It is sufficiently obvious, from the construction of this retort, that, upon its being charged with fresh coal, and already of a red heat, the process of car. bonization will proceed most rapidly at first ; a crust of coke being speedily formed next to the heated metal, and this constantly increasing in thickness, prevents the free transmission of the heat, and the decomposition of the coal is consequently retarded more and more as the crust increases in thickness. As a remedy for this evil, and to which all the above forms of retort are liable (unless made of extremely small capacity, and thereby unfitted for practical use); one of the shape represented In fig. 6 was coustruct ed, having an illiptical cross section, as shown at e, and placed, as regards the grate and brick-work, much in the manner of fig. 2, the letters referring to like parts. When filled about half full of coal, and previously brought to, and afterwards kept at a strong red heat, the quantity of gas produced was about 30 per cent. more than what the retort fig. 5 yield ed, and the illuminating power, bulk for bulk, con siderably increased. And the result of a great num. ber of experiments, continued for a length of time, and under the varying circumstances of rapid and slow distillation, and large and small masses of coal, showed decisively, that the greater the ra pidity with which the distillation was effected, the greater was the quantity of gas produced (from an equal weight of coal), and the more intense the Mu.. initiating power of that gas, volume for volume. The quantity of liquid product was also lessened, and the weight of the residual charcoal not so great as when the process was conducted with less rapidity.
The most advantageous results wereobtained when the retort was heated to a bright red heat; when in creased much beyond this point, so as nearly to ap proach a white heat, the production of gas was ma terially lessened.
We witnessed the above experiments, which were made at Soho, in 1807, and as their results are of the first importance, the particulars of one series are given in the following table. The coal made use of • was from the neighbouring collieries in Staffordshire, and of inferior quality, but that circumstance does not affect the comparative results of the experiments. The best Staffordshire coal will yield about 530 cubic feet of gas from each cwt.
The above table gives the results obtained from 56 lbs. of coal. This quantity was broken into small pieces, and mixed so as to render any portion of it of equal quality with the rest ; it was then •tlivided into four equal parts, each weighing 14 lbs.,• and these submitted separately to the action of the fire. In experiment No. 1, the retort was heated so as to complete the distillation in three hours ; in No. 2, seven hours, &c. as specified in the second column. The third column gives the number of cubic feet of gas obtained in each case ; the fourth, the weight of the charcoal remaining in the retort; and the fifth, that of the liquid products : the last column shows the number of hours which each portion of gas sup..
plied the light of one candle ; and these numbers are, therefore, expressive of the comparative value of the gas obtained in each experiment from an equal weight of coal. Generally, it was observed, that, in keep ing these elliptical, or other flat-shaped retorts at a bright red heat, and introducing the coal into them when in that state, the quantity of gas was increased from one.third to one-half (compared with what the retorts 6g. 5 yielded); the time of distillation great ly shortened ; and the quality of the gas muoh im proved; giving results, on the whole, not less favour able than those particularized in the above table.
The degree of heat, however, which was found to be most advantageous for the production of gas, was very destructive to the cast iron of which the retorts were formed ; and to preserve these from rapid oxi dation, their lower sides were made to rest upon thin firebricks, as represented in fig. 7, where the section of the retort is varied somewhat from the fi gure of an ellipse, as shown at n, to apply with great er exactness to the upper side of the protecting bricks g g; i i, other bricks set on end to support the former, and standing upon an arch f, over the fire place ; the flame, rising through the opening, e, and circulating at h h, underneath the bricks g g, escapes through two side flues, one of which is shown at k ; and, after rising and passing over the top of the retort, enters the chimney : a, 6, c, d, refer as in the former figures.
A somewhat simpler arrangement is shown in fig. 8, where the under side of the retort a is curved up. wards, so as to apply directly to the back of a thin fire-brick arch, e, through which the heat is trans mitted. The flame from the fire d, after acting against the brick arch, divides and escapes through the flues, f f ; and, after uniting again at the upper side of the retort, ascends into the chimney. The conducting pipe is here shown issuing from the low• er part of the door-piece b, and descends, instead of rising or passing away in a horizontal direction, thereby lessening its liability to be choked up by tar condensing therein.
This construction of retort is probably not inferior to any that has been used. When made of the di mensions represented (being drawn to a scale of one fourth inch to the foot) it will contain about one cwt. of coal, when somewhat more than half filled, and produce gas at the average rate of 100 to 150 cubic feet per hour, according to the nature of the coal employed. With proper attention, it will work off six charges of coal in 24 hours ; and last, when in constant use, from nine to twelve months.