398 K T. Allen, F. K Wright and J. K. Clement— 



by the aid of the microscope. The two forms resemble each 

 other so closely in habit and optical properties p that one must 

 rely chiefly on the extinction angle to distinguish between 

 them. In the prism zone enstatite shows parallel extinction 

 on all planes, while the monoclinic variety extinguishes paral- 

 lel only on the orthopinacoid. Inasmuch as crystals of the 

 latter frequently show pronounced development after the 

 orthopinacoid, their monoclinic nature can only be recognized 

 by immersing the crystals in a viscous liquid such as Canada 

 balsam and rolling them, log-like, and observing the extinction 

 in the different positions. 



We have succeeded in preparing enstatite which contains 

 only slight traces of impurities by a method similar to that 

 used in this laboratory for the preparation of pure wollaston- 

 ite,* and which, as may be noted, bears considerable resem- 

 blance to the method of Fouque and Levy. It consists 

 simply in crystallizing a glass of the same composition by 

 beating it to a temperature below 1100°. The practical 

 details of preparing this glass, however, involve considera- 

 ble difficulty owing to the readiness and rapidity with which 

 crystallization proceeds in this silicate. Not more than 10 

 grams of it should be used for a single charge, the tempera- 

 ture of the platinum containing vessel should be raised well 

 above the melting point of the silicate, and the fusion should 

 be instantly chilled by plunging it into cold water. Care 

 should also be taken that the temperature is sufficiently high 

 to melt the silicate (m. p. 1521°) and yet not high enough to 

 melt the platinum crucible (1720°). In our earlier experi- 

 ments, crucibles were ruined so frequently by partial melting 

 that we were compelled to make some rough temperature 

 measurements by inserting a thermoelement under the hood 

 of the Fletcher furnace. Though the temperatures thus 

 observed were not those of the crucible, they indicated the 

 latter approximately, and when the readings ranged from 

 1500° to J 550°, the results were fairly satisfactory. 



Even under these conditions, the most favorable we nave 

 found, the glass is rarely obtained entirely free from crystal- 

 line material. After the glass had been mechanically separated 

 from the latter, small samples were crystallized at various 

 known temperatures and the products examined microscopi- 

 cally. 



To insure the crystallization of a glass as nearly as possible 

 at a given temperature, the following method was used. An 

 empty platinum crucible was placed in an electric resistance 

 furnace, through the cover of which passed a thermoelement, 

 reaching nearly to the bottom of the crucible, together with 

 * This Journal (4), xxi, 89, 1906. 



