ZOOLOGY: R. W. HEGNER 
285 
One hundred 'wild' specimens were first selected at random from a large 
number taken from the pond. These varied in spine number from 7 to 13, 
and in diameter from 23 to 33 units (each unit being 4.3 microns). A marked 
correlation (0.325^0.060) was found between the spine number and diameter 
of these specimens. 
Small families were then reared from 70 'wild' specimens selected so as to 
include large, small, and medium sized organisms. Seven hundred and forty- 
six specimens were obtained in this way, ranging in number from only 2 or 3 
to 149 per family. The mean spine number of the families ranged from 10.40 
to 14.07. Variations in spine number occurred among the descendants of 
Non- 
Before selection 
selection Six selection periods Four nonselection periods 3 selection periods periods 
39 days 64 days 35 days 23 days 1 1 days 
. 7 gen. 22 generations 18 generations 15 gen. 7 gen. 
198 spec. 1218 specimens 1325 specimens 722 specimens 224 spec. 
13.15 
17.54 
12.80 
Uij 
i o/ 
: * 
. 11.68: ■ 
f 
10.87 
/^J138 
< ^># 
10.62 
11.26/ \X 1 
m 
S 
m 
o 
^$32 
/ i 
/ i 
/ N: / 
-4l084 
8.00/ 
/ 1 8.81 
■ 
• 
FIG. 2. DIAGRAM SHOWING THE MOST IMPORTANT HERITABLY DIVERSE LINES DERIVED 
FROM A SINGLE SPECIMEN OF ARCELLA DENT ATA (NO. 58) BY FISSION 
The character used was spine number. The letters indicate the designation of the lines, 
and the numbers are the mean spine numbers. 
single specimens during fission and these variations were in part inherited. 
It was found that the hereditary constitution of the different families was 
different with respect to spine number and the conclusion was reached that a 
'wild' population consists of a large number of heritably diverse families so far 
as spine number is concerned, and also probably as regards diameter, since 
spine number and diameter are closely correlated. 
The main problem was next undertaken, i.e., an attempt was made to isolate 
heritably diverse lines from among the descendants of a single specimen produced 
