ZOOLOGY: R. W. HEGNER 
287 
line (EG) and a low line (EH) were obtained during three selection periods of 
twenty-three days during which 722 specimens were produced belonging to 15 
generations. The average difference between the mean spine numbers of the 
two lines during these three selection periods was 0.30. These selection periods 
were then followed by a nonselection period of 1 1 days during which 224 speci- 
mens were obtained belonging to 7 generations. The difference between the 
means of the two lines during this nonselection period was 0.44. It was con- 
cluded from these studies that two lines heritably diverse with respect to spine 
number, had been isolated from among the descendants of the low line (E). 
Measurements were made during part of the nonselection periods to deter- 
mine whether or not the diversities in spine number in lines A and E were 
accompanied by similar diversities in diameter. One hundred and twenty- 
seven specimens from the high line gave a mean diameter of 27.26 units of 4.3 
microns each and 136 from the low line a mean diameter of 26.92 units. The 
difference of 0.34 unit shows that the two lines were different in diameter as 
well as in spine number, and that on the average the greater the diameter the 
more numerous are the spines. A marked correlation was found between the 
diameters of the parents and those of their progeny, the coefficient of correla- 
tion being 0.489 ± 0.035. A high correlation also existed between spine 
number and diameter, the coefficient of correlation being 0.255 ± 0.042. 
Measurements were also made of the diameters of 384 of the progeny 
of the two branches (EG and EH) of the low line (E). These showed a mean 
difference in diameter of 0.26 unit corresponding to the difference in spine 
number. 
These data prove that the descendants of a single specimen of Arcella 
dentata produced by vegetative reproduction differ slightly from one another in 
their hereditary constitution (fig. 1 , B and C) and that heritably diverse lines 
may be isolated from among them, differing both in spine number and in diam- 
eter, and that these two characters are closely correlated. These heritably 
diverse lines resemble certain of the families that were reared from 'wild' 
specimens, and suggest that differences in the hereditary constitution of these 
wild specimens may have originated in the same way. 
During the course of this investigation several branches were studied that 
arose from what seemed to be "mutations." These are indicated in figure 2 
by the lines EM, ED, EDA , EDB, and EDC. These all appeared in the low 
line. Specimen EM had 8 spines and was only 18 units in diameter (fig. 1, D). 
Its parent had 10 spines and was 27 units in diameter. A large number of 
descendants (403) were reared from this small specimen, but it was found that 
the small diameter and lesser number of spines did not persist, but that the 
progeny of the fourth generation had regained the diameter and spine number 
of the low line from which EM was derived. The origin of this small specimen 
was therefore probably due to environmental conditions. Studies of this and 
certain other small specimens seem to show that it takes three or four genera- 
tions for the progeny of a very small specimen to regain the normal diameter 
and spine number of the parental line. 
