Rate of Change of Hereditary Factors. 13 



but lethal mutations are so much more frequent than the type of 

 visible character variation ordinarily dealt with, that none of the 

 latter were observed in the whole experiment. 



The above figure of 20 in 1,062 has a probable error due to 

 chance of about ± 3 in 1,062. There can, therefore, be no doubt 

 about the correctness of the order of magnitude of the ratio 1 : 53, 

 so far as any error caused by random sampling is concerned. The 

 ratio 1 : 53 is, however, a composite result, for the families were 

 kept in two main lots, one at about 66° F., the other at about 8o°. 

 The 445 grown at the lower temperature produced five lethals, 

 or one in ninety; the 517 at the higher temperature produced 13 

 lethals, or one in forty. The other two were new lethals which 

 occurred in the 100 bottles kept at room temperature, in which 

 the lethals found in the two main series were being tested out. In 

 this connection, it should be pointed out that the high ratio of one 

 in thirty observed in the earlier experiment was obtained in bottles 

 kept at room temperature in the warm climate of southern Texas. 

 Although the absolute numbers of lethals are small, the difference 

 between the two series in the later experiment is probably statis- 

 tically significant, — at least, it may be calculated that if the 

 lots had really been similar, the chances would have been about 

 twenty to one against a difference of this magnitude occurring 

 between figures of the given size. Taking the figures at their 

 face value, we should obtain Qi 0 for mutation between 2 and 3, 

 as is usual for chemical reactions. 



If we accept the one in fifty-three ratio as representing the 

 average frequency for the X chromosome, and if, as there is 

 reason to believe, mutation occurs at the same rate in the other 

 chromosomes as in the X chromosome, then, since the X's form 

 about one fourth of the entire chromosome mass, we may figure 

 that about one fly in every thirteen has a new lethal mutation in 

 some chromosome or other. It is evident that, at this rate, with- 

 out natural selection to weed out the "unfit," the race would soon 

 become filled with lethal factors. For the X chromosome alone, 

 since each female has two X's, and one female in fifty has a new 

 lethal, we may figure that one X chromosome in every 100 con- 

 tains a lethal factor just arisen in the present generation. Or, 

 to put the matter differently, each X chromosome would, on the 



