1905.] 



Study of Conjugation in Paramcecivm. 



381 



each series of measurements with entirely uniform results. The coefficients, 

 measuring direct homogamy in random pairs of (a) conjugants; (b) non- 

 conjugants; or (c) pairs in which one member is a conjugant and the other 

 a non-conjugant, are uniformly zero within the limits of the probable errors. 



It might be maintained that since at different points in the culture and 

 at different times the environment no doubt differs slightly, there would be 

 a corresponding differentiation of the Paramsecia in each local unit of the 

 culture. Then even though the pairing were really quite at random in 

 each locality, yet if the records for several such localities were mixed, a 

 spurious homogamic correlation would arise. Now it is clear that if the 

 observed homogamic correlations were spurious and due to this " local 

 differentiation " factor, we ought to get sensibly as high values if we consider 

 as a pair the two individuals lying nearest in the field of view to each pair 

 of conjugants measured. Such pairs will have come from the same environ- 

 ment and have been killed at the same instant as the actually conjugated 

 pairs. The homogamic correlations for these pairs of individuals, lying- 

 nearest to conjugant pairs, have been determined and found to be sensibly 

 zero, taking into account their probable errors. This I take to prove that 

 the observed homogamic correlations are not to be explained as due to local 

 differentiation within the culture. 



6. The existence of a high degree of assortative pairing with respect to 

 length can be shown in another way. We may consider as a character of 

 conjugant pairs the difference in length between the two members. Forming 

 frequency distributions of this difference for (a) conjugated pairs ; (b) random 

 pairs of conjugants ; and (c) random pairs of non-conjugants (the two 

 individuals lying nearest each conjugant pair measured), I find that the 

 mean difference in microns is from two to three times greater in the 

 random pairs — whether of conjugants or non-conjugants — than it is in 

 the actually conjugated pairs. In more than 60 per cent, of the conjugated 

 pairs, the two members differ in length by less than 8 microns. There is 

 a much greater " scatter " of the variates about the mean in the case of the 

 random pairs as compared with the conjugated pairs. These results for the 

 difference distributions are, of course, what would be expected if there is 

 a high degree of homogamy in respect to length. 



7. In order to determine whether the observed homogamic correlations are 

 due to a real " assorting " in the pairing, or, on the other hand, arise from 

 some process of equalisation in size occurring after union has taken place, 

 the following test was made : — Two series of conjugants were measured and 

 the homogamic correlations calculated. In the first of these series were 

 included only pairs which were known, on the basis of their nuclear condition, 



