PAPERS ON ZOOLOGY AND PHYSIOLOGY 305 
genetic pure line is devoid of one of its most important causes 
when exhibited in higher animals that reproduce sexually that 
is, germinal variability. 
When we subject to analysis those experiments that seem 
to be positive in regard to the effectiveness of selection, and 
contrary to our accepted ideas in regard to the permanency of 
« pure line or the genotype, do we find that they were so con- 
ducted as to be exempt from the criticisms such as Pearson, 
Harris, and Castle have raised in regard to the earlier ex- 
periments that gave negative results? We find that one of 
these common objections, that not enough selections were made 
can hardly apply to the work of Middleton with Stylonycha, 
for he states that in his main experiment on the average 39.86 
plus selections were made in the fast-selected lines; 34.36 
minus selections in the slow-selected lines. Yet in respect to 
the other points Middleton’s work is open, it appears to me, to 
the same objections raised in regard to the earlier experiments 
in pure lines. We find only a single character was studied, that 
this character, the fission rate, is a physiological one and does 
not lend itself as accurately to measurement as morphological 
characters, and is more or less influenced by environment. 
Probably the work of Jennings (1916) comes more nearly 
meeting the various requirements that the critics of the pure 
line work would impose than those of most others, In his 
work with Difflugia corona six different characters were used. 
These were all morphological characters, parts easily counted 
or structures easily measured. Further, they were little 
affected by environment, and their coefficients of correlation 
in regard to variation were ascertained. Jennings has applied 
finer methods of statistical study in regard to variation and 
correlation in a thorough manner to his work; also controls 
were run, checks made, and pedigrees kept. Thus it is made 
possible for other workers to examine his results from different 
angles, and formulate their own interpretations. The work 
of Jennings, it appears to me, must be conclusive in most re- 
spects for Difflugia corona. However, the selections in cer- 
tain cases were carried on for only a few generations, yet the 
mass of data given for such selections might cause one, unless 
making the closest scrutiny, to think that a very large number 
of generations had been obtained. Again, such strong positive 
