Raymond Pearl 
215 
After repeated failures I finally succeeded during the past summer (1905) 
in getting sufficiently abundant material of conjugating Paramecia in the 
Zoologisches Institut at Leipzig. This additional material made it possible to 
extend considerably the scope of the work beyond what had been planned when 
the investigation was begun. As will appear later the Leipzig material fully 
confirmed the results gained from the earlier A A series. 
The main problems with which the present paper specifically deals may be 
stated as follows : 
1. 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 ? 
2. Is there any tendency for like to pair with like (" assortative mating ") in 
the conjugation of Paramecium, and if so, how strong is this tendency ? 
At this point I wish to acknowledge gratefully my indebtedness to those who 
have in various ways aided me in this work. To the officials of the Carnegie 
Institution I am indebted for a grant in aid of this and other biometric work now 
in progress. It is a pleasure to express my heartiest thanks for this aid. To 
Professors Carl Chun, and Otto Zur Strassen I am indebted for the numerous 
facilities of the Zoologisches Institut at Leipzig, which were so freely and kindly 
placed at my disposal during my stay there. The work was brought to completion 
in the Biometric Laboratory of University College, London, and it is a pleasure to 
acknowledge my great debt to Professor Karl Pearson for helpful advice and 
kindly criticism. 
II. Material and Methods. 
The material on which this paper is based is comprised in eight* series of 
measurements including altogether 1894 individual Paramecia. The cultural 
history of the different series is as follows : 
Series AA, Ff; and Fj^. The individuals in these series were contained in the 
mounted slides in the Zoological Laboratory of the University of Michigan 
* Wote aided Nov. 10. In his bef ore-publication criticism of this paper Mr J. J. Lister {Nature, 
Vol. 74, p. 584) suggests that I have mixed and lumped together these different series and that in 
consequence all my results are invalid. The reader of my paper will be able to judge of the correctness 
of Mr Lister's suggestion. I shall be very glad to have a specific instance where I have combined two or 
more series pointed out. I have always supposed it to be a fundamental axiom regarding the worth of 
scientific evidence, that the greater the number of pieces of independent evidence there are leading 
to the same conclusion by so much the more certain does that conclusion become. Acting on this 
principle I spent a great deal of time getting data from as many independent conditions as possible, and 
when, as appears in the paper, they all led to the same result, I began to feel that that result was the 
correct one. According to Mr Lister's new epistemological doctrine this conclusion was wrong and it 
would have been far better to have measured only one series of individuals. As a working biologist 
I cannot but feel that Mr Lister ought in justice to his colleagues to issue a definite statement as 
to whether in his own investigations he follows the principle that the evidence of one witness is more 
trustworthy than that of several independent witnesses. R. P. 
28—2 
