58 H. E. EWING. 



more accurately in dead, mounted specimens than in live in- 

 dividuals. 



(b) The fecundity of the individual is known when the in- 

 dividual is selected, thus obviating the termination of the experi- 

 ment through the unsuspected sterility or low fecundity of the 

 selected individual. 



(c) Permanent mounts of all individuals examined are desirable 

 as a permanent record to be used later in collecting valuable data. 



While making selections in isolation No. 3, I decided upon a 

 system of numbering individuals which was used throughout the 

 series of experiments. According to this system each individual 

 is given in reality three numbers which are written together as 

 one. The first of these, which is written much larger than the 

 second, refers to the number of the isolation (i. e., the isolated 

 subline of the main pure line). The second number, which is 

 written much smaller (a subnumber), refers to the generation 

 to which the individual belongs counting from the stem agamic 

 mother that was first isolated from individuals found in nature. 

 The third number, written the same size as the first, refers to the 

 individual's number in the fraternity to which it belongs. Thus 

 3ie 5, written together as a single number for an individual means 

 that this individual belongs to isolation 3 (the isolation here 

 considered), generation 16, and has been given the number 5 

 as its individual fraternity number to designate it from the other 

 members of the 3i 6 fraternity. 



Before starting the selections in this isolation the mean cor- 

 nicle-length for the line was determined by getting the mean for 

 all the individuals obtained in six previous generations. The 

 mean length of the cornicles of the wingless agamic females was 

 found to be 2.609. The mean length for the winged females was 

 found to be 2.441 . Or in other words the cornicles of the wingless 

 agamic females were found to be 1.065 times longer than the 

 cornicles of the winged agamic females. During the process of 

 selection several winged forms were met with. In such instances 

 the measurements of their cornicles were reduced to the de- 

 nominator of the wingless form simply by multiplying by 1.065. 



Selections in isolation 3 were begun in the I5th generation and 

 continued for II generations, ending in the 26th. In almost 



