(a) A soda-alumina-lime-magnesia glass was continuously melted for sixteen days. The proportions of batch and cullet were equal in weight.
(b) A soda-lime glass was continuously melted for twentytwo days. In this case no cullet was used.
(c) A soda-lime glass was repeatedly melted for thirtytwo times. At the first melting no cullet was used. From the second melting, the water cullet of the previous melting was dried and used.
In each melting, the amount of the total batch was 15Kg at the begining, and the capacity of the clay pot used was 20kin. The specimen of glass which was taken out of each melting was chemically analysed, and its coefficient of thermal expansion and softening temperature were also determined.
From the result of chemical analysis it was found that the increase of the amount of Al₂O₃+Fe₂O₃ was little at the begining, but after a certain days or times of melting, it increased abruptly. When we calculate the result of analysis assuming that this increase of the amount of Al₂O₃+Fe₂O₃ to be due to the pot corrosion and that the ratio of the quantity of SiO₂ and Al₂O₃ in the body material of the pot to be 1.4:1, only little variations in the composition of the glass seem to have been caused by the continuous or the repeated melting.
The coefficients of expansion which were determined by measurements, and those values obtained by calculation applying English and Turners' factor to the result of chemical analysis, were compared, and it was found that while the days or the times of melting were few they agreed with each other fairly well, but after a certain days or times of melting which correspond to the abrupt increase of the amount of Al₂O₃+Fe₂O₃, remarkable differences were observed between them. This disagreement continued a little while and then vanished as the days or the times of melting increased still more. In this period of remarkable disagreement, the glass is thought to be inhomogeneous, and in the later period of agreement the glass is thought to become homogeneous again as the total amount of the glass became smaller.
The X-ray diffraction patterns of several glasses in the series (a) were photographed. There was no trace of crystallization, but two indistinct rings, which were found by Parmellee in feldspar glass, were found.
It was concluded that as the glass becomes ununiform owing to pot corrosion, its properties can not be determined exactly just as in the case of sillimanite pot which was studied by A. A. Childs, V. Dimbleby, H. W. Howes and W. E. E. Turner.

View full abstract

In continuing the previous study, the author reports in present paper on the influence of 0.5% fluorspar in various raw mixture of H. M. below 2.10. Results are: (1) In general, the consumption of the heavy oil decreased 10 per cent and out-put of the clinker increased 10 per cent. (2) The setting time of the cement was retarded as reported in parts 1 & 2. (3) The strength of the cement was not especially increased by adding 0.5% fluorspar.

View full abstract