OF ERRORS OF JUDGMENT AND ON THE PERSONAL EQUATION. 
245 
With the aid of Mr. Horace Darwin I arranged a series of experiments which 
shonlcl test simultaneously the eye, the ear, and the hand, and thus give every 
opportunity for a variety of small causes to influence the errors of judgment. My 
plan was as follows ; A beam of light of very small breadth should traverse a white 
strip and at some part of its course a bell should sound. At this instant the eye 
should judge its position on the strip and the observer should at once divide a similar 
strip by a pencil stroke into parts in the same ratio as he considered the beam to 
divide the first strip. The instant at which the bell would sound was unknown to 
the observers, but it was so arranged that the exact position of the beam when the 
bell sounded could be easily ascertained by another person. 
Mr. Darwin constructed for us a pendulum,'^ consisting of a bar swinging on 
knife edges from an axis through its middle point. At either end of the bar were 
weights, so that by their adjustment very slow or very quick swings could be 
obtained. The pendulum could be released from rest at any angle from the vertical. 
Attached to the bottom of the pendulum was a small bell, which struck a very light 
hammer as it passed through the low^est point of the swing. This hammer was easily 
adjustable and was pulled upright by a string between each experiment, being 
knocked over by the transit of the pendulum. A mirror swinging about a horizontal 
axis had a strut attached to this axis and perpendicular to the plane of the mirror. 
This strut rested on a saddle (a) attached to a similar strut perpendicular to the 
pendulum liar at its axis. By shifting the saddle on the strut the mirror could be 
made to swing through a very small or a fairly large angle, whatever might be the 
amplitude of the pendulum. The whole object of this arrangement was to oljtain a 
great variety of speeds and ranges for the line of light on the strip and so ascertain 
how far these conditions interfered with the independence of judgment which, 
a jn'iori, I supposed must exist. When the first series of experiments showed sub¬ 
stantial correlation in judgment, although the bright line moved in the same manner, 
no further series were then undertaken to determine how this correlation would ]je 
varied by differences of speed and range. Correlation existed wJien all the circum¬ 
stances were alike except the position of the bright line on the strip when the bell 
sounded. I believed that I had evidence that the source of the correlation was 
rather in the observer than in the likeness of condition for each observer in each 
individual experiment, t and this was too subtle to be analysed 1jy simply varying 
speed and range. 
A beam of light from an electric lantern was intercepted by a screen having a thin 
horizontal slit placed in the slide groove ; the selected jiart of the beam reflected from 
the pendulum mirror was received on a black screen at some distance from the 
* See figure 1, yj. 249. 
t I hope later to take a further series of estimates, but it must l)e remembered that 500 experiments 
are the least we can make for our present purpose, and that with varying conditions the labour of making 
them will be greater, while the exhausting work of reduction will not be lessened, 
