162 CARL DOWNEY LA RUE 



been discontinued after three selections it would have appeared that a 

 small, but possibly significant, result had been produced by the selections. 

 The difference between the two lines at that time was 0.267/1, which is 

 probably more than three times its probable error which has not been 

 computed. It is at any rate, more than three times the error of the differ- 

 ence between the plus- and minus-selected lines for the whole experiment. 



When selection is continued for a longer time the apparent effect of 

 selection is soon lost and the minus-selected line at times produces longer 

 spores than the plus-selected one. The intermediate form, which is entire- 

 ly unselected, shows greater fluctuations than either of the other groups, 

 and its mean for the whole experiment is less than that of the minus- 

 selected line. If we assume that the upward and downward swings, so 

 prominent in this experiment, are merely chance fluctuations due to 

 environmental influences we can understand why the intermediate line 

 shows more fluctuation than the others, since it is represented by only one 

 culture in each generation and its generation means are computed from 

 only 100 measurements instead of 1000 as in the case of the other two lines. 



Since the two selected lines cannot be shown to differ from one another, 

 or from the unselected line, we can only conclude that selection has been 

 of no avail in producing lines of Pestalozzia Guepini distinct for length of 

 spore. 



Experiment 2. Selection for length of spore appendages 



In this experiment the same strain was used as in experiment 1 , namely, 

 No. 29. Table 5 presents the means and ranges of appendage length for 

 the eight generations during which the fungus had been studied prior to 

 the initiation of this experiment, and a frequency polygon of this strain 

 is shown in figure 2. 



The plan of the experiment was that explained earlier in this paper 

 and used in experiment 1, and was rigidly followed throughout the whole 

 series of selections which were continued for more than a year. All of the 

 cultures were grown on Hevea agar, made according to the formula already 

 given, from the original lot of Hevea decoction, and used with all the 

 precautions previously mentioned. This work, like that of experiment 1, 

 was all done in Sumatra within a few hundred yards of the place where 

 line No. 29 was originally found. 



In this experiment twenty-five selections were made for length of spore 

 appendages, and twenty-five successive generations were grown of an 

 unselected intermediate line also. The appendage lengths of all the parent 

 spores are shown in figure 6. The extreme cultures of the plus and minus 



