444 l>r, Thomson on [Dec. 



this octahedron, by supposing the four angles of the rhomb 

 A c B rf to be replaced by tangent planes ; but this is unneces- 

 sary, as the two salts differ from each other in their composition; 

 and as the pyramidal termination is quite different from the 

 summits C, D, of the rhomboidal octahedron which constitutes 

 the primary crystal of common carbonate of soda. 



100 parts of water at the temperature of 63° dissolve 63'87 

 parts of these crystals. This is rather more than the quantity 

 dissolved of common carbonate of soda at the same temperature, 

 provided any confidence can be put in an oldset of experiments 

 on the solubility of this salt made in ray laboratory, by which I 

 find that 100 parts of water at 65° dissolve 51*03 parts of the 

 crystals. 



When heated, the salt partly hquefies, but not completely, as 

 is the case with the octahedral carbonate. A portion remains 

 always solid ; and when the salt is cooled, imperfect crystals 

 soon appear. This leads to the supposition that there exists a 

 third species of crystal of carbonated soda containing still less 

 water of crystallization. Its specific gravity is 1*51. 



To determine the composition of this salt, a variety of experi- 

 ments were made, the most important of which I shall briefly 

 state. 



(1 .) 50 grains of the salt were dissolved in water, and neutral- 

 ized with nitric acid. The solution being tested by muriate of 

 barytes was found to contain no trace of sulphuric acid ; but 

 nitrate of silver threw down a quantity of chloride of silver, the 

 weight of which was 1*58 grain. This is equivalent to 0*39 

 grain of chlorine, or 0-65 chloride of sodium; so that 100 grains 

 of the salt contain 1*3 grain of common salt. 



(2.) 50 grains of the salt lost, when exposed to a red heat in 

 three several trials, 28*09 grains. Now common salt being anhy- 

 drous, we must deduct it from the weight of the carbonate of 

 soda employed. This being done, we find that 49'35 grains of 

 pure prismatic carbonate of soda lose, when heated to redness, 

 28*09 grs. ; consequently 100 grains of the salt would lose by 

 this treatment 56-92 grains of" weight. Tliis is the amount of 

 the water of crystallization. 



(3.) Into a small Woulf's bottle furnished with two mouths, 

 one of which was stopped with cotton wool, were introduced 

 through the other mouth 50 grains of the crystals of this salt. 

 The bottle contained a quantity of concentrated and colourless 

 nitric acid ; it had been previously carefully weighed, and it was 

 held in an oblique position when the crystals were introduced. 

 The stopper was immediately put into the mouth of the bottle, 

 and it was left at rest till the solution was completed. The 

 stopper was then withdrawn, and a small sucker introduced, by 

 which I extracted uU the carbonic acid gas contained in the 



