THE MODERN REFLECTlNCi 'I'ELESCOPE. 



29 



determined by experiment; it should be sucli that the convexity of the mirror is 

 slowly and uniformly decreased. 



If the mirror, when first tested, is found to be concave, the cotivexing tool 

 shown in Fig. 6 is used to continue the polishing. 



The concaving and convexing tools often tend to introduce broad slight zonal 

 erroi's; hence recourse must be had repeatedly to the normal tool. When all trace 

 of general cv I- vat are has disappeared, any remaining zonal erroi's are eliminated by 

 the use of the normal tool, and, if necessary, of the small local or figuring tools, (see 

 page 24). 



Fig. 6. Convexing Polishing Tool for P'iguring Plane Mirror. 



If a finished plane miiTor is available which is not smaller than the one being 

 figured, the work is very greatly facilitated l)y continually cold-|iressing the polish- 

 ing tools on the finished miri-or ; every precaution must be taken, however, to pre-. 

 vent injury to the figure of tlie finished mirror by such cold-pi'essing. 



In some of the writer's early work, in which the thickness of mirrors was made 

 only one-twelfth of their diameter, it was found that a /torwa/ polishing tool, as de- 

 sci'ibed above, tended to change the mirror veiy gradually toward a concave. This 

 was undoubtedly due to the fact that the friction of i)oli8hiug warmed the surface 

 very slightly, thus expanding it and making it convex with I'eference to the polish- 

 ing tool ; the tool did not follow this change of form readily, hence the central [)arts 

 of the glass were acted upon in excess. Furthei-more, such thin mirrors, when un- 

 silvered, were so sensitive to slight changes of temperature that the presence of the 



