Apbil 27, 1906.] 



SCIENCE. 



671 



When pure crystallized silica, either quartz 

 or tridymite, is fused in graphite crucibles, 

 there is no difficulty in obtaining quartz glass. 

 The difficulty, as- indicated above, is to free it 

 from enclosed air. After a series of experi- 

 ments at different temperatures, it became 

 evident that there was no probability of ob- 

 taining clear glass by direct fusion at at- 

 mospheric pressure. It was, therefore, de- 

 cided to study the effect of pressure upon the 

 fusion of silica. 



The experiments were conducted in a large 

 bomb furnace under a pressure of 500 pounds 

 of compressed air, the heat being- supplied by 

 passing an alternating current through the 

 walls of a thin graphite box containing the 

 quartz. The heating was at first carried to a 

 much higher temperature than was necessary in 

 an effort to reduce the viscosity sufficiently to 

 release the air bubbles in the normal way, but 

 the attempt failed entirely — the viscosity is 

 but slightly diminished at the higher tempera- 

 tures, and enough silica is reduced to discolor 

 the mass with free silicon. The appearance 

 of the product also clearly showed that gas 

 was being generated at the hottest points in 

 the retaining walls, and the large bubbles 

 formed by the rapid expansion of this gas 

 were always lined with free silicon. At the 

 highest temperatures (above 2,500° C), there- 

 fore, we not only did not get rid of the air 

 bubbles enclosed in the glass, but introduced 

 a new disturbing factor. 



The next step was, of course, to reduce the 

 temperature and lengthen the time of heating. 

 This produced blocks of quartz glass which 

 were quite transparent but which contained a 

 great number of small included bubbles which 

 could not be displaced even when the time of 

 heating was extended over several hours. No 

 large bubbles appeared, however, and no dis- 

 coloration. An effort to explode the enclosed 

 bubbles by turning on the compressed air be- 

 fore the heat was applied and then releasing 

 the pressure while the material was still 

 molten, also failed. The inflated material 

 could not be brought back to a cake again. 



After a number of attempts of this char- 



actei', with slightly varied conditions of tem- 

 perature and pressure, a charge was heated 

 rapidly to a high temperature (considerably 

 above 2,000°) before pressure was applied. 

 After the quartz had begun to vaporize freely, 

 it seemed reasonable to expect that the vapor 

 would displace the air between the grains some- 

 what as mercury vapor is made to displace the 

 air in filling thermometers. Compressed air 

 was then quickly applied to compress the melt 

 into a compact mass, the temperature lowered 

 to the point where it had been found safe to 

 work without discoloration, and held there for 

 perhaps a half hour, after which the current 

 was turned off and the pressure very gradually 

 withdrawn. Plates of quartz glass 3 x 5 x ^ 

 inches were produced under these conditions 

 which were almost entirely free from bubbles, 

 and only occasionally slightly stained by free 

 silicon. The residual bubbles are very small, 

 not more than J mm. in diameter, and are 

 not frequent enough (not more than two or 

 three in a cubic centimeter) to interfere with 

 the use of the glass for lenses, mirrors or 

 other usual optical purposes. It is, further- 

 rqore, very probable that a little more skill in 

 handling, such as could readily be obtained 

 with longer experience, would get rid of even 

 the few remaining bubbles. 



Quartz glass is easily stained by very small 

 quantities of other oxides when present as 

 impurities. In particular, we found that as 

 little as .3 of one per cent, of other oxides was 

 sufficient to make the glass opaque and almost 

 black. It is, therefore, absolutely necessary to 

 start with very pure material, but it does not 

 require to be clear. Pure cloudy quartz serves 

 quite as well. 



The volatility may be due to one of two 

 causes: either the vapor pressure of liquid 

 quartz is very great, or the carbon reduces the 

 silica, with the formation of metallic silicon, 

 which at once volatilizes and is subsequently 

 reoxidized on passing into the surrounding 

 atmosphere. This reaction, that is, the re- 

 duction to the metallic state and subsequent 

 volatilization, is a very common one at these 



