904 
agreement between the thus obtained solidifying-points and the true 
temperature of equilibrium. 
At 14000 M. V. (1371° C.) the whole mass was again a viscous 
liquid, while the cooling-velocity was 3°,5 C. per minute; after an 
undercooling to 1365° C., the liquid erystallised by shaking and 
tapping the crucible, and the temperature increased to 1376° C. 
Another time we observed undercooling to 1349° C. and a erystal- 
lisation at 1364° C.; in yet another experiment, with varied tempera- 
ture-gradient, undercooling occurred down to 1306° C., crystallisation 
setting in at 1326° C.; ete. Thus even in this extraordinarily favour- 
able case the temperature of crystallisation appears to be completely 
dependent upon the preliminary treatment of the molten mass, and 
upon the particular way, in which the heat is withdrawn: cf ds 
possible to find as solidifying-point any arbitrary temperature. 
The complete impracticability of the cooling-inethod is here also 
proved in an indisputable manner. 
§ 6. By the way it merits attention, that the temperatures of 
observation in this and the foregoing case, are close to the extreme 
limit, to which the platinum-resistance-farnaces can again be applied. 
The effect of electricity-leakage out of the heating-coil, which occurs 
by the transport of the ions in the white-hot air-column, appeared 
to be rather appreciable; and it was absolutely necessary now to 
eliminate these disturbances by means of a protecting shield of zero- 
potential round the thermo-elements, by conducting the electrical 
charges to the earth-surface. The conduction-wires of the FARADAY- 
cage were therefore lengthened, and soldered to the platinum-cru- 
cibles by means of the oxygen-flame. Only by these precautions it 
appeared possible to make the final measurements, without being 
troubled by leakage-phenomena any more. 
$ 7. We prepared by synthesis also the ortho-silicates of barium 
and strontium. However the meltingpoints appeared to be here so 
high, that the substances could no longer be melted in platinum- 
crucibles, while all attempts to determine some reliable melting- 
point by means of the optical pyrometer, using a carbon-shorteireuit- 
ing furnace, had to be given up because of the reaction between the 
carbon and the silicates. 
§ 8. Zinc-Metasilicate: ZnSiO,; Zine-Orthosilikate: Zn,SiO,, 
and Willemite. 
These compounds were all prepared from pure ZnO and SiO, 
