426 
looking through the telescope with one eye, 
until the cross hair, which turns with the 
disc, seems to be vertical, or horizontal, as 
the case may be. A piece of ground glass 
in front of the object glass allows light to 
have access to the inside of the telescope, 
but keeps out all disturbing images. An 
assistant then reads off on the scale, and 
records, the deviation from verticality or 
horizontality, and a new setting is made. 
The mean of a dozen settings, ranging in 
the cases of some persons over as much as 
two degrees, suffices to show whether the 
observer has a decided bias and to measure 
it fairly well, if it exists. The subjoined 
table (I) gives the results of the observa- 
tions of ten persons, five of whom are in- 
structors in Harvard University. The last 
observer in the list is a lady. The first 
two and the seventh observers are astig- 
matic. The first two columns give in de- 
grees the means of the deviations from 
level of a number of horizontal settings of 
the cross hair made with the left eye and 
the right eye respectively. The next two 
columns give corresponding means of ver- 
tical settings. In all the cases the positive 
sign implies a rotation from the correct 
position in clockwise direction. The hori- 
zontal determinations of all the observers 
except one, illustrate a fact, familiar to some 
astronomers and physicists, that a cross 
hair which seems level to the right eye of 
@ person generally needs to be rotated 
slightly in counter-clockwise direction to 
suit the left eye. The third line of the 
third column shows the only large bias that 
Ihave found. In a largenumber of cases, as 
the table shows, the deviation of the mean 
was practically nothing. When the ground 
glass in front of the object glass wasremoved, 
a plumb line made of silk fibre appeared 
in the field of the telescope, and many 
settings have been made, by different per- 
sons who have used this apparatus, which 
could not be in the least improved when 
SCIENCE, 
[N. S. Von. X. No. 248. 
the plumb line became visible. Observa. 
tions made by the same person on differ- 
ent days show slight differences in general, 
but unusual fatigue sometimes introduces 
a temporary bias of half a degree or per- 
haps a little more. Ihave not been able to 
discover from settings made with un- 
screened eyes in a room in which many 
horizontal and vertical lines may be seen, 
that observers are much affected by after 
images, or by outside standards not nearly 
superposed upon the line to be set, provided 
that the head is erect. 
TABLE I. 
© | Mean deviation of hori-|Mean deviation of vertical 
a zontalsettings made with settings made with 
4 the left the right the left the righ 
4 eye. eye. eye. eye. 
1 + 0°.6 +1°.2 —2°.3 +0°.2 
2 +0 .1 —0 .9 —0 2 —1 .4 
3 —1.5 +0 .8 —5 .2 —0 .5 
4 —0 .2 +0 .6 +0 .0 +0 .6 
5 +0 .1 +0 .8 +0 .0 +0 .8 
6 —0 .6 +0 .6 —0 .2 +0 .1 
7 —0.1 +0 .8 +0 .4 +0 .5 
8 —0 .2 +0 .7 —0 .2 —0.1 
9 +0 .0 +0 .4 —0 .2 —0 .3 
10 | —0 .7 +0 .0 +0 .0 +0 .0 
To find out whether the accuracy of the 
results depended mainly upon the horizon- 
tality of the seat occupied by the observer, 
a long series of settings were made by a 
person who sat in a chair which rested 
upon a platform inclined about 5°, first to- 
wards the left and then towards the right. 
The feet of the observer were placed upon 
the rungs of the chair. The exact agree- 
ment of the deviations, as shown in the 
following table (II) is, of course, accidental, 
but it is clear that the inclination of the 
seat did not appreciably affect the judgment 
of this person. 
TABLE II. 
Seat inclined to the left. Seat inclined to the right. 
Horizontal | Vertical set- Horizontal | Vertical set- 
settings. tings. settings. ings. 
Left | Right| Left | Right) Left | Right| Left | Right 
eye. eye. eye. eye. eye. eye. eye. eye. 
—0°.3|+0°.7|+0° 5|+0°.7|—0°.3| +-0°.7| + 0°.5|-+.0°.7 
