Upon opening the vessel, the surrounding solution crystallises rapidly, enveloping the first-formed set of crystals with others, which, however, are very readily distinguished from them in consequence of their immediately assuming a white opake appearance. Upon taking out the crystals, those first formed are found to be much harder than the usual crystals of sulphate of soda, and, when broken, it is found that the opacity is not merely superficial, but that it penetrates them to a considerable depth, and even at times throughout.
These harder and peculiar crystals are readily obtained by closing up a solution of sulphate of soda, saturated at 180°, in a Florence flask, boiling the solution in the flask so as to expel the air before closing it. Upon standing twenty-four hours, Bne groups of crystals are formed. When the flask is opened, the solution deposits fresh crystals; but on breaking the flask, the latter may be scraped off by a knife in consequence of the superior hardness of the first set.
The hard crystals when separated are found to be efflorescent, like those of the usual kind, and they ultimately give oif all their water, leaving only dry sulphate of soda. When a given weight was heated in a platina crucible, one half their weight passed off as water, the rest being dry salt; they consequently contain eight proportionals of water, or 72 sulphate of soda, and 8×9=72 water.~ The usual crystals of sulphate of soda contain 10 proportionals of water.
When crystallized sulphate of soda is heated in a flask, a part of it dissolves in the water present, whilst the rest is thrown down in an anhydrous state. The solution at 180° appears to contain one proportional of salt 72, and 18 proportionals of water 162; from which, if correct, it would result, that when the crystals are heated to 180° 59, of the salt take all the water, whilst 3 separate in the dry state.
The object of the paper which I have the honour of submitting at this time to the attention of the Royal Society, is to describe
- ↑ Phil. Trans. 1825. pi 440; and Phil. Mag. lxvi. p. 180.
