I.—PHYSIOLOGY. isi 
distances, can one interpolate between them for an intermediate distance, 
or extrapolate for a distance greater or less? Obviously the answer to 
such questions depends upon the factor which in general terms we designate 
fatigue. Fatigue, however, is a very indefinite and inexact expression ; 
it is necessary to define it quantitatively before we can employ it in a quan- 
titative discussion such as this. There are many varieties of fatigue, but 
of these only a few concern us now. There is that which results in a short 
time from extremely violent effort: this type is fairly well understood ; 
there is the fatigue, which may be called exhaustion, which overcomes the 
body when an effort of more moderate intensity is continued for a long time. 
Both of these may be defined as muscular. Then there is the kind 
which we may describe as due to wear-and-tear of the body as a whole, 
to blisters, soreness, stiffness, nervous exhaustion, metabolic changes and 
disturbances, sleeplessness, and similar factors, which may affect an 
individual long before his muscular system has given out. Of these three 
forms of fatigue the first one only is as yet susceptible of exact measure- 
ment and description. The second type may quite possibly come within 
the range of experiment at no distant date. The third type is still so 
indefinite and complex that one cannot hope at present to define it accur- 
ately and to measure it. Undoubtedly, however, all these types of what 
we call ‘fatigue’ influence—indeed, determine—the results which are 
to be presented. 
Presentation of Data. 
The data will be exposed throughout this discussion in graphical form, 
and in every case but one (fig. 5) the quantities plotted are the speed 
as ordinate and the time, or some function of the time, as abscissa. The 
_ reason for taking the time occupied in a race as one of our variables is 
simple; the problem before us, physiologically speaking, is, clearly, how 
long can a given effort be maintained ? The length of time is given by the 
abscissa as the independent variable ; the magnitude of the effort, or some 
function of it, as represented by the speed (that is, by the average speed 
over the race considered), is given as ordinate. It will be shown below, as 
Kennelly indicated in his paper, that the ideal way to run a race, possibly 
not from the point of view of winning it, but certainly from that of breaking 
_ the record for the distance, is to run it at constant speed. In those 
performances which have attained to the dignity of a world’s record 
itis unlikely that this criterion has been to any very large degree neglected. 
Apart, therefore, from the fact that there is no speed of which we have 
any record except the average speed, we are probably not far wrong in 
using the average speed as a fairly exact measure, or at any rate as a 
function of the effort involved. 
In one case only (fig. 6) the time occupied in the race has been given 
on a logarithmic scale: no great virtue attaches to the logarithm, but 
ff 75 yards and 100 miles are to be shown on the same diagram in a read- 
able form it is necessary somehow to condense the abscisse at the longer 
times. As a matter of fact, from the standpoint of an athlete, one second 
in ten has the same importance as ten seconds in a hundred, as a hundred 
seconds in a thousand; in this sense, therefore, a logarithmic scale of 
_time most truly represents the duration of an effort. Such a scale, how- 
ever, has not been used for any ulterior reason, but only, as in fig. 6, to get 
all the available data on to one diagram. 
