THE INHERITANCE OF A FLUCTUATING CHARACTER. 



By T H. morgan and CALVIN B. BRIDGES. 



[From the Dcparhncnt of Zoology, Columbia University, New York, and the 

 Carnegie Institution of Washington.) 



(Received for publication, May 23, 1919.) 



In the fall of 1909 Morgan observed that the trident-pattern on 

 the thorax of flies of some of the wild stocks of Drosophila melano- 

 ^aster varied both in the intensity of the pigmentation and in the 

 extent to which the pattern was spread over the thorax. The degree 

 of pigmentation and the extent of the pattern seemed to vary- 

 together. A scale of grades was devised (Fig. 1) by means of which 

 it was possible to describe rather accurately the different populations 

 dealt with. Thus, Table I and Curve a give the distribution of 

 individuals in "Falmouth" wild stock (October, 1911). 



Morgan selected throughout the fall and winter of 1909-10 for 

 increased trident-pattern, with no certain progress until January, 

 1910, when a few flies were found that were considerably darker 

 than any hitherto observed. By breeding from these darker flies 

 a race was quickly obtained which showed no individuals as low in 

 grade as those in Classes I and II and very few in Class III. This 

 stock was called "with" and mass selection in it was continued for 

 about a year (thirty generations) with no apparent result further 

 than possibly to decrease the numbers in the lower grades. It 

 seems clear that the various grades observed in the stock prior to 

 January, 1910 were merely the normal fluctuations of the wild type 

 pattern, and that a definite mutation had then occurred which gave 

 rise to higher grades and a characteristic new distribution. A census 

 taken of this new "with" mutant population (October, 1911) re- 

 vealed the condition shown in Table I, and by Curve h, Fig. 2. 

 The most numerous class in the "with" stock was VII, which was 

 not represented at all among the wild flies. 



639 



