Raymond Pkarl 
265 
TABLE XX. 
Variatiuii and Homogavuj in Eatiij and Late Stages of the Conjugation Process. 
Character 
Constant iSeries F, Early 
Series F, Late 
Difference 
Tables 
Length of All Coiijvig; lilts 
15 >) 11 
11 1) 1) 
Mean 
S. 1). 
C. V. 
20S-4SC) + -983 
17-247 + -69.") 
8-272 ± -336 
214-051 + -966 
17-774 + -683 
8-304 ± -321 
+ 5-562 ±1-379 
+ -527 ± -975 
+ -032 ± -465 
Lcng-th of /I 
11 11 
11 11 
Mean 
8. D. 
C. V. 
209-103 ± 1-432 
17-767 ±1013 
8-497 ± -488 
214-497 +1-378 
• 17-927 + -974 
8-357 ± -458 
+ 5-394 +1-987 
+ -160 +1-405 
- -140 ± -669 
F 1 & F 3 
11 11 
1) 11 
Length of /> 
11 11 
11 11 ••• 
Mean 
S. D. 
C. V. 
207-874 +1-345 
16-689 + -951 
8-028 ± -461 
213-604 ±1-353 
17-609 ± -957 
8-244 ± -451 
+ 5-730 ±1-908 
+ -920 +1-349 
+ -216 ± -645 
F\ & F'3 
11 11 
11 11 
Difference between A) 
and B in Length* ] "' 
Mean 
11-794 
12-286 
+ -482 
Correlation between A} 
and B in Length ^ 
r + 
-6797+- -0307 
-6832 ± -0430 
-6212+ -0334 
-6223 ± -0471 
- -0585+ -0454 
- -0609 ± -0638 
2 & 4 
F 1 & F 3 
Number of Pairs 
70 
77 
material it was not possible to do mucli in this direction. The results so far 
as they went were in entire accord with what was found from the statistics. 
I was unable to detect any change whatever in the relative lengths of the two 
members of the pair after conjugation had occurred. 
Altogether, taking into account the results from the F series and from direct 
observations, together with the fact that we get closely accordant results for the 
homogamic correlations in all our series, though the Ann Arbor series were 
collected in a quite different manner from the Leipzig series, I think we must 
conclude that there is no evidence to support the view that these correlations 
arise because of any equalization in size during the conjugation itself It is on 
the other hand quite clear, of course, that the evidence here presented does not 
demonstrate that equalization does not occur immediately after union is effected. 
I can think of but one way in which to get absolutely conclusive evidence 
on this point, and that is to isolate, kill and measure pairs immediately after 
definite union has occurred. If the homogamic correlation in a reasonably large 
sample of this kind is no different from that of conjugants in general, it will be 
demonstrated that homogamy is not due to equalization after union. 
* Without reference to whether A or D is the larger. 
t Coefficient from symmetrical table in which each pair is entered twice. See p. '249. 
X Coefficient from table in which each pair is entered but once. 
