148 



A. FRANKLIN SHULL AND SONIA LADOFF 



TABLE 17 



Showing the number of groups (of various sizes) of successive male-'producers in 

 the families of the two lines recorded in table 16 



in table 16, one examines the daily output. In each day's 

 product occur members of several successive generations. As 

 ShuU (1915) has pointed out, rhytlmiical production of male- 

 producers may often be more easily detected by an examination 

 of daily records than by a study of family records. In table 18 

 is recorded the number of male-producers and female-producers 

 on each day in each Hne of table 16, regardless of the famihes 

 to which they belong. The percentages of male-producers are 

 computed for tri-daily periods, in order to smooth some of the 

 enormous fluctuations in both hnes. These tri-daily percentages 

 are graphically shown in figure 1, where it is seen that the oxy- 

 gen line is plainly more uniform than the control line. 



If it be objected that the daily records should not be combined 

 for comparison, that combination may be rejected, and the dif- 

 ference between the two hnes is still plainly visible. The num- 

 ber of male-producers appearing on successive days in the 

 control line is subject to greater extremes of fluctuation than 

 in the oxygen line. Of the 39 days through which the experiment 

 extended, there were twelve days on which no male-producers ap- 



