parental allelomorphs completely 

 separated from one another.* It is in evidence of this law 

 that I believe that I have provided a "crucial experiment" 

 by showing that in the case of both the male and female germ 

 cells of the heterozygote produced by crossing strains of maize 

 with a white and with a yellow endosperm respectively, this 

 segregation in equal numbers is no theory but a fact. 



The same proof was indeed afforded by Mendel himself in 

 his classical experiment with peas, but too small a number 

 of individuals was then dealt with for the evidence to be regard- 

 ed as final. 



The proof (as stated in Nature) consisted in crossing the 

 female flowers of the heterozygote (each bearing one germ 

 cell) with pollen from a recessive form. As the result of this 



* And if the cross-bred plant ia heterozygous in respect^ of more 

 than 1 pair of allelomorphs, then al> possible combinations of these 

 allelomorphs occur in equal numbers of gamete*. 



