This system has been extensively tried, and indeed used to a certain extent, but the results have not been altogether satisfactory, one of the troubles which have had to be contended with being choking of the retorts.

Of the intermittent processes, the one most in use in America is the "Lowe," in which the coke or anthracite is heated to incandescence in a generator lined with firebrick, by an air blast, the heated products of combustion as they leave the generator and enter the superheaters being supplied with more air, which causes the combustion of the carbon monoxide present in the producer gas, and heats up the firebrick "baffles" with which the superheater is filled. When the necessary temperature of fuel and superheater has been reached, the air blasts are cut off, and steam is blown through the generator, forming water gas, which meets the enriching oil at the top of the first superheater, called the 'carbureter,' and carries the vapors with it through the main superheater, where the "fixing" of the hydrocarbons takes place.

The chief advantage of this apparatus is that the enormous superheating space enables a lower temperature to be used for the "fixing." This does away, to a certain extent, with the too great breaking down of the hydrocarbons, and consequent deposition of carbon. This form of apparatus has just found its way to this country, and I describe it as being the one most used in the States, and the type upon which, practically, all water gas plant with superheaters has been founded.

The Springer apparatus, which is under trial by one of the large gas companies, differs from the Lowe merely in construction. In this apparatus the superheater is directly above the generator; and there is only one superheating chamber instead of two. The air blast is admitted at the bottom, and the producer gases heat the superheater in the usual way, and when the required temperature is reached, the steam is blown in at the top of the generator, and is made to pass through the incandescent fuel, the water gas being led from the bottom of the apparatus to the top, where it enters at the summit of the superheater, meets the oil, and passes down with it through the chamber, the finished gas escaping at the middle of the apparatus.

This same idea of making the air blast pass up through the fuel, while in the subsequent operation the steam passes down, is also to be found in the Loomis plant, and is a distinct advantage, as the fuel is at its hottest where the blast has entered, and, in order to keep down the percentage of carbon dioxide, it is important that the fuel through which the water gas last passes should be as hot as possible, to insure its reduction to carbon monoxide.

The Flannery apparatus is again but a slight modification of the Lowe plant, the chief difference being that, as the gas leaves the generator, the oil is fed into it, and, with the gas, passes through a D -shaped retort tube, which is arranged round three sides of the top of the generator; and in this the oil is volatilized, and passes, with the gas, to the bottom of the superheater, in which the vapors are converted into permanent gases.

The Van Steenbergh plant, with which I have been experimenting for some time, stands apart from all other forms of carbureted water gas plant, in that the upper layer of the fuel itself forms the superheater, and that no second part of any kind is needed for the fixation of the hydrocarbons, an arrangement which reduces the apparatus to the simplest form, and leaves no part which can choke or get out of order, an advantage which will not be underrated by any one who has had experience of these plants. While, however, this enormous advantage is gained, there is also the drawback that the apparatus is not fitted for use with crude oils of heavy specific gravity, such as can be dealt with in the big external superheaters of the Lowe class of water gas plant, but the lighter grades of oil must be used in it for carbureting purposes.

I am not sure in my own mind that this, which appears at first a disadvantage, is altogether one, as, in the first place, the lighter grades of oil, if judged by the amount of carbureting power which they have, are cheaper per candle power, added to the gas, than the crude oils, while their use entirely does away with the formation of pitch and carbon in the pipes and purifying apparatus - a factor of the greatest importance to the gas manufacturer.

The fact that light oils give a higher carburation per gallon than heavy crude oil is due to the fact that the latter have to be heated to a higher temperature to convert them into permanent gas, and this causes an over-cracking of the most valuable illuminating constituents; and this trouble cannot be avoided, as, if a lower temperature is employed, easily condensible vapors are the result, which, by their condensation in the pipes, give rise to much trouble.

The simplicity of the apparatus is a factor which causes a great saving of time and expense, as it reduces to a minimum the risk of stoppages for repairs, while the initial cost of the apparatus is, of course, low, and the expense of keeping in order practically nil.

When I first made the acquaintance of this form of plant, a few years ago, the promoters were confident that nothing could be used in it but American anthracite, of the kind they had been in the habit of using in America, and a light naphtha of about 0.689 specific gravity, known commercially as 76 deg Baume.

A few weeks' work with the apparatus, however, quickly showed that, with a slightly increased blow, and a rather higher column of fuel, gas coke could be used just as well as anthracite, and that by increasing the column of fuel, a lower grade of oil could be employed; so that during a considerable portion of the experimental work nothing but gas coke from the Horseferry Road Works and a petroleum of a specific gravity of about 0.709 were employed.