236 Castle. 



and so we find such workers as Punnett (an uneoniproniising advocate 

 of gametic puiity) assuming whenever a modification has been observed 

 that a distinct modifier has become "coupled" with the ordinary condition 

 of the unit. Thus the Himalayan rabbit is on his view an ordinary 

 albino plus a Himalayan modifier, (not as yet, however, detachable under 

 experimental conditions); and a peculiar strain of dark black rabbits is 

 to him a race of ordinary blacks plus a coupled darkening factor. This 

 assumed darkening factor is, however, demonstrable only in a changed 

 condition of the black ("extension factor") itself. No proof exists that 

 it has a separate existence, as would be shown for example if it were 

 capable of being detached from the black and introduced into a yellow 

 race and then put back on black again. To assume the existence of a 

 distinct modifying factor whenever a unit-character is observed to change 

 is scarcely justified by present knowledge. It affords, it is true, as 

 East has maintained, a workable terminology, tho it seems needlessly 

 cumbersome and voices constantly a theory which may be false and at 

 any rate is certainly unproved. Besides its objectionableness on theo- 

 retical grounds such a terminology is likely to mislead practical men 

 who desire to apply the conclusions of genetic investigation to the 

 breeding of farm animals and cultivated plants. Already we see its 

 consequences becoming evident in the work of agricultural experiment 

 stations. The current talk about "pure lines" and the "ineffectiveness 

 of selection" is leading some to abandon hill selection of potatoes as 

 useless and leading others to look to crossing dairy breeds of cattle 

 for their further improvement rathei' than to selecting within the pure 

 breed. Now such changes in practice may be in the right direction; 

 undoubtedly they are if the "pui'e line" theory is true. I am only 

 pointing out the impoi'tance of knowing whether or not the theory is 

 true, the present lack of demonstrative or even probable evidence in 

 its favor, and the desirableness of devising ways of putting it to an 

 experimental test. 



Bibliography. 



Baticson, \V. and Miss E. R. Saunders, 1902: Report I. to the Evolution Committee 

 of the Royal Society of London, IGO pp. 



Belling, John, 1912: Selection in pure lines. Amer. Breeders' Magazine, vol.3, p. 311. 



Calkins, G. N. and Louise H. Gregory, 1913: Variations in the progeny of a single 

 exconjugant of Paramecium caudatum. Jouru. Exp. Zool., vol. 15, pp. 467 — 525. 



Castle, \V. E., 1905: Heredity of coat characters in guinea-pigs and rabbits. Publi- 

 cation No. 23, Carnegie Institution of Washington, 78 pp., (3 pi. 



