74 NEW YORK STATE MUSEUM 



contains 6.2 per cent water. The commercial plaster containing in 

 the mean 7 per cent water is then almost exclusively made up of 

 this hydrate." 



In the past 10 years a number of chemists have taken up the 

 question of the decomposition of gypsum, the formation of the 

 half hydrate and the anhydrite and their mutual relationships. 

 Among the number are Armstrong, Van't Hoff, vShenstone, Cun- 

 dall, Cloez etc. [see Bibliography iov references]. The most re- 

 cent investigations- are those by W. A. Davis^ who, moved by the 

 umcertainty and lack of uniformity in the results previously ob- 

 tained, has carried out a series of careful experiments, which he 

 presents along with a summary of the work performed by others. 

 This very valuable contribution is presented herewith in abstract. 



At the time Davis entered upon his investigations there were 

 recognized, as derived from gypsum, the half hydrate, formed at 

 128° and decomposed at 163° [Le Chatelier] ; the soluble anhy- 

 drite which according to Van't Hoff was formed directly by heat- 

 ing gypsnm in a vacuum over concentrated sulfuric acid without 

 the intermediate formation of the half hydrate, and natural anhy- 

 drite, which can be formed by strongly igniting the" soluble anhy- 

 drite. The soluble anhydrite is very soluble in water and sets very 

 rapidly to a hard mass. 



Davis first heated a series of gypsum samples at temperatures 

 between 98° and 130° C, and measured their successive loss in 

 weight or loss in water and derived the following curves [fig./]. 



The influence of the state of division of the gypsum is clearly 

 noticeable. 



The striking feature in the curves, however, is that loss of water 

 does not begin immediately after the material is heated. In one 

 experiment a whole hour is shown to elapse before any water is 

 given off. During this period tlie monoclinic gypsum is undergoing 

 a crystallographic change to the orthorhombic system, or in other 

 words CaSO.,. 2 H.O is dimorphous, as are nearly all the hy- 

 drated sulfates of bivalent metals. Further proof of this is shown 

 in the behavior of the half hydrate on setting. When the plaster 

 first sets into a coherent mass, microscopic investigation has shown 

 that all the crystals present have a straight extinction and are 

 probably orthorhombic. Gypsum crystals subsequently appear be- 

 cause the orthorhombic form is labile at ordinary temperatures 

 and, in from a few hours to a few days, changes to gA^psum. This 

 perhaps explains the fact that plaster when first set contracts (while 

 the orthorhombic crystals are forming) and expands (at the change 

 to monoclinic). At first sight the curves might suggest Van't Hoff's 

 idea that no intermediate half hydrate is formed, but owing to the 



^ Soc. Chem. Tnd. Jour. 1907. 26:727. 



