160 



CARL DOWNEY LA RUE 



given generation parallel each other. When this is the case the deviation 

 between the two selections is small. When the fluctuations, for any reason, 

 do not parallel each other the deviations between the two selections may 

 be considerably increased or considerably decreased. 



From an inspection of figure 4 lit is difficult to tell whether or not the 

 parent cultures of the plus and minus strains are really different. It is 

 absolutely essential in selection studies that all selections be significant, 

 that is, the plus selection must be greater than the minus selection, not 

 only apparently, but statistically. It is by no means certain that this 

 has always been the case in previous investigations, and in many cases 



FIGURE 5. Graph of means of spore lengths of generations in experiment 1. The plus- 

 selected line is designated by P; the minus-selected line by M. The ordinates are lengths in /u; 

 the abscissae, the generations of the experiment. 



it is difficult if not impossible to know whether or not the differences are 

 significant, especially where individuals are selected without knowledge of 

 the mean of the generation. 



Pestalozzia is a very favorable organism in this respect since one can 

 employ the mean spore length of the selected cultures and know at the 

 same time the means of all the other cultures of that generation, as well 

 as the mean of those means. In this experiment it was found by computing 

 the errors of the differences between the plus and the minus selections that 

 these differences were significant in all cases. The smallest difference found 

 in the course of the experiment, .900 + .206^ in the ninth selection, is more 

 than four times its probable error and may be considered significant. 



