226 A Biometrical Study of Conjugation in Paramecium 
TABLE II. 
Variation in the Length of Paramecium. 
Series and Class 
Series A, All Conjugants 
Selected Ancestry, 300 hour, controlt 
Series A, All Noii-Conjugaiits 
Selected Ancestry, 200 hour, controlt 
Series F^, All Conjugants ... 
Series C, All Non-Conjugants 
Selected Ancestry, 300 liour, sugart 
Series Fj^, All Conjugants ... 
Series F, All Non-Conjugants 
Selected Ancestry, 200 hour, sugart 
Series AA, All Conjugants ... 
Selected Ancestry, 100 hour, controlt 
„ „ 100 „ sugart 
Simpson's Series* 
Ann Arbor Seriest 
Total 
Mean 
168-143 
176 015 
182-200 
189-976 
207- 080 
208- 489 
209- 356 
213- 340 
214- 051 
214-470 
217-380 
217-675 
221-800 
224-980 
229-050 
246-080 
± -522 
± -621 
± -480 
± -724 
± -518 
± -983 
± -906 
± -601 
± -966 
±1-074 
± -592 
± -645 
± -587 
± -533 
+ 
+ -983 
Standard 
Deviation 
11 -212 ±-369 
13 094 ± -439 
15 -917 ± -340 
15 -549 ±-512 
17-171 ± -366 
17 -247 ±-695 
19 -099 ±-641 
19-936 ± -425 
17 - 774 ±-683 
18- 291 ±-759 
19- 630 ± -419 
19-112 ±-456 
19-457 ±-415 
17 -680 ±-377 
19-152± 
23-041 + -695 
Coefficient of 
Variation 
6 - 668 ±-220 
7 - 439 ±-251 
8- 736± -188 
8 -185 ±-271 
8-292±-178 
8 - 272 ±-336 
9 - 123 ±-309 
9-345 ± -201 
8 -304 ±-321 
8 - 529 ±-357 
9 - 030 ±-194 
8 -783 ±-211 
8-772±-189 
7- 859 ±-169 
8- 361 ± 
9- 363+ -285 
so that, as a result, the coefficients of variation, measuring the amount of variation 
relative to size, cluster well together in value. It may be concluded, I think, 
until equally extensive series showing a different result are forthcoming, that the 
usual or "normal" value for the coefficient measuring variation in the length of 
Parameciam caudatum is 8 — 9 %• The good agreement in the values of the 
coefficients of variation for the different series is very satisfactory, and is something 
which probably no biologist would have predicted before measurements were made. 
One has been accustomed to think that Paramecium because it is a soft-bodied 
creature is likely to show great and altogether irregular fluctuations. As a matter 
of fact Paramecium is distinctly less variable in size than is, for example, Arcella\ 
(coefficient of variation = 10-2676 or the crab, Eupagurus prideaiiasi^ (coefficient 
of variation for carapace length = from 12 to 19°/^), or the ophiuroid, Ophiocoma 
nigra\\ (coefficient of variation > 20 for both disc-breadth and arm-length), all 
which organisms have a more or less firm exo-skeleton. Furthermore it is perhaps 
of some interest to note that the degree of variation in length of Paramecium is 
of the same general order of magnitude as that in the capacity of the human skull. 
There can be no doubt I think of the substantial homogeneity of each of the 
series. Especially does this impress itself when we compare the variability of 
* Biometrika, Vol. i. p. 405. 
t Pearl, R., and Dunbar, P. J. : Seventh Report, Michigan Academy of Science, pp. 77 — 86, 1905. 
t Pearl, R., and Dunbar, F. J. : Biometrika, Vol. ii. p. 327. 
§ Schuster, E. H. J. : Biometrika, Vol. ii. p. 195, Table viii. bis. 
II Mcintosh, D. C: Biometrika, Vol. ii. pp. 403—473. 
