Proceedings of the British Association. 119 



which would be required for making any effectual progress in the re- 

 examination of the nebulae, the first experiments were undertaken, in 

 the hope of obviating the difficulties which had previously prevented 

 the application of the brilliant alloy which may be formed of tin and 

 copper in proper proportions to the construction of large instruments. 

 The manner in which the difficulty had been met, was, by adding an 

 excessive proportion of copper to the alloy, but the mirror was no 

 longer susceptible of a durable polish, and, when used, its powers 

 declined rapidly. It appeared to me, therefore, to be an object so 

 important to obtain a reflecting surface which would reflect the 

 greatest quantity of light, and retain that property little diminished 

 for a length of time, that numerous experiments were undertaken 

 and perseveringly carried on. After a number of failures the dif- 

 ficulties appeared to be so great that I constructed three specula, 

 where the basis of the mirror was an alloy of zinc and copper in the 

 proportion of 1 zinc to 2' 74 copper, which expands with changes of 

 temperature in the same proportion as speculum metal. This was 

 subsequently plated with speculum metal, in pieces of such size as 

 we were enabled to cast sound. These specula were very light and 

 stiff, and their performance upon the whole satisfactory; but they 

 were affected by diffraction at the joinings of the plates ; and although 

 very brilliant and durable, defining all objects well under high 

 powers, except very large stars, still as the effect of diffraction was 

 then perceptible they could not be considered as perfect instruments. 

 In the course of the experiments carried on while these three specula 

 were in progress, it was ascertained that the difficulty of casting large 

 discs of brilliant speculum metal arose from the unequal contraction 

 of the material, which in the first instance, produced imperfections in 

 the castings, and often, subsequently, their total destruction ; and it 

 appeared evident, that, if the fluid mass could be cooled throughout 

 with perfect regularity, so that at every instant every portion should 

 be of the same temperature, there would be no unequal contraction in 

 the progress towards solidification, nor subsequently, in the transi- 

 tion from a red heat to the temperature of the atmosphere. Although 

 it was obvious that the process could not be managed so that the 

 exact condition required should be fulfilled, still, by abstracting heat 

 uniformly from one surface (the lower one), the temperature of the 



