~ 
oO 
rolling melted glass. The best are found in the thickest 
patent plate glass, caused by the expansion of minute 
quantities of incandescent gas or impacted vapour. Some 
curious phenomena are connected with these spherical 
cavities, of a physical character, which, however, are out of 
place here. It suffices to state that enclosed between the 
surfaces of glass, ground to such accuracy as to show Newton’s 
rings under slight pressure, or only under the weight of a 
second layer, some of these exquisite little gaseous lenses 
exhibit images of great precision and clearness. They differ, 
however, essentially from the solid spherical lenses by the 
focal point being placed within the sphere, instead of without 
it, as in glass sphericles; and in the enormous aperture 
embraced in the focal picture (nearly 180°, the limit). 
Remembering that when a pencil of parallel rays passes 
through a denser into a rarer medium (supposing common air 
were enclosed), the focal point for a refractive index p= 
would be found to lie upon the posterior curvature of the 
minute hollow sphericle, i.e. on the surface furthest from the 
eye of the observer. If the contained gas or air were more 
attenuated or w diminished, it would approach the centre. 
This appears from the formula 
ep 
Brinn 
The field of view presented is generally three fifths of the 
diameter of the gaseous sphericle, and is independent of the 
aperture of the objective, whilst change of aperture has a very 
surprising effect upon the visual characters of the solid glass 
sphericle, however small, as also upon cylindrical refracting 
fibres. This change, so decided and important in minute 
research, has been a cause of much surprise and pleasure to 
the writer, as it appears to open a new mode of changing 
and selecting definition under novel conditions. 
Experiment 1.—Placing minute glass sphericles 1-1000th 
of an inch in diameter, no black outline is visible with 
an aperture exceeding 83°. As the aperture is reduced a 
black annulus appears and broadens continually (figs. a, 4, c, 
d, e), and for rows of sphericles in contact and very small 
their outlines are similar, obscure or developed according to 
the aperture and obliquity of light used. 
Experiment 2.—Spun glass threads 1-3000th in diameter. 
Jet black bands, similarly, are developed, only with an aper- 
ture of 83° and under, and their breadth increases rapidly as 
the aperture is reduced (figs. 2, 3, 5). 
Remarkable is the result that large aperture destroys the 
Bs 3 (writing here - for »): 
