180 ILLINOIS STATE ACADEMY OF SCIENCE 



3031 4 are based on the five lines of ± strain A brood and 

 the ± B brood from which selection was made. 



However, the A brood is often small, sometimes only 

 four or five individuals, and therefore not so good to base 

 selection upon. The method was changed again so as 

 to average together the A and B broods of each of the five 

 plus lines and of each of the five minus, and select five in- 

 dividuals from the ± C brood of the most favorable aver- 

 age. This was done to start generation 3033 6 and has 

 been continued to date (generation 3037 included). Cal- 

 culations for all curves from 3032 to 3037 are based there- 

 fore on ±(5A's + 5B's + C) used for selection. 



The results are shown in the table of Means and Dif- 

 ferences each with their probable errors. It shows that 

 the least positive results are a difference which is still 

 three times the probable error, and two times as great 

 as the difference between the first selected individuals. 

 In one case the difference is twenty times the probable 

 error. The frequency curves are interesting. They 

 seem to show that whereas at the beginning the difference 

 is often due partly to odd individuals at the extremes, 

 later these are more or less eliminated and the population 

 more even. The mode of the plus curve seems to be mov- 

 ing to the right of that of the minus curve. 



Environmental conditions have important bearing on 

 the results and probably explain the fluctuations of the 

 difference. When the food water becomes concentrated 

 other algae get the upper hand, one occurring particu- 

 larly on which Daphnia could not subsist and in the long 

 strings of which it became entangled. A sudden rise in 

 temperature kills them rapidly, such as is caused by the 

 sun shining on the bowls. The food must be replenished 

 continually or they do not thrive. The writer has looked 

 for indications that size of brood or size and vigor of 

 animal causes a difference, but has been unable to find 

 any signs of this. The vigorous animal is larger both 

 ± but the length of spine has the same relation to body 

 length. All individuals were measured at about seven 

 days old. One cannot tell exact date of birth unless ma- 

 terial is under constant surveillance. The size at the 

 same age varies considerably, but not so the index. 



