92 NATURAL SCIENCE [August 
In order to prevent the necessary length of the strip being 
excessive, the diagrams are made to overlap (Fig. 2). If the strokes 
are flat at the top it is frequently not easy to determine at sight 
which is the ‘ finish’ corresponding to a particular ‘ beginning.’ This 
is, however, discovered at once by measuring the length of the first 
or last stroke, from which the approximate distance of the ‘finish’ 
from any ‘beginning’ is known. A stroke thus individualised can 
then be measured up. The measurements taken are :— 
(1) Length of stroke in degrees: which is afterwards reduced 
to distance moved through by the handle of the oar. 
(2) Greatest height of diagram. 
(3) Area, measured with an Amsler planimeter. 
(4) The shape of the curve gives the style of stroke. 
Fig. 2. Part of an Indicator Card (two-thirds actual size). 
To determine the style of an oarsman under any particular 
circumstances, the idiosyncrasies of the strokes are eliminated by 
superposing several 
diagrams (Fig. 3), 
and then drawing a 
mean line through 
the result. In this 
way the ‘Charac- 
teristic Diagram’ 
is obtained. Simi- 
larly, in finding 
an oarsman’s power, 
Fig. 3. Deduction of Characteristic Diagram (actual size). 
the mean of several stroke measures is taken. 
Having now described the indicator, the diagram, and the 
method of treating it, some account will be given of the results 
obtained from twenty-seven experiments involving some 2000 
strokes of rowing. ‘The principal oarsmen experimented on are 
denoted by A, B, C, D, #, F, G, H, I. 
Fig. 6 represents a series of characteristic diagrams of rowing 
on sliding seats. AK was obtained with the earlier indicator, but 
is introduced here as the stroke-form differs from any others in no 
small degree. This stroke although very short is nevertheless very 
powerful. 
