262 A Biomctrleal Study of Conjugation in Paramecinm 
itself, and not a really significant thing at all. Our data for conjugants are 
obtained from conjugated pairs, and it might be maintained that the pro-con- 
jugants were simplj' a I'andom sample from the general population having equal 
variability with it. Then if we supposed that during the conjugation act itself 
there was a pronotinced tendency towards equalization in size of the two members 
of a conjugant pair, we might get a reduction in the variability of conjugated 
individuals as a result of the act of conjugation while the pro-conjugants were 
not less variable than the general population. Now data were especially collected 
to determine whether any such process of equalization occuri-ed during the act 
of conjugation, and anticipating results to be presented in detail farther on, it 
may be said that the most careful search has failed to find any evidence supporting 
this view. One has been compelled to conclude that such a process of equalization 
in size during conjugation does not occur to any appreciable degree. That there 
is absolutely no tendency to equalization cannot of course be said but, so far as 
the available data indicate, if any such tendency does exist it is far too slight in 
amount to account for the changes in variability and correlation observed. Further- 
more, such a tendency to equalization would not help us at all in explaining the 
reduced mean size of conjugants, and it seems altogether probable that the same 
set of causes are responsible for the changes in means and variabilities. 
The first problem with which this paper has to do was stated at the beginning 
(p. 215) to be "Is the portion of the Paramecium population which is in a state 
of conjugation at a given time differentiated in respect of type or variability, 
or both from the non-conjugating portion of the population living in the same 
culture at the same time ? " So far as our present material goes the answer to 
this question is unequivocally that there is such a differentiation in both type and 
variahility. The morplwlogical differentiation of conjugants is held to he probably 
due to fundamental physiological changes whicii, precede and lead to conjugation. 
The further question as to the cause of these physiological changes falls outside 
the scope of the present investigation. It is a subject on which there is urgent 
need for experimental work. 
We may turn now to the question of homogamy in the conjugation of Para- 
mecium. We have seen that when measurements are taken on the individuals 
of conjugated pairs, there is found to be a high degree of direct homogamic 
correlation. Our problem is to find, if possible, a reasonable explanation for the 
existence of this correlation. It is evident that there are at least two possibilities 
here which suggest themselves immediately. One is that there is no real pairing 
of like with like, but that after two individuals have become definitely united 
in a syzygy, an exchange of substance takes place until there is what we might 
call an equilibrium of internal pressures. In this process it might be conceived 
to result that the smaller individual of the pair becomes larger and the larger 
smaller, so that at the end both would be more nearly alike in size than they were 
at the beginning. If then we measured the coefficient of homogamic correlation 
after this had occurred, it might be found high without any real selection having 
