A New Befradometer. By Dr. Boyston-Pigott. 299 
changes of colour in the central part, reckoning from the black 
spot of perfect contact (half-millionth of an inch thick) and the 
final evanescence of the last colour by slight movements of the 
screw. 
For the purpose of illuminating the point of contact of the 
" screw-lens " with the plate under examination, I inserted a 
minute right-angled prism behind the lens. This giving totally- 
reflected light, provided nothing adhered to the reflecting surface 
of the prism, afforded the means of making observations by ordi- 
nary daylight, and observing the rings of Newton, though very 
pale, by transmitted light. 
Another great advantage of the prism-lens attached to the end 
of the steel measuring screw, is the bright illumination of the con- 
tact surfaces, the contact of an opaque extremity of the screw being 
with difficulty ascertained. 
The whole method of finding the refractive index of a thin 
plate of a given refractive material resolved itself, then, into the 
best instrumental means for advancing a fiducial visible and illu- 
minated point truly and steadily through measurable intervals, and 
observing with a good microscope the precise position of evanish- 
ment in and out of focus, and determining the focal points under 
correct collimations. 
After many trials the following form was adopted : 
A steel cylinder very accurately turned between dead centres 
(i. e. the centres being fixed and the object revolving between 
them), about 5 inches long and f^o inch in diameter ; upon this 
screw-threads (very nearly 101 "3 per inch) were very patiently 
formed. The front part of the cylinder passes smoothly (at first 
air-tight like a piston) through a collar of brass, into which it had 
been very slowly and carefully ground (with the finest cutting 
powder and oil) ; at about two inches of the other end was formed 
a screw as described, with a very exact apparatus lent to the writer 
by a celebrated optician. 
It was found that in so delicate an operation as dealing with 
coloured films, touching any part of the instrument caused them to 
flash a new colour. It was necessary to obviate, then, all varying 
mechanical strain. Springs so common in micrometers to obviate 
" loss of time " were found to introduce, from their varying pres- 
sure, very variable errors : after a time I was compelled to abandon 
their use altogether. 
The constant force of gravitation and dead unvarying weight 
was now introduced ; and in order to compensate possible deviations 
in the true spiral form of individual threads (every one of which 
was carefully examined with a strong magnifying power), a nut 
was formed, so as constantly to embrace (unlike ordinary micro- 
meters of the usual form) precisely the same number of threads in 
