viated only slightly from the individual trials, showed that the gas required 1' 7" to fill the balloon, and therefore that 43304-76 cubic inches have passed through the muriatic acid. The examination of the muriatic acid used in the experiment gave 0.198 grm. chloride of platinum and ammonium, corresponding to 0.0152 ammonia. If we assume as the composi tion of the gases that formed at a depth of eight feet, we can easily calculate the quantity of coal necessary to produce the above 43304-76 cubic inches of gas. According to analysis, 1000 cubic centimetres of this gas contain 547-7 cubic centimetres of nitrogen. We have already seen that nitrogen is not produced from the materials introduced into the furnace, and hence all the amount present must have been introduced by the blast as atmospheric air, which, burning before the tuyère, mixed with the gaseous products of distillation in the upper parts of the furnace, and produced the above 547.7 cubic centimetres of nitrogen. But as this amount of nitrogen is derived from atmospheric air, it implies that 143-84 cubic centimetres, or 0-2066 grm., of oxygen has been consumed in the lower part of the furnace by uniting with 0.1549 grm. of coke, in the formation of carbonic oxide. But, as has already been shown by a previous experiment, 0.2304 grm. of coal must have been distilled to produce the 0.1549 grm. of coke; and as the above quantity of coal is required to generate one litre of the above gaseous mixture, 163.5 grm. must have been employed in the generation of the 43304 cubic inches of gas washed by the muriatic acid. Hence it follows that only 0.0093 grm., or 3.77 per cent. of the ammonia generated from 100 parts of the coal (which according to our experiments amounts to 0.2463 grm.) pass over along with the gases; so that the remaining 0.2370 grm., or 96 23 per cent. of ammonia, must have been condensed in the water of distillation found in the tube. In fact, we ascertained that the lead tube contained a clear liquid so strongly charged with ammonia as instantly to render blue reddened litmus paper held over it. These experiments prove how easily the ammonia might be condensd, even without the intervention of an acid.

It will be observed that the gases from the inferior parts of the furnace contain cyanogen, the presence of which is highly interesting, not only in a theoretical, but also in a practical point of view. This gas appears immediately over the point of entrance of the blast, and again disappears at a small elevation above it, so that at the top of the boshes only traces of it are observed. The compound of this substance with potassium appears to play a most important part in the furnace, although its functions have apparently been altogether overlooked. This is the more surprising, as it has long been known that cyanide of potassium effloresces on the walls in certain states of the furnace. We have been fortunate enough to elucidate the conditions of its formation, and to fix its region in the furnace. In obtaining information with regard to the formation of this cyanogen gas, it was necessary to withdraw the gases from the vicinity of the hearth of the furnace, and through the kindness of Mr. Oakes, we were enabled to bore a hole over the Front of the furnace, two feet nine inches above the level of the tuyère. As soon as this hole was made a gas issued from it, possessing strong illuminating powers, and burning with a yellow flame, from which came abundant vapors of white smoke. On introducing an

iron pipe into the hole, without allowing it to pass into the furnace, it was retained sufficiently cool to prevent its fusion, and we were enabled to collect the volatile products. The gases which poured out of this tube under a pressure of several feet of water, were so richly laden with vapors of cyanide of potassium, that we were obliged to use precautions in approaching its opening, so as not to suffer injurious consequences from this poisonous material. Although the conducting tube was twenty-two feet in length, the amount of cyanide of potassium carried along with the gas was so considerable as to fill, in a very short time, glass tubes of one-eighth of an inch in width, and we therefore endeavored to obtain an approximative result as to the amount which thus passed over with the gases and escaped condensation in the long tube. The opening of the iron tube was connected with an empty Woulf's-bottle, to which another was attached containing water, in such a manner that the gas had to stream through a layer of four inches of the latter. The first of these bottles became quickly filled with a rich white sublimate of dry cyanide of potassium, while the water in the second became a tolerably concentrated solution of the same substance. It was now necessary to determine the quantity of gas which passed through the bottles, and this we ascertained by accurately noting the time employed in the experiment, and the exact period necessary to fill a balloon of known capacity attached to the second bottle.

1. Duration of the experiment, 24 minutes.

2. Mean time required to fill the balloon, 25 seconds. 3. Capacity of the balloon, 380-8 cubic inches.

Hence it follows that 21,933 cubic inches of gas passed through the bottle. The cyanide of potassium in the Woulf's-bottles and their connecting tubes were made into one solution, which weighed 381 024 grms., and 129.211 grms. of this solution yielded 0-62208 grm. cyanide of silver, which was easily decomposed by fuming sulphuric acid. Hence, in the 21,933 cubic inches of gas which had passed through the flasks, there must have been 0.8944 grm. of cyanide of potassium held in mechanical suspension. We have already seen that the gas possessed the following composition:

The 21,933 cubic inches of gas, admitting only its approximative estimation, its temperature being neglected, must contain 1192.97 grains of carbon, corresponding to 1774-79 grains of coal. Hence, out of 100 parts of coal at least 0.778 of cyanide of potassium are generated; and as 31,200 pounds of coal are consumed every day in the furnace, it is obvious that at least 224-7 lbs. of cyanide of potassium are generated daily in the Alfreton furnace, and hitherto have been altogether lost.

When the iron tube used in the experiment was withdrawn from the

furnace, it was found to be encrusted with melted cyanide of potassium, which speedily deliquesced in the air. On bringing it in contact with water, a considerable quantity of hydrogen gas was evolved, obviously due to the presence of reduced potassium, or to its compound with car bonic oxide. In the tube itself at least three or four times the amount of cyanide of potassium was condensed, so that we may be quite certain that the amount formed is far more considerable than we have stated. With these unexpected results before us, it became of importance to determine the origin of the large quantity of potassium in the furnace. At first, we conceived that it might be present in the limestone, which not unfrequently contains carbonate of potash, according to the researches of various chemists; but on examining as much as 30 grammes, we were unable to detect in it the smallest trace. However, we were informed by Mr. Charles Oakes that he had detected the presence of potash in the iron ore, and we are glad to be able to confirm the result of this talented young chemist. We have subjected an average sample of the calcined ore to analysis, according to the methods usually employed in such cases. The quantity used in the analysis was 2.324 grms., which yielded 1-400 peroxide of iron, 0.153 alumina, 0.145 carbonate of lime, 0.202 phosphate of magnesia, and 0.599 silica. In order to estimate the amount of potash, 17,936 grms. were ignited with carbonate of barytes dissolved in muriatic acid, and the bases, after separation of the silica, were as much as possible precipitated by carbonate of ammonia. The solution was then freed from barytes by means of sulphuric acid, and the excess of the latter removed by evaporation with chloride of strontium. In this way the remaining bases were converted into chlorides soluble in alcohol, and the solution mixed with chloride of platinum and evaporated to dryness in the waterbath, left a residue, which, treated with alcohol to dissolve out the other chlorides, consisted of pure chloride of platinum and potassium, and weighed 0-689 grm. This analysis gives the following composition for the cal cined ores:

Another source of the potash was found to be in the coal, although to less extent than in the ore; 1.627 grm. of the coal, dried at 2120, yielded 0.122 grm. of water; 0.2865 grm. gave 0.7865 grm. of carbonic acid and 0.153 grm. of water; 2.887 grms., heated with the mixture of soda and lime, gave 0.0801 of chloride of platinum and ammonium. The experi ment, repeated with 5.687 grms., gave 0.175 of the above salt; 13.059 grms. of coal yielded 0.3505 grm. of ashes, which did not effervesce with acids; and this quantity of ashes, treated as in the case of the iron ore,

gave 0046 grm. of chloride of platinum and potassium. The coal, therefore, is composed as follows:

The quantity of ironstone consumed by the furnace every twenty-four hours is 33,600 lbs., and that of coal 31,200 lbs., so that the furnace receives every day, in these materials, 271.48 lbs. of potash, corresponding to 377-3 lbs. of cyanide of potassium. Thus these analyses render intelligible the large quantity of potash which we observed in the inferior parts of the furnace.

But we have yet to discuss the most interesting and important question bearing upon the presence of cyanide of potassium, viz. the origin of its cyanogen. We know how easily ammonia, in contact with carbon at high temperatures, is converted into cyanide of ammonium. Hence we should be apt at once to admit that the formation of cyanogen is due to the ammonia so freely evolved from the coal during its distillation; and if this view were correct, the existence of one must arise from the destruction of the other. But when we view more closely the circumstances under which the cyanogen is produced, we are compelled to admit that the ammonia cannot take part in its formation. The hearth, at which the formation of cyanogen takes place, is the deepest and hottest part of the furnace, and it would be absurd to suppose that the coal which reaches this part could contain a trace of ammonia, exposed as it has been for eighty hours to a red heat, and in one part to a temperature sufficient to reduce potash.— Hence we are compelled to adopt the only remaining conclusion, that the nitrogen of the air introduced by the blast combines directly with carbon to form cyanogen. This direct formation has been argued for by various chemists, and supported in this country by the experiments of Fownes and Young. But as it has been objected to experiments of this kind, that they were instituted without reference to the ammonia of the air, which is apt to be taken by most substances exposed to it, it is scarcely to be wondered at that the direct generation of ammonia is still doubted by distinguished chemists. We have, therefore, thought it necessary to determine this disputed question by an experiment which seems to banish all sources of error. We have led simultaneously, and under exactly the same conditions, a stream of carbonic acid and another of nitrogen, at a very high temperature, over a mixture of two parts of charcoal from sugar, and one part of chemically pure carbonate of potash, and have subjected the products to careful examination.

The apparatus used by us in these experiments is represented in fig. 9; a is a gasometer, from which a uniform stream of air is made to pass through a bottle filled with sulphuric acid (b), and then through a gun-barrel (cc)

filled with copper turnings. The gun-barrel is kept in a furnace, so that the air passing through it is thoroughly deprived of oxygen, and passes into the gun-barrel (dd) filled with the mixture of charcoal and potash, and heated to a temperature sufficient to reduce potassium. In the same furnace is placed another gun-barrel (ee), filled with the same mixture, and over which is passed a stream of dry carbonic acid from the apparatus fg.

When both the systems were completely filled, one with nitrogen, and the other with carbonic acid, the streams of gas were allowed to pass slowly over the mixture of potash and charcoal, both the tubes in the same furnace being kept at a temperature sufficient to reduce potassium. The gas passing out of the tube filled with carbonic acid had all the characters of pure carbonic oxide, being transparent, inodorous, and burning with a pale blue flame, without depositing any kind of sublimate. The tube over which nitrogen passed emitted a gas richly laden with a white smoke of cyanide of potassium, which sublimed in such quantity as to stop the conducting tube. When the nitrogen was passed so slowly through the sul phuric acid that the bubbles passed only once in a second, its absorption by the potash was complete, and no gas appeared at the mouth of the gunbarrel; but as soon as the temperature was lowered, so as to be under that necessary for the reduction of potassium, the absorption of nitrogen ceased. The contents of the tube over which carbonic acid had passed were examined after cooling, without the detection of the smallest trace of cyanide of potassium. The mixture treated with nitrogen, on the other hand, dis