94 Allen and White — Polymorphic Forms of 



was preserved clown to the corners and edges, which remained 

 as sharp as ever, and the microscope proved that all had 

 been converted into pseudo-wollastonite. We were therefore 

 unable to reproduce Brim's results or to explain them. 



To locate the inversion point, we experimented in two differ- 

 ent ways. On the one hand, we established a point below 

 winch inversion would not occur, by heating powdered wollas- 

 tonite in contact with the hexagonal form at measured tempera- 

 tures for more or less protracted intervals. At a temperature 

 of 1100° there was no sign of inversion after a period of 60 

 hours. We relied upon microscopic evidence for our conclu- 

 sion." 



A charge made up in the same way and held at 1170° for 

 an hour likewise showed no indication of change, but the 

 same charge returned to the furnace and kept at 1190° for 

 another hour showed that the transformation had begun. 

 The change revealed itself in the formation of small grains 

 and patches of pseudo-wollastonite, embedded in the original 

 wollastonite, which were plainly visible between crossed nicols. 

 1190° is therefore the lowest temperature at which we can 

 certainly say that inversion takes place. Changes in the solid 

 state are well knovn to be very sluggish, and this one, as will 

 appear farther on, is a rather extreme case. It may therefore 

 well be that if experiments were continued for very long 

 periods, the true inversion temperature would be found to fall 

 somewhat lower clown. 



By the use of Frankenheim's method, f in which the tem- 

 perature is observed at regular intervals as heat is continually 

 supplied to the mass, we were able to detect a small absorption 

 of heat corresponding to the physical change revealed by the 

 microscope. 



A departure from previous practice was made in the use of 

 a control element. The heating of a furnace can not be made 

 perfectly regular even with storage batteries of large capacity, 

 but the temperature rises with continual slight fluctuations, 

 which, of course, are shared by the charge in the crucible, 

 where they are indistinguishable in their effect on a thermo- 

 element from slight evolutions or absorptions of heat in the 

 charge itself. The control element, which gives the tempera- 

 ture of the furnace, enables a correction to be applied for the 

 furnace fluctuations. It is read alternately with the element 

 in the charge. In general, it is not easy to tell exactly what 

 effect upon the charge is produced by a given fluctuation in 



* For the examination of these mixtures the authors are indebted to Mr. 

 W. Lindgren of the U. S. Geological Survey. 



f Day and Allen, Isomorphism and Thermal Properties of the Feldspars, 

 this Journal, xix, 93, 1905 ; Publication No. 31, Carnegie Institution of 



Washington. 



