328 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13, No. 14 
and the furnace used had a zone of 4 cm. in which the temperature was 
uniform to one degree. 
In the quenching experiments the charge was placed within 1 mm 
of the bare thermocouple junction, and the whole carefully centered 
in the furnace. The furnace was kept at a constant temperature 
by a Roberts type thermoregulator,’ and readings were made every 
minute. The two curves 2349 E and 2349 G in Fig. 1 give minute 
by minute readings for two of these experiments for 30 minutes; the 
maximum e.m.f. on 2349 G is 10642, the minimum 10640, with the 
exception of drops to 10639.0 and 10639.5 at the end of the first and 
fifth minutes, respectively. 
The results of the quenching experiments are given in Table 1. 
TABLE 1—ReEsvtts oF QUENCHING EXPERIMENTS 
EXPERIMENT NO. TEMP. IN MICROVOLTS CONDITION 
2449C 10632 Crystalline 
2449G 10641 “ 
2449F 10650 Glass 
2449E 10645 ae 
2449D 10643 Both crystals and glass 
The melting point of this sodium metasilicate can with certainty be 
placed at 10644 microvolts with this particular element; the actual 
temperatures do not concern us here. 1 microvolt is equivalent at 
this temperature to about 0.09°C. 
Two heating curves were run before the quenches, 2349 A and 
2349 B, and two after, 2350 A and 2350 B. Readings were in some 
cases taken every 30 seconds, in others, every minute. The experi- 
mental results are shown by the curves of Fig. 1, on which the deter- 
mined points are represented by crosses; the ordinate for the melting- 
point as determined by quenches, 10644, has also been drawn. ‘The 
curves drawn thru the observed points are of course wholly empirical 
hence the exact location of the point of inflection is impossible. 
Examination of the data shows that in each case the ordinate for 
10644 cuts the curve in the region of steeply ascending slope. In 
2349 A, the increase of e.m.f. for the time interval which includes the 
ordinate 10644 is 44 microvolts, while in the preceding interval the 
e.m.f. increased 23 microvolts; similarly, in 2349B, the increments are 
38 and 11; in 2350A, 17 and 9; in 2350B, 10 and 8. Theprecise location 
3H. S. Roberts. This Journal 11: 401. 1921. 
