302 Transactions of the Eoyal Microscojoical Society. 
Eeplacing the glass and again repeating the process the initial 
reading at the lower contact was now found not to be 0 * 0014, but 
I2 = 0-0045 (an extra tooW inch). 
Viewing an exceedingly minute scratch on the upper surface by- 
re-focussing upon it, the reading for thickness was 
T 0-0163 
Initial reading = 0-0045 
Thickness =0-0118 
A = 0 • 0042 elevation of image. 
V = 0' 007 6 distance of image from upper surface. 
Therefore !^ = I = ^T^f^ = 1-55 nearly. 
I obtained from Mr. Browning several square pieces of flint 
glass of varying density and colour, and found by a series of 
measurements : 
Value of IJ. for mean rays. 
A. Clear white flint 1-537 
B. Yellow and heavy 1 - 6626 
C. Yellowish, and very heavy 1-723 
D. Strong yellow, and the heaviest of them all .. 1-7555 
2. By reflected Light. — Far more interesting are the phe- 
nomena developed by reflected light, throwing the light laterally 
upon the inclined transparent plane within the microscope, the 
object-glass of which acts as its own condenser. The most beautiful 
colours are developed on contact, either with a film of air or 
kerosine. 
The central spot of final contact is of a grey-black, surrounded 
with its well-known succession of Newtonian rings of great beauty 
and perfection, flashing through numerous changes of colour for 
each thickness of film varied by the micrometer-screw. 
The instant of the formation of the black spot film (thickness 
0 • 000005 inch) determines the place of contact, i. e. the fiducial 
point of zero, with a precision scarcely equalled by any other 
known method of linear measurement by optical means. It can be 
readily obtained with great accuracy. This important point (the 
zero) of the observations depends not upon a thin spider line or 
engraved line, but upon the formation of a bold black circular spot, 
whose diameter is variable and dependent upon the curvature of the 
lens employed. Where great endurance is desired, a small plano- 
convex sapphire lens can be cemented to the prism at little 
expense. 
The behaviour of minute microscopic kerosine oil-drops per- 
