Degree of Injury and Rate of Regeneration 517 
greater injury did not recover their ordinary activities as soon as 
those with the lesser injury. 
10 Departure of the living conditions from the optimum. If 
the two compared groups are exactly alike in original choice of 
animals and in treatment there is still a source of error that must 
not be neglected. The object in view is the determination of 
the effect under optimum living conditions. If the living condi- 
tions are poor it is evident that one group may suffer more than 
the other. Under poor conditions such as foul water, presence 
‘of pathogenic organisms, insufficient food, etc., it is natural to 
suppose that an individual with a greater degree of injury is 
likely to suffer much more than one with a lesser injury. The 
same may be said of differences inherent in the experiment, such 
as loss of blood, nervous shock, and the greater difficulty experi- 
enced by individuals with the greater injury in seizing and hand- 
ling their food. ‘Therefore under identical treatment of the two 
compared groups the series with greater injury is at a disadvantage 
whenever there is any departure from optimum living conditions. 
11 Relation of the degree of injury to the optimum degree. It 
is obviously true that with a degree of 1 injury so great that the 
animal is barely able to survive the operation a rapid rate of 
regeneration cannot be expected. Accordingly near the upper 
limits of degree of injury a decrease in rate is found. The general 
question therefore resolves itself into the location of the optimum 
degree of injury. Is it coincident with a very low, a very high 
or with a medium degree of injury? If we compare two groups 
of animals with different degrees of injury both of which are 
near the upper limit of degree the result will be different from 
that obtained near the lower limit of injury. Furthermore if the 
lower injury is below and the upper injury above the optimum 
there may be no difference in rate of regeneration. In most 
cases the present experiments deal with medium degrees of 
injury. 
12 Individual variation. In most experiments differences 
classed as individual variations are due not to any inherent differ- 
ences in the individuals themselves but to uncontrolled differences 
in treatment or in age of the animals. The principal sources of 
: 
