246 SCIENCE PROGRESS 



of dust ; the air entering the room was filtered ; the walls and 

 ceiling were coated with shellac ; canvas was suspended above 

 the grinding machine, and the floor was kept wet during 

 polishing. It was also necessary that a 6o-inch plane mirror 

 for use in testing the paraboloidal surface of the ioo-inch 

 mirror should be constructed. 



The problem of casting the disc was carefully considered 

 by the St. Gobain Company. One of the chief difficulties was 

 how to anneal thoroughly so large a slab of glass. If the disc 

 were allowed to cool too suddenly or unevenly, internal strains 

 would be set up which, even though they might not cause the 

 slab to fracture, would almost certainly produce irregular 

 changes of figure of the completed mirror with change of tem- 

 perature. It was necessary after casting the slab for it to be 

 raised to such a temperature that internal strains would be 

 relieved and then to be allowed to cool extremely slowly and 

 uniformly. 



After repeated trials first on smaller discs and then on large 

 ones, a disc was obtained in 1908, which the glass company 

 considered satisfactory and which was therefore forwarded to 

 Pasadena. On arrival there, it was found that there were 

 numerous flaws within the disc and it was immediately rejected. 

 The company expressed their willingness to bear the loss and 

 to construct another disc. In consultation with Prof. Ritchey 

 a new glass furnace, as well as a new annealing oven and acces- 

 sories, were constructed ; the new furnace and melting-pot were 

 capable of holding 20 tons of material, and improved methods of 

 annealing were introduced. Meanwhile, a perfect 60-inch disc 

 for the plane mirror for testing had been made. The further 

 attempts to produce a 100-inch disc were destined to failure. 

 In 1910a successful disc was cast, but it broke during annealing. 

 The same fate happened to a further disc in 1 9 1 1 . It was there- 

 fore decided in the autumn of 19 10 to commence work upon the 

 disc already received and temporarily laid aside. The glass 

 company agreed to this being done, payment for the disc being 

 conditional upon its proving suitable for use as an astronomical 

 mirror. Examination of the disc showed that whilst there were 

 several sheets of air-bubbles running through it, they did not ap- 

 proach the surface sufficiently closely to interfere with a perfect 

 paraboloidal figure. It appeared probable that the internal 

 strains would not be sufficiently harmful and the process of 



