ON THE REFRACTIVE INDEX OF FLUIDS. 109 
point, the optical centre, is in the compound lens, and at a 
depth varying with the thickness of the layer and refracting 
power of the fluid which constitutes a part of the plano-con- 
vex lens. Henceit is hardly possible to measure the distance 
of the object with the degree of accuracy required for the 
subsequent calculation. 
Il. Harting’s Method. 
The following method may be followed with any micro- 
scope and without the addition of special apparatus ; and 
although comparatively limited in its application, which is _ 
restricted to certain fluids, it affords indications of extreme 
exactitude when due attention is paid to the manipulation. 
It is free from the above-mentioned disadvantages of 
Brewster’s method, and has the further recommendation that 
a very small quantity of fluid is required for each observation, 
even a few milligrammes amply sufficing for the determination 
of the index of refraction. 
This method is founded upon the different dimensions of 
images of the same object placed at like distances from air- 
bells of like size in fluids of different refractive power. That 
this difference in the size of the images is rather considerable, 
the following examples will show: 
Water . n=1:336 Diameter of image = 1000 
Sulph. acid . > L416 ah 3 =e fo 
Canada balsam ,, 1°504 95 es == 5o2 
To enable us to calculate the index of refraction, it is 
necessary that there should be— 
Ist. A thin layer of the fluid between two plates of glass 
with parallel surfaces ; also some air-bells in the fluid to act 
as dispersing lenses and form images of objects situated 
beneath them. To prepare the fluid for the observation, let 
a drop be placed on a thin glass plate, and some air-bells 
formed by blowing air into the fluid through a small glass 
tube drawn out very fine in the blowpipe flame. <A small 
ring of paper is next laid round the drop, and on its surface 
a thin glass covering-plate is placed. It will then be found 
that some of the air-bells have lost their spherical form and 
are consequently useless for the purpose in hand. This will 
be quickly seen from the distortion of the images which they 
form—there wi!l, however, always be present a few which will 
exhibit correct and sharply defined images. The thickness 
of the glass object-plate should not exceed 3th of a milli- 
metre = ;45th of an inch nearly: for a thicker plate would 
exercise on the course of the rays an influence which in the 
