INDIVIDUALITY OF MINEKALS IN MELTING 147 



one instance may be cited. More than one observer has sought to deter- 

 mine melting temperatures by watching for the moment when the sub- 

 stance appears fluid and runs in the crucible, and measuring the tempera- 

 ture at which this appearance of fluidity occurs. Moreover, this method 

 has served successfully for the measurement of melting temperatures of 

 low-melting salts and of metals, and will serve again in the case of single 

 minerals^ whenever it can be shown that this appearance of fluidity occurs 

 coincidently with melting; but it so happens that in minerals this does 

 not always hold true. A number of minerals, like quartz and the feld- 

 spars, melt to form liquids which are so viscous^ that they show little or 

 no tendency to take the form of the containing vessel, even after melting 

 is complete, and therefore give no evidence, to the observer who merely 

 watches for the mineral to run, of the beginning of melting. In the 

 accompanying photograph (plate 3) a crystal of orthoclase was bent with 

 a wire when partially melted. The amount of bending is shown by the 

 cleavage crack. Dark areas are already molten; bright areas unmelted 

 crystals. It is obvious that the melted (dark) material is not squeezed 

 out or otherwise displaced by the bending more than the unmelted orig- 

 inal orthoclase. 



To the research student who has become experienced in such phenom- 

 ena this plainly indicates that the particular property chosen, namely, 

 the appearance of mechanical fluidity after sufficient heat has been ap- 

 plied, is not a suitable one with which to determine the melting points 

 of substances which form viscous liquids, and that his ingenuity must 

 devise a more appropriate means with which to approach these sub- 

 stances. In other words, experience teaches him that he must have more 

 than one method at his disposal for the determination of melting points 

 of minerals if he would obtain a competent record, and that he must 

 make a study of each mineral from this viewpoint in order to assure 

 himself that the method which he uses will locate the desired point with- 

 out ambiguity in each. individual case. In our experience, some minerals 

 have been found to melt sharply to a thin liquid, and with these almost 

 any method is competent to determine the melting temperature. Others 

 can be more readily determined by noting the disappearance of crystal 

 structure (indicated by abrupt changes in some optic property like bire- 

 fringence) when a fragment of the mineral is observed in a furnace 

 mounted on the table of a microscope. With still others the melting 



^ It will not serve for the study of a mixture of minerals, and Is therefore of little or 

 no value in approaching the subject of rock formation. 



"^ Hyperviscous fluids differ essentially from solids. A fluid, unless extremely under- 

 cooled (glasses), yields continuously, though sometimes very slowly, even to the smallest 

 pressures. A solid does not. 



