l52 Arend. L. Hagedoorn and A. C. Hagedoorn. 



This is exactly what would be required, to make the experiments 

 trustworthy, and fit to show, whether or not the genes, influencing 

 the distribution of pigment in rats are modifiable by selection. Every- 

 thing centers around this question: were the rats, with which Castle 

 began his selection-experiments at least all homozygous for all the 

 factors which influence the distribution of pigment, or were they 

 possibly not? For, of course, if there is a possibility, that the rats 

 may have been heterozygous for such factors, or even that they did 

 not all contain the same genetic factors, the results prove nothing 

 whatever. 



What proofs does Castle advance to make acceptable the required 

 supposition, that his rats constituted a material, homogeneous in res- 

 pect to genetic factors? He simply states that "Hooded rats differ 

 from Irish ones in one unit-character". And this he seems to think 

 sufficiently proven, by the fact that "Irish and Hooded mendelize". 



Now, in the first place, we have to ask ourselves this question: 

 does the fact that one type differs from another in having one gene 

 more or less, necessarily imply that it be pure in respect to all other 

 genes? Posing this question is answering it. 



A looseness in applj'ing the term "unit-character" is to blame, 

 or rather, the term "unit-character" itself is to blame, being all- 

 together too elastic and too easily misapplied. If we say, that black 

 and agouti mice differ in one pair of "unit-characters", this simply 

 is a statement of the fact, that the agouti hybrids from a black and 

 an agouti parent, produce, inter se, three times as many agouti as 

 black offspring. If we want to express, that all agoutis have a 

 genetic factor, which all blacks lack, it does not suffice to say, that 

 black and agouti differ in one unit-character. There are at least 

 two kinds of agouti, differing because of the presence in one of 

 another gene. Those mice having this gene are richer in colour than 

 the others. There are also two corresponding kinds of black, with 

 and without the same factor. The statement: "Black is a unit- 

 character" has no significance. It is possible, by selection, to make 

 a population of only black mice, either on the average darker or 

 fader, by changing the proportion, in which the animals with or 

 without the second genetic factor find themselves. 



The fact that the Hooded rats have one unit-character, or, to 

 state it a little more correctly, the fact, that Hooded and Irish rats 

 differ in one pair of unit-characters, does not necessarily imply, that 

 all the Hooded rats must therefore constitute one genotype. The 



