REVIEWS 30 1 



jected was estimated to vary from 1600 down to 500 C. Two open- 

 ings, half a foot long each, were arranged in the side of the furnace so 

 that crucibles could easily be inserted and removed. The temperature 

 within the entrance chamber was much less than in the heart of the 

 furnace, and by placing a crucible first in the innermost glow, then at 

 the inner mouth of the chamber, and lastly, a short distance within the 

 chamber, conditions of gradual cooling and crystallization could be 

 brought about. From day to day at certain periods there were varia- 

 tions in the temperature of the furnace itself due to the requirements 

 of glass manufacture which went on as usual in the huge crucibles of 

 the factory, and these changes affected to a certain extent the crystal 

 structures obtained. Fire-clay crucibles of various sizes were used, the 

 melting being done in large crucibles, the crystallization in smaller 

 ones of 150 c. c. capacity. The crucibles when filled were carefully 

 covered and placed on refractory tiles. They were first warmed to 

 dark red heat and then thrust into the position of maximum tempera- 

 ture. After a few hours they were drawn to the second position at the 

 inner mouth of the opening, and finally, after remaining there for sev- 

 eral days were drawn within the small chamber where they finally 

 cooled. 



Crystallization lasted commonly from one to three weeks, but in 

 exceptional cases the crucibles were left in the furnace as long as two 

 and one half months. A few experiments were made on a very large 

 scale in the great factory crucibles where over a hundred pounds of 

 mineral matter was molten at a time. It was found that certain mix- 

 tures corroded the crucible violently, while others remained unaffected 

 by contact with the crucible walls. Magmas with high magnesia and 

 low alumina and alkalies acted violently upon the clay, because mag- 

 nesia has, at these high temperatures, a very strong affinity for alumina, 

 and in the absence of alumina from the mixture combines readily with 

 that which forms the containing vessel. Mixtures of lime and the alka- 

 lies, rich in alumina, do not affect the crucible, even after long exposure 

 to the highest temperatures. About two hundred experiments were 

 made in all, and of these 25 per cent, failed owing to various causes. 

 The others produced coarsely crystalline mineral masses in many cases, 

 so that isolation of the minerals for analysis could be accomplished. 

 The mixtures used were prepared usually from pure chemicals Silica 

 was used in the form of the hydrate Si0 2 .3H 2 0; alumina as hydrar- 

 gillite (A1 2 3 .3H 2 0) ; lime, magnesia, and the alkalies as carbon- 



