208 A Biometrical Study of Conjvijatiou in Pwameciiim 
not more than 10 per cent, of the total difference in length between the 
individuals of conjugant pairs is due to difference at the anterior ends. I do 
not present measurements in support of this statement for the simple reason 
that the magnification used in the measuring was not sufficiently high for the 
accurate determination of this " anterior end difference," so small is it in the great 
majority of cases. I think, however, that the statement that in definitely con- 
jugated pairs of Paramecium the mouths and anterior ends of the two individuals 
approximately coincide will not be questioned by anyone who has ever examined 
closely large numbers of such pairs. Finally if {d), and the comments which have 
just been made regarding it be true, then (c) must necessarily be true, because if 
we observe that all successful conjugants have the mouths and anterior ends 
approximately coinciding, then clearly all pairs which did not fulfil these con- 
ditions when the individuals came together by chance, must have either separated 
or died. The probability of separation being what actually happens is strongly 
indicated by the observations of Gruber and Jennings cited above. Similarly {f) 
necessarily follows from (d), if there be positive correlation between total length 
and distance from anterior end to mouth, the existence of which can hardly be 
doubted. 
All things considered, I think that, at least till further evidence is forthcoming, 
the conclusion is justified that the observed homogamic correlation arises as a 
result of the factors set forth above, (a) to (e). On this view the real essence of 
the matter is the necessity for anterior end.« and mouths to " fit " reasonably well 
if conjugation is to be successful. The mechanism by which individuals which do 
"fit" are brought together is readily to be accounted for by well-established 
factors in the " action system " of the organism, and involves no assumption of 
consciousness or any vitalistic hypothesis. 
It is the belief of the writer that a considerable degree of homogamy will 
probably be found to exist in the conjugation of other Protozoa than Paramecium, 
though of course no other form has yet been investigated in connection with this 
matter. An observation recorded by Calkins*, concerning the infusorian Lionotus 
fasciola, is of interest in thi.s connection. He says that "In conjugation a large 
form unites with a smaller one, the mouth parts being united." Such a conjugant 
pair is found in his Fig. 33. It would be very interesting to study this case 
biometrically, in the light of the results on Paramecium. There would seem to 
be two possibilities in the case of Lionotus ; one that there is a dimorphic condition 
of pro-conjugants paralleling sexual differentiation in higher forms, either with or 
without positive assortative pairing : the other that thei'e is a negative assortative 
pairing in the conjugation. In any event the case seems well worth investigating. 
According to the explanation which has been given for the origin of the 
homogamic correlation in Paramecium, the assorting in the pairing is held to be 
concerned only with the lengths of the individuals. If this explanation be true 
the homogamic correlation between the breadths of the two individuals of a pair 
* Bulletin of the U. S. Fish Commission for 1901, pp. 413—468. 
