258 
A convenient method of proving the above 
is as follows: Darken the experimental room 
slightly by closing the windows, but leave a 
erack between the shutters, showing a strip of 
bright sky. Now gaze with one eye, say the 
right, on the crack until its image is branded 
on the retina. If we now turn about and look 
at the wall in various directions the after-im- 
age, of course, follows all the motions of the 
eye. Even if we shut the eye and look about 
the field of darkness the after-image follows 
all the motions of the eye. But if, with eyes 
still shut and looking straight in front, without 
changing the direction of looking, we press in 
the external corner of the branded eye the 
after-image does not move. It still remains di- 
rectly in front. 
We have given this experiment as most con- 
venient, but we may use a retinal brand pro- 
duced by the setting sun with still more con- 
spicuous results. 
One more experiment to show the behavior 
of after-images in the movements of the eye. 
Gaze with both eyes on the crack of the previous 
experiment, until its image is strongly banded 
on the vertical meridian of both retine. On 
turning about and looking at the wall the 
after-image is distinctly seen and follows with 
exactness all the motions of the eyes in looking 
about. But now converge the eyes until they 
look at the root of the nose. Of course, each 
eye changes its direction at least forty-five 
degrees, but the direction of the after-image 
does not change. It is still directly in front. 
The reason is that, while each eye individually 
changes its direction, the binocular observer 
looks in the same direction, though at a nearer 
point. Thetwo external images of the retinal 
brands cannot separate, as the images of an 
object do, because the brands are on corre- 
sponding points and have the same spatial 
representative and, therefore, must be seen 
single. This is the reason, as I have fully ex- 
plained in my volume ‘Sight,’ pp. 199 and 200, 
why after-images cannot be used to test the 
motions of the eyes by rotation on the optic 
axis in convergence, although they are such 
accurate tests in parallel motion. 
JOSEPH LE CONTE, 
BERKELEY, CAu., July 29, 1897. 
SCIENCE. 
(N.S. Vou. VI. No. 137, 
SCIENTIFIC LITERATURE. 
Leitfaden der Praktischen Physik, mit einem An- 
hange, Das absolute Maas-system. Von F. 
KouirauscH. Achte Auflage, B. G. Teub- 
ner, Leipzig. 
Each new edition of this book has been char- 
acterized by such considerable additions that 
the modest guide to elementary laboratory 
work in physics, which first appeared under 
this title more than twenty-five years ago, has 
become a reference volume of some five hun- 
dred pages, as valuable to the advanced worker 
as to the beginner. 
A laboratory manual should cover the entire 
field without undue specialization in any one 
direction, and without any omissions. Kohl- 
rausch’s book does this for physics more satis- 
factorily than any other. In fact, the manuals 
edited in America are too evidently, and often 
avowedly, nothing more than a compilation 
from the course given at the particular institu- 
tion at which the author is teaching, and hence 
possess all the peculiarities and limitations of 
the work at that laboratory, and too often are 
of slight value elsewhere, with different condi- 
tions and facilities. 
Kohlrausch sets forth not what is done at 
Wurzburg or Strasburg, or even at Berlin, but 
what ought to be done under ideal conditions 
in a complete laboratory. The book isa guide 
to the instructor as to what experiments should 
be performed, as well as to the student as to 
how they are to be performed. 
As regards the explanation of the operations, 
the author seems to have followed most suc- 
cessfully the principle laid down in the preface 
to this edition, ‘‘to carry the scheme, the ex- 
planation, and the setting-up of the apparatus 
for an experiment no further than is rendered 
necessary for the. successful operation of a 
laboratory attended by a large number of stu- 
dents.’’ He has achieved the happy mean be- 
tween allowing the pupil to flounder too long 
in the working out of an experiment, and giy- 
ing him such minute instructions that the neces- 
sity for originality of thought on his part is 
entirely eliminated. 
It is difficult to select parts of this book as 
worthy of special mention, and yet certain sub- 
jects are treated in a manner in especially 
