A few additional directions for testing will probably be of service to the reader.

My mirror was supported on a large board, the lower edge of the mirror resting on a ledge screwed to the board and deep enough -to carry the screens as well. Both mirror and testing apparatus should be kept as firm as possible - a heavy tripod about 3 ft. high is convenient for the latter, and the observer can work sitting; while a stout easel forms an efficient support for the mirror. The mirror can also be suspended by a strap from a nail, and rest against the wall; but that plan does not give so much power of adjustment. The mirror may, with advantage, be at a somewhat lower level than the testing apparatus, so that the tube of the latter (which, of course, points to the center of the mirror) is inclined downwards.

While the best place in which to carry on testing is undoubtedly a long cellar or underground passage with a stone floor, excellent results can be got almost any where, provided sufficient care is taken: (1) That the temperature of the air is uniform as possible; this, of course, means no fire, and (2) that the floor is as free as possible from tremors. Dr. Common did the testing of his 5 ft. entirely out of doors and found the conditions excellent; and, in my own case, as I had no cellar or stone floor at my disposal, the testing had to be carried out on an upper floor of a "timber-framed " house. By using a combined screen, and keeping perfectly still, and making every one else do so, I succeeded in getting very concordant readings. The least movement of anyone else in the house made the image dance about in a surprising manner.

Having placed the mirror on some firm support, we arrange the testing apparatus so that the distance of the pinhole from the surface of the mirror is approximately equal to the radius of curvature - i. e., twice the focal length - and so that the tube is pointed as accurately as possible to the center of the mirror. The pinhole is removed and the lamp lighted, and all light from the lamp is screened off as far as possible, except that issuing from the 1/2-in. hole and falling on the mirror, the room being darkened, of course. The lamp ought to be so adjusted in height that the brightest part of the flame is level with the pinhole, or the center of the larger hole, and draughts must be carefully avoided, or the illumination of the mirror will flicker in a disagreeable way.

The mirror is adjusted until the reflection of the flame is thrown back to the tube - this will take some time at first, as the image will be faint; but the difficulty is soon overcome by the use of a large sheet of paper and patience. A cap or disc of white card to fit the end of the tube will prove useful, as the image can be thrown on this, and when the paper cap is removed the light will pass through the tube as required.

The brass plate with the pinhole is now placed in position, and the observer looks through the tube at the mirror. If the head is moved back a few inches the image of the pinhole should be seen through the tube - and the testing apparatus may be moved about until the image of the pinhole is accurately centered.

We may now make a preliminary examination of the image of the pinhole with an eyepiece or, if the eyepieces are not yet procured, with a simple pocket lens. The apparatus must be moved to or from the mirror until the image can be brought into good focus, when, if the curve is fairly regular, the image should be sharp and well-defined. If the grinding and polishing have been carried out as recommended in my previous letter, the figure will probably belong to class A, the oblate spheriod; but if the stroke in polishing is too long, or the pitch too soft, it may tend to B or C. Whatever the appearance inside and outside the focus may be, one thing is essential: the image, both in focus and out of focus, must be absolutely circular and symmetrical. To judge of this the lens must be held accurately at right angles to the axis of the tube or distortion will be introduced, and it is better in every way to use a low power eyepiece which can be screwed into the tube. If one is not at hand a lens may be mounted in a paper cap to fit the tube.

A Cheap Nine Inch Reflector VI Continued Testing o 5

If the image is not quite circular, the surface of the mirror is not a perfect " surface of revolution, " and there is nothing for it but regrinding or starting afresh. Unless the curve is the same, whichever way it is taken across the mirror, good definition is quite impossible.

It is well, however, not to be in a hurry to condemn the mirror at the first glance. While I was working at the refiguring of my first 9 in., I used for a short time, to support the mirror, the back of a marble washstand - the mirror was therefore resting against marble which reached about 6 in. up its back. I was dismayed to find on testing, immediately alter polishing, that the image was hopelessly distorted and unsymmetri-cal. This, however, disappeared in half an hour or so, and the testing was quite satisfactory. The cause of the temporary distortion of the mirror was apparently that the mirror having been polished in a warm room had not had time to take the temperature of the much colder room where the testing was being done, and when the mirror began to cool down, the temperature of part of it in contact with the marble would naturally fall more naturally from being in contact with the upper part.

After examining the mirror roughly with the eyepiece - and it may be mentioned that the eyepiece test is not absolutely to be relied on, as many eyepieces have a certain amount of spherical aberration of their own which tends to make the mirror appear nearer to class A than class C - we may proceed to the screen test. Remove the lenses of the eyepiece and screw its mount into the tube. Then, with the eye a few inches behind the end of the tube, move the head from side to side and note whether the image and diaphragm in the eyepiece move relatively to one another. If they do, move the testing apparatus to or from the mirror until there is no relative motion, and then move it to the left so that the image is just not extinguished by the right-hand edge of the diaphragm. If now the eye is brought close up to the tube and the mirror viewed, it will be seen as it were flooded with light. Depress the left side of the baseboard by a gentle pressure of the hand, and the screen (the right edge of the diaphragm) will be seen to cut off the light from the mirror, the appearance being that described in Chapter IV. Care should be taken, of course, to have the screen exactly at the focus.