188 



on the acceptance of the idea of factorial constancy, which a priori need 

 not necessarily be a valid hypothesis. For it is, indeed, strange, if fac- 

 tors are the complex chemical bodies so many have assumed them to be, 

 that they should display such constancy in their relations with each other 

 in spite of the intimate contact which exists between them in the cell. 

 Castle, therefore, has pointed out that assuming that factors vary, it is 

 possible to account for the phenomena exhibited in size inheritance with- 

 out postulating the existence of so great a number of factor differences. 

 The case in point is that of the hooded pattern in rats. During the 

 progress of the experiments, over 25,000 rats have been reared and the 

 color patterns studied, so that this case has been studied as extensively 

 perhaps as any bearing on the subject. The rats have been graded ac- 

 cording to an arbitrary scale which is designed to express the extent of 

 pigmentation in such a way that the results of the experiments may be 

 analyzed statistically. The set of arbitrary standards employed together 

 with some rats which have been classified according to it are shown in 

 Fig. 88. As a result of these investigations, Castle has drawn a number 

 of conclusions of which the following seem most pertinent in this 

 connection: 



1. The hooded pattern of rats behaves as a simple Mendelian char- 

 acter in crosses with either the Irish pattern (white belly) or the wholly 

 pigmented condition of wild rats. 



2. Though behaving as a unit, the hooded pattern fluctuates that is, 

 it is subject to plus and minus variations. 



3. Selection, plus or minus, changes the position of the mean and 

 mode about which variation occurs. 



4. The results of such plus or minus selections are permanent, for 

 return selection is not more effective than the original selection, and dur- 

 ing return selection regression occurs away from the original mode, that 

 is, toward the mode established by selection. 



5. During the progress of the original selection variability as measured 

 by the standard deviation was somewhat diminished. 



6. Upon crossing the selected plus and minus races with each other, 

 the variability was somewhat increased in F\ and was further increased 

 in F z . The extreme condition (plus or minus) of the grandparents rarely, 

 if ever, recur in this generation. Only one individual among 378 F 2 

 young has been recorded in a grade as extreme as either grandparent. 



On the basis of these and other facts Castle argues that we must 

 recognize three types of inheritance: 



1. Typical Mendelian Inheritance. Factors of allelomorphic char- 

 acters may meet each other generation after generation in a common 

 zygote, but segregate in gametogenesis without any apparent modifica- 

 tion following their conjugation in the zygote. 



