306 ILLINOIS ACADEMY OF SCIENCE 
parent-progeny correlations as were obtained in family 305, 
were found in a family that appeared abnormal in several re- 
spects. 
But has the pure line theory been overthrown by these re- 
cent experiments? I think far from it. The strongest evi- 
dence against the concept is that found in Jenning’s work with 
Difflugia. In fact, I believe that this is the one piece of work 
that shows definite positive results in regard to somatic change 
through selection (in pure lines), that is not open to the seri- 
ous objections raised against the earlier work in this field. Jen- 
nings himself, apparently does not claim that his results neces- 
sarily overthrow his conclusions based on his earlier work with 
Paramecium, or the pure line theory as it applies to higher 
plants or animals that reproduce by self fertilization or par- 
thenogenesis. In fact, he states that there is a srong probabil- 
ity that selection was effective in Difflugia because of a peculiar 
role the mechanism of inheritance plays in these forms. In 
his discussion we read: “It would appear therefore that the 
substances determining the hereditary characters may be dis- 
tributed with less accuracy than in higher organisms, so that 
the two products of fission may often receive parts that are not 
equivalent. As a result, the two products of fission would 
differ in hereditary characters; and in time diversities of 
strains would be brought about such as are described in the 
present paper. The possibility that this is the state of affairs 
is entirely open, so far as our present knowledge is concerned.” 
The writer in his work with Aphis avenae (Ewing, 1916), 
attempted, as did Jennings, to meet the objections that had 
been raised in regard to the earlier experiments in pure lines. 
My results, I believe, in the main, have met these objections. 
There is one phase of the work, however, that was not reported 
in the recent summary of the experiments with Aphis avenae 
in the “Biological Bulletin” for August (1916); that is, the 
statistical study of selection, regression, and the parent progeny 
correlation. The data for most of these were collected and in 
part put in manuscript form at the time the published results 
were first written. In order to forestall criticisms of the omis- 
sion of these features in the work with Aphis avenae, and in 
order to help us obtain a more intelligible interpretation of the 
facts in our pure line work, I submit them in this paper. 
