326 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13, No. 14 
PHYSICS—A comparison of the heating-curve and quenching methods 
of melting-point determinations... GrorGE W. Morny. Geophys- 
ical Laboratory, Carnegie Institution of Washington. 
The temperature of equilibrium between liquid and solid phases 
of a pure substance, or of the beginning or end of melting in a mix- 
ture, is a datum point frequently determined, and any information 
bearing on its determination is of interest to the investigator. The 
following experiments were carried out to compare the heating-curve 
method, which is the method followed almost exclusively in the study 
of metals and of salts which crystallize readily, with the quenching 
method, used chiefly in the study of substances which are difficult to 
crystallize, such as the silicate minerals. In the latter method, a 
tiny charge, of a few milligrams, is wrapped in thin platinum foil, and 
held at a definite temperature until equilibrium is reached. The 
charge is then suddenly cooled, sometimes by dropping into mercury, 
sometimes by merely lifting out of the furnace, and examined under 
the microscope. If the heating has been above the melting-point, the 
examination will show only glass; if below the melting-point, only 
crystals; and repetition at the same temperature for varying times 
will show how long it is necessary to wait for equilibrium. This 
method is the choice of all familiar with it, whenever it can be applied, 
because of the unequivocal nature of the evidence it supplies; it is, 
however, only applicable to substances which are sluggish erystallizers, 
such as the silicates, borates, and phosphates, and is not applicable 
to most salts. 
In the case of the silicate minerals, cooling-curves are rare y of 
value, because of the great tendency toward undercooling, and the use 
of heating-curves is made difficult by the same sluggishness. In 
metals and in easily crystallizable salts such as sodium sulfate the 
heating-curve shows a sharp break both at the beginning and end of 
melting, and usually an actually flat portion between. In the case 
of the silicate minerals the heating-curve is rarely sharply broken, and 
often is oblique. ‘The causes of the obliquity have been fully dis- 
cussed by White;? but the exact location of the melting-point remains 
a matter of some difficulty. It therefore seemed desirable to compare 
the two methods on a substance which gives a fair heating-curve, and 
also enables the melting-point to be located exactly by means of 
quenches. Sodium metasilicate is such a substance, and accordingly 
! Received July 11, 1923. 
2 W. P. White, Am. Journ. Sci. 28: 453-73. 1909. 
