During the polishing the pressure should be very moderate, or the lenses will partially sink into the surface of the polishing tool, and become rounded at those parts of the edges which are unsupported by the neighbouring lenses. This evil may be partially remedied by cutting off a portion of the circumference in the manner alluded to on page 1229. But in order to avoid the rounding as much as possible, the more accurate the lenses are required to be, the less the pressure that is employed in rubbing them on the cloth polisher.

The edges of the lenses are finally ground circular, and of course the axes of the lenses should when put into the tubes of the instruments, be perfectly parallel with the axis of the tubes; to attain this they are cemented upon a chuck in the lathe, and before the cement has set the lathe is put in revolution, and the reflection of any fixed object such as a candle, or a bar of the window, is watched, and the lens is adjusted until the image appears strictly stationary, notwithstanding the revolution of the lens, and which shows the axis of the lens and that of the mandrel of the lathe to be in agreement. The edge is then ground circular with a piece of brass supplied with emery and water. The piece of brass being placed beneath the lens, and gradually elevated by a screw tapped through one end, while the other rests upon any convenient prop on the lathe bearers.

Concave lenses are ground and polished in the same manner as convex lenses, except that they are fixed in the concave tools and ground upon the convex, which as before mentioned is always the lower tool, when several glasses are operated upon together.

In Mr. C. Varley's lathe for grinding and polishing lenses and specula,* instead of the lower tools being mounted upon a fixed post, they are mounted upon a revolving axis, placed vertically. This considerably expedites the process, which is conducted in exactly the same manner in all other respects, but the necessity for walking around the lower tool is removed. It is however generally considered that the method of grinding lenses of medium and large sizes, with a tool mounted on a rapidly revolving axis, is less accurate than when the tool is fixed; and that when circular motion is given to the tool, it should be so slow as only to give change of position, leaving the abrasion to be effected principally by the elliptical or circular strokes.

In manufactories where large quantities of common lenses are ground and polished, these operations are principally effected by machinery. The block of lenses is mounted upon a slowly revolving axis, placed vertically, and the upper tool has an eccentric motion given to it, by means of a small crank fixed on the lower end of a second vertical axis, that is placed a little on one side of the central line of the lower axis. A pin, fixed in the center of the back of the upper grinding tool, enters a socket in the crank, and the revolution of the latter causes the upper tool to describe small circles, which, combined with the slow revolution of the block of lenses, causes every point of the grinder to describe epicycloids upon the surface of the lenses, much the same as in the circular strokes employed in grinding lenses by hand. The radius of the crank admits of adjustment to give various degrees of eccentricity to the upper tool, and the pressure is regulated either by a spring, or by adjusting the weight of the grinder.

* Described in a communication to the Society of Arts. See Trans., Vol. XLIX., page 91.

As previously mentioned, the best lenses for object glasses of telescopes are ground and polished singly by hand; in this case the lens whether concave or convex is kept in the hand, and the grinding tool is fixed. The glass if small is held by a cement handle, and if large is cemented to a metal handle, as wood is liable to swell with the moisture.

The grinding is performed in exactly the same manner as when several lenses are ground together, but greater care is taken with every successive step, and these lenses are in general polished upon a piece of thick silk, the kind known as lutestring being preferred.

The silk cut to the width of about seven eighths the diameter of the lens, is stretched across the middle of the brass tool, and the lens is rubbed backwards and forwards in straight lines along the silk, and instead of the operator walking around the post, ' the lens is continually twisted round in the hand, and at the same time traversed gradually sideways until the center of the lens is brought to the edge of the silk, when the direction of the traverse is reversed. The single thickness of silk stretched across the tool assumes the form more correctly than the cloth polisher, and the lens is traversed partly off the silk in order that the center may be acted upon equally with the margin.

The putty powder and water with which the silk polisher is supplied, are kept ready mixed in a corked bottle to avoid the contamination of dust, and at the time of application the bottle is shaken up, and its contents allowed to subside for a few seconds, a small quantity of the water is then taken put with a clean stick and thrown upon the polisher, and thus only the suspended portions of the putty powder are used. The most carefully finished lenses are polished on a pitch tool prepared in the same manner as for polishing specula.

It has been stated at page 1265, that with ordinary lenses accuracy of spherical form is of much greater importance than the radius of curvature, but in making the object glasses of achromatic telescopes it is requisite to be enabled to measure accurately the radii of curvature of the lenses, which are first tried experimentally, and afterwards made as nearly as possible to the radii obtained by calculation, in order to correct the chromatic and spherical aberration.

In 1841 Mr. Andrew Ross, (from whose practice most of the foregoing particulars on grinding and polishing lenses have been derived,) received the silver medal from the Society of Arts for his instrument, called a spherometer, for measuring the curvature of the grinding tools. The instrument is shown in fig. 1130, and the following description by Mr. Ross is extracted from Vol. 53 of the Transactions of the Society: -