connected to each other by means of special bends. Before good mastic was available, these bends had conical ends, which were ground into funnel-shaped sockets on the receivers (see Fig. 10). Later on the sockets had the usual Fig. 10.

form (Fig. 11), and the bends were luted in by a mastic or putty made of asbestos and silicate of soda, sometimes with the addition of a little barium sulphate, to harden it. Care has to be taken that the sodium silicate is diluted with

Fig. 11.

about 20 parts of water, or else the mastic absorbs moisture, and by its swelling cracks the sockets. There is a good deal of manual labour connected with these receivers, as they have to be emptied from time to time, and as condensers they are by no means perfect. A simple consideration will show that it is only the outer surface of the pot which effects condensation, and therefore the whole of the inner core is useless. At the same time these receivers are nearly always partly full of warm acid, which consequently largely reduces the available condensing surface, and affords constant contact for gases, so that any impurities are readily absorbed by the liquid. It is also natural that each receiver yields a different strength of acid. For a retort of 10 cwts.

of nitrate there were always at least a dozen of receivers required.

A better condenser was that of Plisson and Devers (Fig. 12), which may still be found in some old works. It consisted of a number of pots fitted one in the top of the other. The first acid ran into an independent receiver, and connected at the bottom by means of an overflow pipe, so thence into a separate carboy. The lowest receivers were that the acid could only reach a certain level, and was drawn off from the last receiver. In a similar way the condensation is sometimes effected by means of receivers arranged in steps (Fig. 13).

Marchal made a good improvement on the Plisson and Devers system, which is shown in Fig. 14. There is only one range of receivers, but each of them has a conical bottom with a trap, and these bottoms fit into a main discharge pipe. By these means the receivers were always empty, and the whole of the acid obtained is of the same strength. In order to further increase the condensation, a short pipe was interposed between the receiver and the bend.

In many factories earthenware worms are used to partly replace the large number of receivers. Such worms, when placed in a tub of water, are a very effective means for cooling, especially when they have a corrugated section, such as the sample shown on the table. They have on the other hand great drawbacks. Unless they are made from very good material, they are liable to crack with the difference in temperature between the cooling water and the hot gases. And also it is, with the rough usage they get in a factory, very easy to knock off either the inlet or the outlet piece. But the greatest objection is in the fact, that the gases enter them at the top, and unless the gases are entirely condensed in the worm (which as a matter of fact they never are), there will be either a stoppage of the draught, and consequently slower distillation, or a certain pressure on the worm to enable a mixture of liquid and gas to pass through. In either case all the impurities will be absorbed by the acid, and all the water will be condensed, so that a weaker acid will be the result.

In this connection I may mention a process patented to the Chemische Fabrik Griessheim. They put the collecting vessel between the retort and the condenser, which is on the Liebig's principle, or a worm, where the gases enter at the bottom. Compressed air is blown into the collecting vessel, and this bleaches the acid, whilst the nitrous acid and chlorine pass through the condenser.

Another system is that of Valentiner in which a vacuum is formed behind the condenser, and the temperature of distillation thereby reduced.

In order to get a largely increased cooling surface, and to do away with the receivers, which by reason of their shape are unequally strained in different places, which makes them liable to break, I have devised a system of condensation by means of long earthenware pipes. This has only been possible by using the excellent material supplied by Rohrmann, such as you see here.

Originally each pair of pipes was connected at the bottom by means of a trapped bend, (Fig. 15), and the outlet of this bend went into a main discharge-pipe. The many joint necessitated a good deal of attention and lead to dripping of acid. I then devised the "chamber-pipe," this is a pipe with a number of sockets for the insertion of the vertical condensing-pipes, and a division between each two sockets, so as to allow the gases to pass from one pipe into the other, whereas the acid runs constantly off by means of small bends connecting one chamber with the other. A section of one of these pipes is exhibited here, and the whole arrangement of the condensing battery, &c. can be seen in Fig 16. This condensing battery works very well, as the condensation is rapid whilst at the same time the steam and chlorine have a perfectly free passage to the tower without coming into contact with the acid. The distillation can be carried on very quickly, the usual time being only 12 or 14 hours. These pipes stand any reasonable variations of temperature, in fact the cylindrical form is the most favourable one for resisting such stresses as are thus set up, and only gross illusage will cause breakage.

Although the acid made in this condensing batteries is quite free from chlorine and sulphates, and contains comparatively little hyponitric acid, yet with very strong acid, and fairly quick distillation the hyponitric acid is rarely less than 2 per cent. This is more than is desirable, because now a days manufacturers of dynamite, smokeless powder, &c., require acid, which whilst having the maximum possible strength contains as little hyponitric acid as possible. Subsequent bleaching would weaken the acid again, and therefore very slow distillation would have to be

resorted to, which could only give a small proportion of strong acid. It is well known, that it order to convert nitrous acid into nitric acid the presence of air and water are necessary. The reaction 3 NO2+ H2O=2HNO3 + NO takes place, and the nitric oxide is by the air constantly oxydised and again transformed by the water. Now, even the strongest nitric acid, which is made on a large scale, contains about 4 per cent. of water, which is ample for a much larger quantity of nitrous acid. If, therefore, hot air is blown into the gaseous mixture, before any of it has commenced to condense, the nitrous acid should be transformed into nitric acid. This consideration led me to employ an earthenware injector (Fig. 17) on the top of the Fig. 17.

retort, and it has been a complete success.

A small coil of gas pipes is placed into the flue of the retort, and compressed air goes through it to the injector. This simultaneously causes a reduction of pressure in the retort, thereby reducing the temperature of distillation, and transforms the nitrous acid into nitric acid. The expansion of the compressed air also assists in cooling the gases after leaving the injector, and the air carries the impurities away from the condensers. This injector enables one to produce nitric acid of 96 per cent. monohydrate, with under per cent. hyponitric acid, and also to make nitric acid of 1.420 sp. gr., which is free from every impurity except a slight trace of iron. With other condensing systems there is at best but 80 per cent. of the total yield obtained as strong acid, with my system the whole of the yield contains 96 per cent. of monohydrate. In distilling with waste acid from the manufacture of nitroglycerin, which contains a great quantity of organic compounds, the nitric acid produced without the injector has as much as 16 per cent. hyponitric seid, whereas with the injector it is reduced to below 1 per cent., which shows conclusively its good work.

With the necessary modifications this battery is admirably adapted for such operations as the denitration of waste acids, the manufacture of arsenic acid, &c., where a prolongated contact with air and steam is required.

Finding Mr. Rohrmann's pipes to be of such excellent quality made me bolder, and I have recently devised a water-cooled condenser (Fig. 18) on the same principle as the battery just described. It has enabled me to reduce the number of pipes from 20 to 5, and yet to get about 5 per cent. more strong acid out of it. In fact this condenser yields now 97 to 98 per cent. of the total possible quantity of acid at a strength of 96 per cent. monohydrate,

with about per cent. of hyponitric acid only, and the remaining 2 or 3 per cent. go to the tower. The cost of this condenser, including injector, collector, &c., is only half of the large battery, viz., about 45. per retort of 12 cwt. nitrate. A plant erected to this system takes now only about a quarter of the floor space, which is required with others for the same quantity of production, and a charge is easily made every day from each retort, as against one in 48 hours with other systems. The coal consumption is only 1 cwt. per charge, or about 0175 lb. of coal per lb. of nitric acid.

It is a matter of course, that no system of condensation is perfect, and that the stronger the acid the more fumes will ultimately escape it. As the nitric fumes are in most cases a nuisance to the neighbourhood, and as in this country the Alkali Act provides, that the gases escaping should not contain more than a minimum of free acid, special contrivances have to be resorted to, which are commonly known as towers. In the Plisson and Devers system, the last column of receivers was filled with pumice stones, on to which water trickied, and the uppermost receiver was connected with a cooling worm.

A more effective way was the use of coke towers (Fig. 19), these had a Segner's wheel on the top to distribute the water. They are still largely used, but sometimes they choke, and sometimes the acid finds a straight and short path through the coke, and so minimises the action of the tower. A coke tower to serve six retorts of 10 cwt. each must be 48 ft. high and 3 ft. diameter, which means a costly structure to contain it.

Since then attempts have been made to construct towers similar to the column apparatus in the distillation of alcohol. Such a tower is shown in Fig. 20. It consists of a number of saucers with a hole in one corner through which a small pipe is fixed. The acid always stands as high as the pipe, thus presenting a large surface for absorption. The liquid falls on to a plain part of the next saucer, the pipe being wide enough to allow the gases to pass through from below. By far the best is the plate tower designed by Professor Lunge and made by Rohrmann, which is now used for a large variety of purposes. It is shown in Fig. 21. You see there is the usual bottom vessel with a number of cylinders placed on the top of each other, but instead of being filled with coke, it contains a large number of thin earthenware plates, each of which has 1,200 small perforations, each perforation having a small rim so as to cause the liquid to reach a certain height before falling through. The enlarged section of a plate is shown in Fig. 22. Each plate rests on a thin earthenware ring. Both plate and ring will probably strike you as clever pieces of workmanship, considering that they are both 26 in. in diameter and only in. thick. The plates are so placed, that a hole in one corresponds to a plain part in the next, so that the liquid dropping down is scattered into small drops, and consequently the whole of the tower is filled with a cloud of minute drops, which offer a very large surface for absorption.

The tower as it is designed here, is only 10 feet high without the base, and has the same absorbing power as the coke tower 48 feet in height mentioned before, but whereas the coke tower yields at best an acid of 1.260 sp. gr., the acid from the Lunge-Rohrmann tower marks 1·380. This is due to the fact, that with a coke tower every drop of water falling through the distributing cover on the top runs without stoppage to the bottom, whereas in the LungeRohrmann tower it only causes a drop from each plate to fall on to the lower one, and ultimately one from the last plate to fall into the bottom vessel. With the coke tower it is therefore necessary to have a fair stream of water running down to prevent escape of gas, whereas with the Lunge-Rohrmann tower the water can be admitted even drop by drop, and therefore does not block the draught.

The proper working of a nitric acid plant can easily be regulated by a sight glass, which is placed into a pipe just before the gases pass into the chimney. They should only show the faintest yellow tinge here.

Having thus given a broad outline of the manufacture of nitric acid, I have only a few more words to say as to the quality of the acid. With the invention of smokeless