406 E. T. Allen, F. E. Wright and J. K. Clement— 



In addition to this substance small amounts of both the 

 orthosilicate and quartz were observed. The crystals of the 

 former were frequently well developed and exhibited the char- 

 acteristic crystal habit of fosterite ; they extinguished parallel 

 to the prism edge, had an average refractive index of about 

 1*66, and much stronger birefringence than either the m eta- 

 silicate or quartz. 



The quartz crystals were ordinarily very small and recog- 

 nizable only by their refractive index. In one preparation, 

 however, where the sodium silicate in the form of large lumps 

 was heated with magnesium ammonium chloride for 3 days at 

 400°-450°, numerous quartz crystals were found, the longest 

 of which measured nearly 3 mm . They were water-clear, doubly 

 terminated, sharply defined crystallographically, and resembled 

 crystals of the natural mineral. Both optical and crystallo- 

 graphic measurements were made on these crystals, the results 

 of which have been described at length in another paper* on 

 the minerals of the lime-silica series. 



In the products much of the material was too fine even for 

 approximate determinations and may have contained other 

 substances than the three cited above. It is noteworthy, how- 

 ever, that the pyroxene form of the metasilicate was not pro- 

 duced, but apparently the low refracting amphibole. 



IY. — Orthorhombic Amphibole. Eupjferite.j- — If magnesium 

 silicate be melted and cooled without any special precautions, 

 there may usually be detected, under the microscope, in the 

 crystallized product, patches of a substance which differs 

 markedly in appearance from either of the forms I and II. 

 The fact that it occurred more frequently in the outer zones of 

 the mass near the walls of the crucible, led to the suspicion that 

 rapid cooling favored its formation. Experience soon proved 

 the correctness of this conjecture. After a very considerable 

 amount of work on the conditions necessary for its formation, 

 we find (1) that the mass must be cooled rapidly, but not so 

 rapidly as to form glass ; and that (2) for this reason, probably, 

 the quantity of material must not be too large. A satisfactory 

 rate of cooling is attained if a platinum crucible containing 

 15 or 20 grms. of the molten silicate is quickly drawn from the 

 furnace and held, uncovered, by the operator until crystalliza- 

 tion is complete. If the hot crucible be set at once on a tile, 



* Day, Shepherd and Wright, The Lime-Silica Series of Minerals, this 

 Journal, xxii, Oct., 1906. 



f The name Kupfferite was applied originally to the natural monociinic 

 amphibole of this composition. In the course of time, however, the name 

 has been shifted to designate the orthorhombic form of the same composi- 

 tion and like enstatite represents the pure magnesium member of the 

 amphibole series kupfferite-anthophyllite. Hintze, however, objects to 

 this change. Compare, Hintze, Handbuch d. Mineralogie, Bd. II, p. 1196, 

 and E. S. Dana, System of Mineralogy, p. 384-385, 1900. 



