GRIMSLEY. | Technology of Gypsum. | 95 
So first, the plaster partially dissolves in contact with the water, 
as Landrin pointed out in his second principle, and as accepted 
by Chatelier. Next, some change takes place whereby, accord- 
ing to Marignac’s experiment, the liquid becomes turbid and 
crystallization begins. Landrin thought evaporation took place 
as a result of the heat formed by chemical combination, and 
that a crystal was then formed which started the crystallization 
through the entire mass. Chatelier showed by experiment that 
evaporation was not necessary, and he argued that by the tak- 
ing up of this water the solubility of the hydrate was decreased, 
and so, on account of the resulting supersaturation, crystalliza- 
tion ensued. 
The solution of the hydrate in these experiments is certainly 
saturated, and all that is needed is something to start the crys- 
tallization. From a study of saturated solutions in the labora- 
tory, it is well known that if crystals are introduced into such 
solutions, crystallization will result and go on until the salt has 
crystallized out. 
The effect of heat on gypsum in the burning of plaster, as 
we have shown, is to remove a certain percentage of water, and 
to break up the small masses of the rock into finer and finer 
particles, microscopic and even ultra-microscopic in size. If 
the heat is not carried too far certain particles through the mass 
may still possess their crystalline form as shown in Figure 12, 
and so they are true crystals though very small. These minute 
crystals in the saturated solution would start the process of 
crystallization. Their growth would cause the turbidity of the 
solution noted by Marignac, and would result in a precipitation 
of small gypsum crystals, thus forming the crystal network 
which constitutes the set of plaster. 
If the plaster is underburned the gypsum is not reduced to 
the proper fineness and uniformity, and so would not permit the 
crystallization to go on in the way it would in the properly 
burned plaster. But of more importance, the hydrate repre- 
sented by plaster of Paris would not be formed. 
If the plaster is overburned, the plaster will be so completely 
comminuted that no minute crystals will be left to start the 
