638 report — 1884. 



•with so much difficulty, that it has occurred to me that, although my own experience 

 is far inferior to that of my late father, a few ohservations may be of interest. 



The process for preparing and completing a paraholic speculum, after tliat the 

 casting, rough grinding, &c, have been completed, consists of two operations : first, 

 the ' grinding ' with emery of finer quality and smaller quantity as the operation 

 proceeds, by which so tine and true a spherical surface may be given as may quickly 

 receive a polish under the action of the polisher; and, secondly, the ' polishing' proper, 

 which not only imparts the polish but changes, by the action of the elastic surface 

 of pitch covered with a paste of rouge and water, the spherical into the paraboloid 

 surface The principal causes of uncertainty in the process are the variable hygro- 

 metric condition and temperature of the air. If the air be too moist, that gradual 

 drying-up of the moisture of the polishing material which is essential to success 

 will not take place ; if, on the other hand, it be too dry, it will be difficult to 

 preserve that uniformity of moisture which is equally desirable. Again, if the 

 temperature be too high for the quality of pitch employed, the surface will be 

 untrue in detail from undue wearing down of the softer parts, and the pitch will be 

 pressed out too thin before the process is complete ; while, if the pitch be too hard, 

 the rouge will not become sufficiently embedded in it, and the polish will be 

 imperfect, and also the pitch will not be able to adapt itself to the speculum, and 

 an imperfection of figure will result. In short, we have three variables — tempe- 

 rature, moisture, and quality of pitch, and these must be so relatively adjusted as 

 to give a satisfactory result. Two kind3 of stroke were employed by my father, 

 that known as the straight stroke, produced by the action of two eccentrics, one 

 making many (say 15) revolutions to one of the other, and that known as the 

 circular stroke, employed also by Mr. Lassell, where both eccentrics rotated in 

 equal times, their radii being parallel to one another. 



The straight stroke involved a considerable overhang of the polisher during two 

 or three consecutive strokes, and a consequent difficulty in preventing too great local 

 compression of the pitch, while there was some difficulty in preventing an unequal 

 evaporation of the moisture when the ah was a little too dry. 



The circular stroke also involved a tendency to unequal drying, and a tendency 

 to annular elevations, and depressions of the surface of the speculum could only be 

 removed by a change at intervals of the distance from the centre of the circle de- 

 scribed by the centre of the polisher from the centre of the speculum. This was 

 effected in the case of the 3-foot speculum by a somewhat complicated contrivance, 

 and in the case of the 6-foot speculum simply by hand. 



The late Mr. Lassell, in the 'Phil. Trans, for 1875,' describes a machine 

 in which the relative angular motions of the eccentrics differ a little only, but as 

 at the time he had nearly given up observing owing to age and infirmity, the 

 machine probably was not tested as it deserved. 



I have, during the last two years, reconstructed the two polishing machines at 

 Parsonstown, and have adapted them for giving to the polisher the stroke of Mr. 

 LasseU's machine. 1 I have polished two 3-foot specula and one 6-foot speculum. 

 In each case the result has been much above the average. There has been no ap- 

 preciable unequal dicing, and no tendency to annular depressions. 



It should be observed that in all the machines the speculum has a slow move- 

 ment of rotation on its axis, and also the polisher turns round, at a different rate. 

 This was effected by Lassell by means of a ratchet. We have found no necessity 

 for a special contrivance, as the polisher being grasped by a loose hoop, with 

 perhaps i to 1 inch of play, rotates by rolling inside the hoop. 



With regard to the polishing of flat mirrors, as the required figure can be 

 imparted in the grinding, it appears to be much the simplest to take pains to 

 obtain as flat a surface as possible before polishing, and to impart as fine 2 a surface 

 also as we can in the grinding. Care in this is well repaid by the shortness of time 



1 On reference to Mr. LasseU's paper, it appears that the relative angular velocities 

 of his eccentrics are as 1 : 0-922. In my machine the relative velocities are as 

 1 : 0806, but by means of change wheels any other relative velocities can be given. 



2 A bed of hones, made by cementing pieces of Welsh hones on to a metal plate, 

 is well adapted for giving a line surface before polishing. 



