Raymond Pearl 
275 
correlation between the lengths of the individuals of conjugant pairs has about 
the same value in all the series studied, and it may be concluJed that the true 
value for the coefficient measuring assortative pairing in respect to length in 
Paramecium lies between "5 and '6. 
6. A considerable degree of homogamic correlation between the bi'eadths 
of the members of conjugant pairs was found. It is believed to be a result of 
direct assortative pairing with respect to length, and of the organic correlation 
between length and breadth in the individual. The same explanation is to bo 
given for the cross assortative correlations which were observed. 
7. The homogamic correlations are not due to any local environmental factor, 
tending to make all conjugants relatively alike, because if conjugants be paired 
together at random, all homogamic correlation disappears. 
8. The homogamic correlations are not greater in late stages of conjugation 
than in early, hence it is not probable that they owe their existence to any process 
of equalization of size due to the act of conjugation itself. 
9. Evidence is presented to show that the homogamic coi'relation arises 
through the necessity for the anterior ends and mouths of the two members 
of a pair to '"fit" reasonably well in a successful conjugation. The method by 
which this is probably brought about is discussed. 
10. One general result of the work is to indicate strongly that the relation 
of the Protozoa to evolution problems is not that usually assigned on the 
Weismannian hypothesis. With reference to the evolutionary history of the 
race, the conjugant individuals of the protozoan population appear to be in certain 
respects comparable to the germ cells of metazoan. In order that there should 
be any real evolutionary progress in the Protozoa, it seems to be necessary that 
there be definite changes in these differentiated conjugant individuals, or in what 
has been called in this paper the conjugant type. What Paramecium may acquire 
during a cycle of fission generations as a result of environmental action is of 
no consequence in the evolution of the race if at the end of the cycle the 
individuals must come back to a relatively fixed type (the conjugant) before 
starting on the next cycle. 
Appendix of Measurements. 
In this section are given the correlation tables from which the constants have 
been deduced. In order to facilitate reference to these tables those of a particular 
series are grouped together, and are designated with the letter of the series and 
the number of the table within the series. The order in which the tables are 
given for each series is as follows : 
1. The organic correlation tables for conjugants and non-conjugants. 
2. The direct homogamic tables. 
3. The cross homogamic tables. 
4. Random or other tables. 
For series B and D the measurements are given in fidl, as no correlation tables 
were formed for these very short series. 
