190 PROCEEDINGS OF SECTION A. 
being mounted in a dense medium, stand out clearer and sharper 
than those which have been plumbagoed and mounted in balsam.” 
The complete series of measurements upon which the calcula- 
tions in the extract just quoted are based have been kindly for- 
warded to me by Mr. Nelson, and are subjoined. | 
SCREW MICROMETER VALUES. OBJECTIVE 4 O11 IMMERSION APO- 
CHROMATIC, 1.43 N.A. ILLUMINATION 3/4 AXIAL CONE. 
001 Inch . 
Measurements. 
A—451.0 
B—450.0 
C—450.0 Difference of maximum intervals F and 
D—450.0 J from mean = + 1.25. 
E— 450.0 
F—451.5 Difference of minimum interval H from 
G—449.5 mean = — 1.95, 
H—449.0 
LORE This reduced to inches = : inch. 
J —451.5 
360,200 
450,25 mean reading. 
.01 mm. MEASUREMENTS. CONDITIONS SAME AS BEFORE. Two 
INTERVALS TAKEN FOR EACH MEASURE. 
A—354.5 
B—355.5 
C—356.5 Difference of maximum interval C 
D—355.5 fr —— BD 
rom mean = + 1.72. 
E—354.0 
I —354 0 Difference of minimum space H 
G—353.5 fr disteery 27 
3555 rom mean = — 1.27. 
T — 355.5 ; Reidy) ce 
These values are equivalent to +.097087 | 
9)3193.0 and —.072033 p respectively. 
2) 354.7 = .02 mm. 
177.38 = ,01'mm. 
Only one other matter remains for me to refer to, viz., my 
experiments with, or attempts to produce, diffraction gratings. 
To effect this, it was necessary to render the ruling machine 
completely automatic. This was conveniently done, experimen- 
tally, by means of an electric motor driven by current from four 
large Daniel’s cells. The capacity of my screw with the ratchet 
used cnly enabled me to rule some 13,000 lines without a stop- 
page. To rule this number of lines occupied some twenty 
hours, more or less ; 600 lines per hour gave good results. Un- 
fortunately, I possessed no suitable speculum metal for ruling 
upon, and therefore had to fall back upon plate-glass, which 
