Minerals of the Composition Mg Si O z . 399 



an unglazed porcelain tube of about l cm inside diameter, the 

 lower end of which was directly over the uncovered platinum 

 crucible. The upper end of this tube was covered by a porce- 

 lain lid. (An arrangement similar to tig. 10.) The temperature 

 of the furnace having been raised to the desired temperature, 

 the glass, in small fragments, is then dropped through the porce- 

 lain tube into the crucible. Only small fragments can be 

 used, for the initial temperature of the crucible is materially 

 lowered by the introduction of any considerable mass of the 

 cold substance, after which the heat of crystallization of a 

 large mass of the glass will raise the temperature of the 

 portion last to crystallize above that of the furnace. At 

 1150°-1200°, when small particles, each of a few milligrams 

 weight, were used, no perceptible change in the temperature 

 of the crucible was noted. Under such conditions about 20 

 seconds elapsed, and we estimated that the temperature of the 

 crucible was reached, before crystallization began. The tem- 

 perature of the glass particles, however, does no doubt rise 

 above the temperature of the crucible, for they become more 

 luminous during the process. It is interesting to watch the 

 crystallization ; it begins at the surface and proceeds inward 

 with rise in temperature, Avhich is especially noticeable when 

 the particles are large ; the outer portion contracts and squeezes 

 out the still liquid interior in the form of hollow, rounded 

 projections, which crystallize last, and at a higher temperature, 

 for they are always of coarser grain. These grains bear a strong 

 resemblance to a kernel of ordinary popped corn. 



Table IY contains the results of the experiments on the crys- 

 tallization of magnesium silicate glass at different tempera- 

 tures by the process above described. 



Table IV. 

 Crystallization of Magnesium Silicate Glass. 

 750° No crystallization in 45 minutes. 

 900 Too fine-grained for identification. 

 990 " " " " " 



1075 Nearly all enstatite ; presence of monoclinic form 



doubtful. 

 1125 Enstatite, with some monoclinic form. 

 1175 Enstatite and monoclinic form. 



u a a u 



a (c a tt 



1300 Monoclinic form and probably some enstatite. 

 2325 " " u " " " 



It is plain from these figures that below 1100° the product is 

 enstatite ; while not far above that point the monoclinic form 

 begins to appear. To prepare a nearly homogeneous aggre- 



1225 

 1275 



