REPORT OF COMMITTEE ON OBSERVATORIES 1 69 



found very injurious to silver on glass, but which did not perceptibly 

 affect the brilliancy of the surface of the grating. Doubtless you 

 will be able to learn the exact composition of the alloy used for the 

 Rowland gratings ; but probably the combination of 4 atoms of 

 copper with i atom of tin, recommended by the late I,ord Rosse in 

 his account of the construction of the great telescope at Birr Castle, 

 would yield an alloy capable of retaining its polish for many years. 

 From Lord Rosse' s own account it seems that he himself used a 

 somewhat softer alloy, with the consequence that the larger mirrors 

 made by him required repolishing about once every two years ; but 

 I may mention that the night air at Birr Castle Observatory is 

 usually very damp, and that owing to the great mass of the mirrors 

 they are very liable to get dewed. This brings me to another part 

 of the subject. 



A serious objection to metallic specula, as usually constructed, is 

 their great weight ; but this difficulty may be largely remedied by 

 giving the metal a more suitable form than that of a simple circular 

 slab or disc. By the use of suitable sand (such as is used in the produc- 

 tion of the highest class of bronze castings) there should be no diffi- 

 culty in casting a speculum with deep ribs on the back, which would 

 be much lighter and relatively stiffer than a disc of the same diam- 

 eter. For a mirror 54 inches in diameter, I would suggest making 

 the ribs and face of a uniform thickness of ^ inch. By giving the 

 speculum a total depth of 8 inches, it would probably be stiffer than 

 any mirror yet cast, and with a suitable arrangement of the ribs 

 would weigh about 1,200 pounds. Of course, I assume that the 

 ' ' metal ' ' would be cast face downwards on a " bed of hoops ' ' of the 

 proper curvature, as practiced by the late Lord Rosse, to insure that 

 the surface to be ground and polished should be perfectly sound and 

 almost exactly of the desired form. The possibility of casting a 

 speculum of this shape was, in a great measure, set at rest by an 

 experiment made by the present Lord Rosse, who cast an elliptical 

 flat mirror some 1 1 inches by 8 inches with a ribbed back, some 

 thirty years ago. This mirror was perfectly sound and took a very 

 high polish. It was used as a diagonal mirror for the 6-foot New- 

 tonian reflector. If I were trying the experiment I should be in- 

 clined to honej^comb the mirror after this fashion, adding a "web" 

 round the outside, but taking care to make every part, including 

 the face of the mirror, of the same thickness to facilitate uniform 

 contraction in cooling. It is almost needless to say that a casting 

 of this kind would need to be carefully annealed. Provided the 

 12 



