MELTING POINTS AND MELTING INTERVALS 



159 



substances) superheat the whole charge. Small and narrow crucibles 

 (figure 4) liave the effect of diminishing this difficulty also. This whole 

 question of apparent anomalies in the determination of melting points 

 due to small quantities of impurity and irregular heat distribution has 

 been treated in considerable detail, both theoretically and practically, by 

 W. P. White in two recently published papers. ^^ 



The third difficulty with the sharpness of melting points 

 is inherent in the participating materials themselves, and 

 the temperature-measuring device is merely offering a 

 faithful interpretation of what actually occurs. 



Earlier in the paper it was noted that a liquid mineral 

 on cooling often does not crystallize promptly, but through 

 the iniluence of other properties; chief among which is the 

 viscosit}^, the crystallizing temperature is lowered by an 

 uncertain amount. This w^as explained as a kind of molec- 

 ular inertia, inherent in certain su.bstances, through the 

 operation of which any molecular rearrangement, such as 

 crystallization from a molten mass, is seriously hindered. 

 Xow it sometimes happens that what is true of crystal for- 

 mation in a cooling mass may also be true of crystal de- 

 struction in a melting mass — the crvstalline structure can 



^ • FiGVKE 4. — Small 



only be broken down with considerable difficulty and delay. Crucihiefor the ac- 

 Tn such a case the expenditure of the energv required to ^"'"^^^ nefermma- 



^ fc). ^ ^.o„ f^f fi,p Melting 



break down the molecular system (latent heat) will in fact Point of Minerals 

 le distributed over a temperature interval, and will be §,) ("^"''^^''■^"^'**^^•• 

 recorded by the thermometer. Just as in the case of the delayed solidifi- 

 cation, where the solidifying temperature appeared Ioav and variable with 

 the rate of cooling, so here the delayed melting may result in too high a 

 temperature, which is equally dependent upon the rate of heating. In 

 extreme cases, of which quartz and albite again serve as familiar exam- 

 ples, this effect is so pronounced that no melting "point," in the proper 

 sense of the word, can be obtained — that is, the change of state itself 

 actually extends over a greater or less temperature interval, and the accu- 

 rate interpretation of such a melting curve is a matter of some uncer- 

 tainty. A series of actual curves showing the long temperature intervnl 

 (100 to 200.. degrees) over which the heat absorption of melting ortlio- 

 clase is distributed, with different rates of heating, is shown in figure 5. 

 The dotted line indicates the interval during which heat is absorbed. In 



K^ 



15 w. P. White: Melting-point determination. American Tournal of Science (4), 28. 

 1909, pp. 453-473. Melting-point methods at high temperatures. American .Journal of 

 Science (4), 28, 1909, pp. 474-489. 



