FLAMES OF ATOMIC HYDROGEN 143 



dication of reduction of the oxide to metal. A water-cooled, polished copper 

 surface held in the flame just over the heated oxide rapidly became coated 

 with oxide, but no metal was observed. It may well be, however, that the 

 apparent volatility of the oxides is in large part due to reduction of the 

 oxides and evaporation of the resultant metal, but as the temperature of 

 the flame gas is lowered by cooling, the metal vapor reacts with the water 

 vapor or oxygen derived from the original oxide. 



Quartz glass seems peculiarly difficult to fuse. At least, with the most 

 intense heat of the arc it retains a very considerable viscosity. For ex- 

 ample, when the atomic hydrogen flame is directed against the central part 

 of a thin piece of quartz glass about 3 mm. thick, it does not seem possible 

 to melt a hole through the sheet of glass. Right under the arc, however, the 

 silica evaporates rapidly so that after perhaps half a minute, but without 

 any further increase in temperature, the glass becomes so thin that a hole 

 forms. The hottest part remains clear and the surface clean, but this region 

 is surrounded by a ring where the surface is covered by a fairly thick film 

 of silicon or lower silicon oxides. On cooling the sample by removal from 

 the flame, this ring appears black or dark brown with a peculiar luster, but 

 in the flame the central part appears dark because of its low emissivity 

 while the surrounding ring is bright. Pyrometer readings of the sample 

 when strongly heated give black-body temperatures of 2530° K. for the 

 central part and 2690° K. for the inner edge of the ring. 



The inability to get the silica in a more fluid condition is due partly to 

 its inherently high viscosity. Cooling by rapid evaporation, and perhaps 

 failure of the quartz to catalyze the reaction involving the recombination of 

 the hydrogen atoms, probably limit the temperature to which the quartz can 

 be heated. 



EVAPORATION OF METALS 



Pieces of various metals weighing 3 to 6 grams were placed upon cupels 

 made from pure magnesium oxide, and were heated as strongly as possible 

 by an atomic hydrogen flame from a 60-ampere arc. Cracks which formed 

 in the cupels, through which molten metal could escape, were repaired by 

 fusing small amounts of lime into the surface. Of course, a flame directed 

 against the surface of a liquid cannot raise the temperature of the liquid 

 to the boiling point, for cooling by evaporation will hold the temperature 

 below this point. Thus, a Bunsen burner flame playing on the surface of 

 water in a small dish heats the water to only about 85° C. Table IV gives 

 maximum temperatures to which various metals could be heated by the 

 atomic hydrogen flame. 



These temperatures were calculated from the pyrometer readings by 



