326 SOLUTION OP CARBONATE OF MAGNESIA. 
These results accord tolerably with those of other inquirers 
who have examined this compound ; the variation or want of 
perfect accordance, probably chiefly depends on the degree of 
dryness of the preparation examined, or on the quantity of 
water retained in the powder admitting of expulsion at 212°, 
which water, being hygrometrical, at least in part, must vary 
in the degree of dryness of the atmosphere to which it is ex- 
posed. 
The method by which these two compounds of magnesia 
and carbonic acid were analysed, was a simple one, admitting 
of considerable accuracy. 
The quantity of water expelled at a temperature of 212°, is 
determined by exposure of an hour or more to the heat of a 
steam bath ; the quantity of carbonic acid, by acting on the 
compounds, very carefully weighed by muriatic acid, satu- 
rated with carbonic acid over mercury in a graduated tube ; 
and the quantities of magnesia and of water expelled at a 
higher temperature than 212°, by the action of a red heat, 
continued for two or three hours, till no further loss of weight 
was produced by a continuance of the high temperature. In 
estimating the proportion of carbonic acid, the calculation was 
made on the ground that 100 cubic inches of this gas weigh 
47.262 grains. 
A few words relative to the properties of the first-mention- 
ed carbonate. Its tendency to crystallize is remarkable ; 
however obtained, even when rapidly separated by the expul- 
sion of the excess of carbonic acid by heat, it has been depo- 
sited in a crystalline form. This form is not obvious to the 
naked eye , but, when the powder is examined by the micro- 
scope, each particle is found to be a distinct prismatic crystal. 
And the persistence of this form is no less remarkable ; it is 
not destroyed by decomposition ; the powder after ignition, 
after the expulsion of the whole of the water and carbonic 
acid, under the microscope shows no alteration ; each particle 
is still prismatic, and, when moistened with water, is trans- 
parent. 
It is asserted that this carbonate readily loses the water with 
