ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 
521 
outlines and blank interiors. This is frequently noticeable in photo- 
graphs of bacteria. 
By a proper choice of a screen, if a screen is required, a photograph 
should show any object as clearly as we can see it in the Microscope. 
Colour-sensitive plates may be said to be indispensable in the photo- 
graphy of rock-sections with polarized light. 
The yellow solution devised by Professor Zettnow, of Berlin, is used 
with much favour by many workers. It is composed as follows : — Copper 
sulphate, 175 grm.; potassic bichromate, 17 grm. ; water, 1000 ccm. 
The true function of the colour-screen should be to give definition 
and detail, not to increase contrast between the object and the field, as 
many observers seem to believe. 
(5) Microscopical Optics and Manipulation. 
Amplification in Micrometry.* — My attention has quite recently 
been drawn to this subject in connection with the celebrated Dr. Cronin 
case. It may be taken for granted that one cannot measure what 
he cannot see. But how high an amplification is necessary in a 
given case is a matter of much importance. In the measurement of 
blood-corpuscles in medico-legal cases the late Dr. Richardson advocated 
the use of a very high power, viz. a 1/25 or 1/50 objective. In my own 
measurements of blood-corpuscles I have, out of respect to authority, 
always used a high power, from 1500 to 1800 diameters. Decent 
experience has, however, qualified my views upon the subject, and in the 
case of the comparison of the ultimate subdivisions of a micrometer, 
ruled on metal, I am now of opinion that practically the same result may 
be obtained by the use of a 1/4 objective as with a 1/18 or 1/25. 
In December 1885, I commenced the investigation of the 1/100 mm. 
spaces of “ Centimeter A ” ; but was unable to finish it. Two series of 
measurements were then made with a Bausch and Lomb opaque illumi- 
nating objective, and a Bulloch filar micrometer. Eecently I have 
measured the same spaces with a Spencer 1/10 and 1/25, and with a 
Zeiss 1/18. The results of these measurements are given in the table 
below, each correction being the mean of from three to twelve readings 
of the filar micrometer at each end of the measured space. 
It will be observed that the agreement between the several series of 
the writer, and the results obtained by Prof. Hilgard is quite close, the 
discrepancy being practically insensible. 
Provided the amplification is sufficient to render the object to be 
measured of a sensible size, and to render the difference between the 
sizes of two objects visible, my own judgment is that little, if anything, 
is gained by the use of a power so high as to impair the definition, even 
though such impairment be but slight. Quite as much, in other words, 
is lost by impairment of definition as is gained by increase of amplifica- 
tion. The practical conclusion then is that no higher power should be 
used than is consistent with perfect definition. 
Diffraction Rings and Diffraction Spectra. — There appears to be 
still some confusion between the diffraction “ spectra ” of the Abbe 
theory and the diffraction bands or fringes and spurious lines seen 
* By Hon. Marshall D. Ewell, LL.D. 
I 1890. 2 o 
