75 NEW YORK STATE MUSEUM 



lization. Where the pilaster is shghtly overburned, the crystals are Jj 

 extremely fine and crystallization goes on very slowly and im- 

 perfectly." 



While the presence of any unbiirned gypsum may hasten or other- 

 wise influence the setting of some plasters, it does not appear that, 

 the process is absolutely dependent upon that condition for its start. 

 In the case oi soluble anhydrite there is rapid setting on addition of 

 water, which is hardly explainable by the view taken by Grimsley. 



The approximate solution of the problem is undoubtedly to bej 

 found in the work of Davis. " It has always been assumed that] 

 the setting of plaster is due to the regeneration of gypsum by the 

 action of water on the half liydrate. If, however, the setting of 

 the half hydrate be carefully observed by means of the polarizing 

 microscope, not a single gypsum crystal can at first be detected in the 

 set mass ; the cake of set material, during the first quarter of an hour 

 after it has hardened to a coherent mass, which is only slightly 

 indented by the finger nail, is made up of crystals- showing a straight 

 extinction only, and therefore probably orthorhombic. The first 

 product of the setting of the half hydrate (or soluble anhydrite) is, 

 indeed, the same orthorhombic dihydrate as is produced in the first 

 stage of the dehydration of gypsum. Gypsum crystals subsequently 

 make their appearance within the set mass, owing to the fact that 

 the orthorhombic form of the dihydrate is labile at the ordi- 

 nary temperature, and undergoes change more -or less rapidly 

 during the course of several hours or several days, the time vary- 

 ing greatly — into the more stable form of gypsum. The series 

 of changes 



a CaSO.. 2H2O '^I^ fi CaS0..2.H O "^I^ CaSO^. ^ H.O 



Gypsum (monocHnic) (orthorhombic) Half hydrate (orthorhombic) 



is, indeed, strictly reversible. Before gypsum can undergo dehy- 

 dration to form the half hydrate, it passes into the orthorhombic 

 form of the dihydrate, and the latter is also the first product of 

 the hydration of the half hydrate." 



Some recent experiments have been made by Leduc and Pellet^ 

 on the relation of calcining temperature to the setting of plaster. 

 They calcined for an hour or more pure alabaster at various tein- 

 peratures and mixed the plaster formed with 85 per cent water. 



The results of their experiments are as follows : 



» ■ \ ^ 



> Le Genie Civil. 1906. 49:253, 



II 



