176 CARL DOWNEY LA RUE 



lonychia might be due to the accumulation of waste products in the cyto- 

 plasm of the slowly-dividing, large-sized group." The fact that MAST 

 (1917) secured two lines with different fission rates mDidynium nasutum 

 without selection is suggestive that mutation may offer an alternate 

 explanation of MIDDLETON'S results. 



ROOT (1918) working on Centropyxis has secured data worthy of consid- 

 eration. In one experiment, using mass selection, he found a decided 

 effect of selection on the number of spines after only four selections. 

 However, the number of individuals studied was very small, 88 in the high 

 series and 77 in the low series. In another mass selection the results were 

 exactly opposed to those of the first; that is, the parents with a low spine 

 number produced offspring with more spines than those from parents with 

 a large number of spines. The latter result obviously cannot be due to 

 selection, though had it stood alone it might have been so interpreted. 

 The two experiments show the danger of drawing conclusions from a small 

 number of individuals and a few selections. 



In a third experiment in which individuals were selected according to the 

 character of their progeny no considerable effect was gained until the 

 fourth selection. From the fourth selection a large effect was secured. 

 One selected individual gave offspring all of which had a high spine number. 

 This individual (5ala of ROOT'S series), the offspring of which are respon- 

 sible for a decided increase in the number of spines in the plus series, may 

 have been a mutation. The number of spines in the low series did not 

 appreciably decrease during the selection period. The whole experiment 

 included only a small number of individuals, 56 in the plus series and 51 

 in the minus, and only four selections were made. 



In the fourth experiment in which mass selection for number of spines 

 was practiced, two populations were segregated which varied in respect to 

 spine number. The difference between the two populations was small and 

 fluctuated from generation to generation. In the second selection period 

 the difference was less than in the preceding period. The number of 

 individuals observed was larger than in the other experiments but still 

 relatively small. It would have been interesting to see what the result 

 of a continuation of the cultures would have been. As it is, one suspects 

 the separation to have been only a temporary one, such as is shown for 

 Pestalozzia in table 4 of this paper. 



HEGNER (1919) carried on extensive selection experiments on Arcella 

 and finds that it is possible to separate by selection within families groups 

 which are distinct in regard to the selected character. In his principal 

 experiment, however, the effects of selection on number of spines are by no 



