240 
PROFESSOR K. PEARSON ON THE ^lATHEMATlCAL THEORY 
observers, and was practically absorbed, Ijeing invisible until a white strip was placed 
on the l)lack screen ; then the bright line was visible in a half-darkened room so long 
as it fell on the white strip. This white strip was 32’6 centims. long and 6 centims. 
broad ; it could be placed anywhere on a scale painted in red on the black screen, and 
quite invisible to the observers. 
The method of experimenting was as follows : The pendulum was brought to rest 
in a vertical position and the hammer was then moved up so as to touch the bell 
without resting against it, he., it did not change its position Avhen the pendulum was 
withdrawn. The line of light from the lantern now reflected from the stationary mirror 
fell on the scale on the black screen, which was adjusted by a fourth person so as to 
give a definite equilibrium position. The pendulum was now drawn back and clamped 
at a definite angle, which gave a very considerable range to the line of light. The 
three observers looking at the screen now saw ]io light at all, only 6 feet by 2 of 
black cloth. The fourth person now attached the strip of white card to the black 
cloth by aid of a drawing pin, so that its top coincided with any division on the scale 
known to himself only. He was thus able to make a record of the position on the 
stiij) occiq)ied hy the bright line when the hammer struck the bell. No doubt slight 
errors of adjustment occurred, but they were of much higher order than the errors 
of judgment. The equilibrium position of the beam was tested at the end of every 
twenty experiments, as well as the proper contact of the hammer. 
A series of positions for the bright line on the strip were selected so as to cover 
fairly well the possilde range, but the order in which these were taken was quite 
unknown to the observers. Of course, if the bell rung when the bright line just 
appeared on the strip, the latter was not moving as fast as if it rung when the bright 
line was just leaving the strip ; but the range of the bright line was very considerable 
compared with tlie length of the strip, and I doubt whether this difference of speed 
was sufficient to sensibly influence the judgment.* The shifting of the strip on the 
screen was only adopted after it had been found that to adjust the equilibrium 
position of the bright line between each experiment to a fresh position on the screen- 
scale would mean an expenditure of time which it was impossible to provide for. It 
was easy enough to shift the equilibrium position, but it required two persons, one 
at the pendulum and one at the screen, to adjust the equilibrium position to a definite 
point of the scale, and the one at the screen instructing the other at the pendulum 
how to raise or lower the line of light in adjustment was likely, besides the evil of 
tediousness, to have far more influence upon the judgment of the observers than the 
fairly small shift of the strip while it was hidden from sight by the body of the 
adjuster. 
Each observer was provided Avith a Avhite sheet of paper on which were tAventy 
* If the correlations of judgments had been solely due to an “ external cause ” such as this, then it 
Avould not have been possible for the correlation to have been sensibly zero bctAveen two observers, but 
finite between the third obserA^er and each of them. 
