68 
Variation in Chilomonas 
variation for length and breadth in Chilomonas have values ranging roughly 
between 10'5 and 11 '5. There is no sensible difference in relative variability 
between length and breadth of body. For Arcella we have for the diameter of 
the shell a coefficient of variability of 10"27, and for the diameter of the " mouth" 
opening a coefficient of 13'66 (cf. Pearl and Dunbar, loc. cit). The mean value 
of the coefficient of variation in length of body for a series of Paramecia reared 
under various environmental conditions and including all told 4900 individuals 
is 8"45 (cf. Pearl, loc. cit). All these values cluster well together, and point to a 
value of roughly 10 per cent, for the coefficient of variation in size characters of 
this kind in Protozoa. 
Discussion of Results. 
It now remains to consider the meaning of the facts set forth in the preceding 
sections. These facts may be summarily stated as follows : comparing two popu- 
lations of the same local race of the flagellate Infusoriau GJiilomonas paramecium, 
one of which populations was living under the most favourable of environmental 
conditions and the other under the least favourable conditions, we find : 
(1) That in respect to absolute length and breadth of body and in shape of 
body as measured by the length-breadth index, the types of the two populations 
are significantly different. Those living in the least favourable conditions are 
smaller and relatively slenderer than the individuals in an optimum environment. 
(2) In respect to the characters dealt with, both populations are equally 
variable, and have their parts correlated to an equally high degree. There is a 
slight tendency for the individuals living in the unfavourable environment to be 
more variable and more highly correlated, but in view of the probable errors the 
differences cannot be said to be certainly significant. 
(3) The individuals living under unfavourable conditions vary symmetrically 
about their typical condition, while the group from the optimum environment 
exhibit an unsym metrical or skew variation about the type. 
(4) There is a sensible correlation between the absolute size of the body and 
its shape as measured by the length-breadth index. 
The first of these results is exactly what we should expect to find, on general 
grounds. There can be little doubt that one of the chief factors which induce 
saprophytes like Chilomonas to disappear from a culture is that the medium no 
longer furnishes proper food (either in amount or kind, or both). The Series A 
individuals are in all probability to be regarded as " starved." We should in 
consequence expect them to be smaller than the flourishing individuals of 
Series B. Similar cases of diminution in size in organisms living in unfavourable 
environments have recently been described by Dimon (1902) for Nassa, and by 
Warren (1902) for Hyalopterus. 
It is of more intei'est to find that in spite of the great change in the type 
between the two populations there is no marked difference in the amount of 
