190 H. G. SMITH. 



somewhat opaque, but reverted again to the transparent 

 condition at the melting point of the original substance. 



(c) When the original crystals were boiled in water they 

 softened and apparently fused at that temperature, and, 

 when the solution had become saturated, remained as fused 

 globules or masses in the boiling water, but soon solidified 

 into a semi-crystalline condition when the water had 

 sufficiently cooled, showing that complete fusion had not 

 taken place, because when fused by dry heat the mass 

 always remained as a glass, and there was no sign of 

 crystallisation during the many weeks it remained under 

 observation. If, however, this glass was dissolved in 

 alcohol it again readily crystallised from this solvent, and 

 the crystals were also readily formed from water when the 

 glassy form was boiled directly in the usual way. When 

 thus recrystallised, the melting point of the crystals, both 

 from the alcohol and from the water, had reverted to that 

 of the original crystals, although the melting point of the 

 fused material from which they had been derived had only 

 been 62-63°. 



(d) If the melted glassy substance was broken up into 

 small spangles, but not powdered, these became, after 

 several weeks, opaque and yellowish in colour. The melt- 

 ing point of these opaque spangles had then considerably 

 increased, showing that the tendency is to revert to the 

 higher melting point in all cases, which may thus be con- 

 sidered the stable condition. How many weeks or months 

 it would take for the melted unbroken glassy lump to revert 

 to the higher melting point is not yet known, as sufficient 

 time had not elapsed. So far, this has been found to be 

 62 - 03°, and in one case three months had passed between 

 the fusion of the substance and the determination of the 

 melting point. The method of observing the melting point 

 of the spangles was to place them on a thin glass microscope 



