168 CONSTRUCTION OF LARGE TELESCOPE LENSES. 



lens about upon the spindle of the lathe until exact coincidence is 

 reached between the optic axis (common axis of curvature of the two 

 surfaces) and the mechanical axis of the spindle. Recognition of 

 this condition depends on observing the reflections from the two 

 glass surfaces, and accurate centering is reached when these reflec- 

 tions do not move with the rotation of the spindle. When the right 

 adjustment is made the edge of the disk is turned off true by means of 

 a grinding band fed Avith emery and water, and bj^ this means the 

 lens is reduced to the proper diameter. 



After the centering follows the fine polishing on a special polishing 

 machine. The process is much the same as that of rough polishing, 

 excepting that instead of a cloth-covered tool there is provided for 

 each face a series of great pitch-covered plates. Frequent trials of 

 the surfaces are made by means of so-called " test glasses.'' These are 

 small glass plates ground and polished accurately to fit the desired 

 curve; that is to say, convex for a concave surface, and vice versa. 

 Their employment in testing depends on the following principle: 

 If two closely fitting polished surfaces are laid one upon the other 

 there is retained between them a thin film of air which exhibits the 

 so-called " Newton's colors,'- seen in soap bubbles and similar thin 

 transparent structures. The color is the same over th(> whole surface 

 only when the thickness of the inclosed film is everywhere uniform, 

 which only occurs when the lens has the same curvature as the test 

 glass. At the beginning of the polishing the Newton's colors appear 

 as rings of more or less width. By the proper use of i)()lishing tools 

 of different sizes, and by suitable regulation of the stroke and velocity 

 of the machine, the condition is finally reached wIumi a uniform 

 color supersedes the rings, no matter where the test glass is laid upcm 

 the lens. By such methods of measurement in terms of the wave length 

 of light, deviations of thickness of only one ten-thousandth millimeter 

 (one two-hundred-and-fifty-thousandth inch) can be accurately de- 

 tected, a magnitude scarcely appreciable to the lay mind. It is 

 obvious that the fine i)olishing in such conditions is an exceptionally 

 difficult task, the more so that care must always be exercised to avoid 

 all blemishes on the surfaces, such as scratches and the like, and only 

 the most competent and experienced workers can succeed w^ith it. 

 A concei)tion of the difficulties to be overcome may be found when it 

 is said that the fine polisliing of a single lens surface takes several 

 months. 



^^^len l)oth lenses have passed through the processes of fine grnid- 

 ing and polishing they are inserted in brass or iron mountings which 

 have meanwhile been prepared for tlieni and in which they lie sepa- 

 rated by a small free space. Cementing together with Canada balsam 

 or turpentine, as generally practiced with small lenses, and formerly 

 with hirge ones also, has more recently l)een discontinued on account 



