ON THE REFRACTIVE INDEX OF FLUIDS. lil 
those of organic origin, causes the diameter of the air-bell 
after a certain time to become notably diminished. We 
must not therefore rest satisfied with a smgle measurement, 
but take each set of dimensions again and again, and use in 
the final calculation the mean results of all. 
Let us now suppose : 
Distance between object and air-bell 
Diameter of the object 
Diameter of image > 
Diameter of the air-bell 
I fl Wl 
a 
b. 
c. 
d 
The index of refraction will then be obtained by means 
of the following formula, for which I am indebted to my col- 
league, Van Rees: 
i b—c)d 
= = + Ae ee 
4ac 
But as c may be regarded as infinitely small when com- 
pared with 4, 
1 1 bd 
as Vv; * Fae 
The use of this formula enables us to deduce the refractive 
index with certainty to the third decimal place, but only 
when the above-mentioned conditions are attended to, and 
the final mean of several measurements is used as the basis 
of calculation. When a thicker object-plate, and especially 
when a larger object is used, there arises the necessity for 
different corrections, which cannot be neglected, and which 
render the computation troublesome and its result less 
exact. 
If, as in the arrangement which I have recommended, 
a =100, b= 20, or, in general, if 4 3 then, 
n=5 a: Vite 20 
Some results obtained by this method may be here sub- 
joined, in order that the reader may form an estimate of the 
degree of accuracy of which it is susceptible.* 
* The first two examples taken from Harting are all that we give 
here. (T.) 
