

OF PARENTS BY THOSE OF THE CHILDREN. (209) 17 
selection. CASTLE would demonstrate that by selection the genes them- 
selves vary, the same genes being continually together in the zygotes. 
The object of experiment was the hooded rat, i. e. a white rat with a 
black head and shoulders up to tbe forelegs and often with a black 
stripe on the back. The hooded property is a mendelian factor, is f. 1. 
recessive with regard to self-colour and with regard to the property 
Irish. CASTLE succeeded in breeding two series of hooded rats, a plus- 
series and a minus-series of which at the end (1916) the extreme plus- 
variation was almost quite black and the extreme minus-variation 
quite white. When recrossing with wild rats and with Irish ones, it has 
been shown that after crossing, the RR exemplars as well of the plus- as 
of the minus-series have turned back in the direction of the hooded 
type. And after some repeated crossings (1919) the result of the selec-. 
tion has been reduced to zero. 
From this result which agrees with the interpretation of selection 
of Huco DE VRIES (1901, S 52—91), CASTLE (1919) concludes that the 
genes have not varied and now he attributes the effect to the residual 
heredity. This residual heredity depends upon the working of modifica- 
tion-factors and agrees as theoretical explanation rather well with that 
of multiple factors. (NILSSON-EHLE 1909 S. 118). In my opinion there 
are for the explanation of selection still great difficulties, we after- 
wards come back to it (p. 45—52). 
So these exetensive selection-experiments of CASTLE have not fur- 
ther cleared up the nature of selection. Still after CASTLE's reply (1914) 
to HAGEDOORN it must be considered as a difficulty that they are not 
enough analytic; they have not been sufficiently begun as a mendelian 
experiment, but as a selection-experiment. So the children of the par- 
ents have not been continually pursued, but out of a group of off- 
spring at every turn some extreme variations have been chosen. 
- It is also to be regretted that CASTLE has not preceded the heredit- 
ary-analytic elaboration of his work through an elaboration according 
to the biometric methods (PEARL 1911). 
Of great importance is CASTLE’S assertion (1919) that through the 
-inbreeding on a large scale which has taken place in his material, the 
offspring of his animals is going to suffer very much. He foresees that 
before long his material will become extinct. 
_ A 2nd experiment of CASTLE regards the heredity of polydactylism 
of caviae (1906). Polydactylism (BATESON 1913, p. 34, 228) is a Mendel- 
(14) 2 
