XIV, 5 Dar Juan and Elicano: Philippme Glass Materials 471 
ing cullet with the batch is to utilize the fragments of glass 
left over during the process of manufacture, and to serve as a 
flux and to lower the temperature of reaction of the materials 
employed. Mixing the batch with cullet also considerably re- 
duces the cost of production. However, since the main object 
of this investigation was to ascertain the suitability of Philippine 
materials for glass making, no cullet was used in our experiments. 
PROPORTIONS OF CONSTITUENTS 
The proportions in which the different ingredients of glass 
are mixed vary considerably, depending upon the quality and 
composition of the raw materials available, the quantity of broken 
glass mixed with the batch, the temperature, and the quality and 
color which it is desired to give the finished product. 
Table IV gives the proportions of the materials used for the 
different batches, the percentage composition of glass materials 
in the mixture, and the percentage composition of the glass 
obtained. 
The mixtures used in the first and second batches were rather 
difficult to melt and gave a glass that at white heat was not 
sufficiently fluid to expel all the gas bubbles. If to mixture 1 
a small quantity of arsenious or manganese oxide is added, as 
shown in the table, a practically colorless glass is obtained. 
Tubing and bottles made from these mixtures are hard and 
strong, but they become cloudy after a few weeks. Mixture 3 
is not difficult to melt, the gas bubbles are easily expelled, and 
the glass remains perfectly transparent. Mixture 4 gives a glass 
practically free from bubbles, and bottles made from it have an 
excellent appearance. Also in this case the addition to the mix- 
ture of a small amount of arsenious or manganese oxide decol- 
orizes the product almost completely. Mixture 5 gives a greenish 
glass excellent for soda-water or beer bottles. The addition of 
small quantities of arsenious oxide (batch 6), or of manganese 
oxide (batch 7), produces a lighter-colored glass. Mixture 8 
gives a green glass suitable for soda-water or beer bottles, and 
mixtures 9 and 10, a glass rather too dark for soda-water, but 
good for beer and wine bottles. 
The furnace used in these experiments was built after the 
ordinary Japanese glass pot furnace and has over-all dimen- 
sions of 188 by 79 centimeters at the base and a height of 132 
centimeters. Text figs. 2 and 3 give a general idea of the fur- 
nace construction. Its most important feature is the air pas- 
sage, represented in fig. 2 by the letters la, 2a, and 3a, which 
