caustic) |3·4773 |34·773|34-4/5 | 86-7/8 |173-7/8 |347-3/4 | | | | | | | | LIME (pure, caustic) |1·2727 |12·727|12-3/4 | 31-3/4 | 63-5/8 |127-1/4 | | | | | | | | MAGNESIA (pure, | | | | | | | anhydrous) | ·90909| 9·091| 9-1/11 | 22-3/4 | 45-1/2 | 91 | -----------------------+-------+------+--------+---------+--------+--------+ In this ingenious method of alkalimetry it is absolutely necessary that the whole of the alkali in the specimen tested should be in the state of neutral carbonate. If a sample of potash contains any caustic alkali (as the potashes and pearlash of commerce generally do), Fresenius and Will direct it, previously to being tested, to be triturated with its own weight of pure quartzose sand, and about one third of its weight of carbonate of ammonia; and the resulting mixture, placed in a small iron capsule, or a porcelain crucible, to be moistened with water, and exposed to a gentle heat until it becomes quite dry, and all the ammonia is expelled. If the sample contains any bicarbonate or sesquicarbonate, it must be heated to dull redness before being placed in the apparatus and tested. In the case of crude soda (particularly soda ash), the proportion of carbonate of ammonia should be equal to at least one half the quantity operated on. With both alkalies, if the sample contains sulphides, sulphites, or hyposulphites, the same method is to be followed, except that solution of ammonia, instead of water, is to be employed for moistening the powder. To remedy the error which would arise from the apparent amount of carbonic anhydride liberated during the assay, being swelled by the disengagement of ‘sulphuretted hydrogen’ or sulphurous acid from these substances, a small quantity of neutral (_i. e._ yellow) chromate of potash may be added to the alkaline solution in the flask (_A_); by which they will be converted into sulphates, sulphur, and water, which will remain in the apparatus, the carbonic acid only being evolved. “As most sorts of soda of commerce contain one or other of the substances (just) named, and as it is far more simple to add at once some chromate of potassa to the soda solution, than to test the latter for either of the three salts, it is always advisable to make it a rule, in the examination of SODA, to add some chromate of potassa.” (Fresenius.) If the sodium or other carbonate under analysis contains much chloride, the addition of more sulphuric acid than necessary must be avoided, and the carbonic anhydride expelled by gently heating over a warm bath, and not by the addition of excess of acid. [Illustration]