70 OPTICAL PROJECTION 



poses. They are sold in tlie capsules at 10s. each. For 

 long experiments I have found them very acceptable, in 

 spite of their inferiority in illuminating power. 



Only very recently, in fact since the above paragraphs 

 (which have been modified in consequence) were actually in 

 type in a more pronounced form, my hopes have been revived. 

 Observing in a paper l by Mr. G. H. Bailey, D.Sc., of Owens 

 College, that his purest zirconias prepared for other purposes, 

 were distinguished by luminosity and a slight bluish tinge in 

 the incandescence, I communicated with him, and he was so 

 kind as to send me three separate samples, weighing 12, 15, 

 and 18 grains respectively. There were difficulties in the 

 manipulation, chiefly from shrinkage when first ignited, which 

 have prevented me (by premature breaking up) from as yet 

 ascertaining the full incandescence of these samples ; but I 

 found them all absolutely free from any trace of silica * glaze,' 

 and satisfied myself completely that the light was in truth 

 perfectly white (if anything slightly bluish) and far superior 

 in quantity to that of any previous zirconias I had obtained. 

 I also proved unmistakably that the incandescence of this 

 material was much superior when the small thin disc is only 

 supported by a thin ring of platinum, to that when the sample 

 was mounted in one of the above-named capsules emptied for 

 the occasion. Such discs, about ^-inch in diameter (more is 

 useless) will each require about 1 gramme or 15 grains of 

 zirconia, previously calcined in the heat of the OH flame 

 itself to complete the shrinkage. 



Provided therefore that samples of zirconia of approxi- 

 mately equal purity to these can be supplied at any price 

 commercially practicable, the material does hold out hopes. 

 The real difficulty is to entirely get rid of the silica, and these 

 specimens alone, of all I have had, showed no trace of glaze 

 after heating. As already stated, my own experiments are 

 not yet completed, and it remains still to be ascertained how 



* See Proc. Royal Society, xlvi. 74. 



