Fletcher — The Melting -Points of some of the Rarer Minerals. 449 



like oil, refusing to flow, depending upon prevailing conditions of surface- 

 tension. 



In few of the minerals examined was the melting-point definite ; that is 

 to say, nearly all showed longer or shorter periods of fusion before complete 

 liquefaction. The smallness of the particles ensures comparative freedom 

 from errors due to conductivity, and renders unnecessary considerations of 

 the latent heat of fusion. The delicacy of the method reveals the lack of 

 sharpness in the melting of complex minerals. 



It is important, as other observers have shown, to use only the finest powder, 

 as experience shows that comparatively large particles of moderate fusibility 

 may melt only on their undersides, and show no change under the microscope 

 at high temperatures. With substances revealing more sharply defined 

 optical changes, the temperatures may be obtained accurate to within one 

 or two degrees. The following results are probably correct to 20°C. above or 

 helow; approximations as close as was considered were required. 



Colour-change in glasses : — An interesting colour-change is to be observed in 

 the glasses from certain minerals, especially tantalites. Melted tantalite, 

 which is a glass, presents when cold varying shades of yellow according to 

 the thickness of the enamel. When heated, this darkens in intensity to an 

 extent depending on the specimen, and sometimes to opacity. When cooled 

 slowly, the melt retains the darker colour-shade. On heating, however, to a 

 certain temperature — usually a bright red — and suddenly cooling, the original 

 lighter yellow tint is obtained. This change may apparently take place 

 indefinitely in the same glass. 



Production of sublimates : — In many cases the appearance has been 

 described of a growing ring, often molten, collecting round particles at 

 temperatures near their melting-points. Its first appearance is heralded by 

 . a single speck, situated at some distance from the particle, which is rapidly 

 joined by others as the temperature is raised until a definite peripheral ring 

 is formed. 



This has also been observed with elements such as nickel and cobalt, the 

 former furnishing a ring at about 1460°O. — its melting-point — which grows 

 outwards and eventually disappears, and may be the deposition of a heavy 

 vapour or sublimate on the cooler parts of the ribbon away from the 

 particle. 



Reactions witli fluxes: — With the meldometer substances may be con- 

 veniently examined in presence of fluxes, and as only minute quantities need 

 be used, colour tints are observable with smaller amounts than are necessary 

 for an ordinary borax bead. Delicate tints are the more easily detected since 

 ravs reacliing the observer have passed twice through the enamel by reflection 



