forms of nourishment which agriculture demands in the combinations of its fertilizers, nitrate of soda and sulphate of ammonia, are derived.

At the present time the agricultural industry depends almost entirely for the former of these two substances for plant life, upon the nitrate of soda deposits to be found in Chile, but the supplies available from that country are by no means adequate for our needs. Under these circumstances in order to avert that famine which Sir William Crookes in concert with many other distinguished savants predicts as being in store for the world, further sources of supply have become urgently requisite. In view of the fact that the atmosphere surrounding us is roughly composed of seventy-nine parts of nitrogen to twenty-one of oxygen-which overwhelming preponderance of the first named element demonstrates the extent of our dependence upon nitrogen-it will be seen that an enormous quantity of this gas envelopes the earth. Consequently it is only natural that scientific investigation should endeavor to find some means of producing nitrates from these immediately accessible reserves. The economical and commercial practicability of such a scheme in the interests of science, industry, and agriculture opens up vast possibilities.

Ever since Cavendish first discovered the presence of hydrogen gas and the combustibility of this gas-which when ignited in the atmosphere gives rise to water-as far back as 1781, scientists have been wrestling with this problem.

oxide of nitrogen with the water, is nitric acid.

During the past few years several attempts to turn Cavendish's momentous discovery to practical account have been made, but success has been only mediocre, and not sufficient to warrant the commercial use of the system. Two great difficulties confront the realization of such a project, as these exploiters have found to their cost. In the first place, owing to the fact that the yield of nitric oxide per unit of electric power consumed is so small, the process is impracticable unless an abundant and cheap supply of electric energy is available, such as is afforded by water power; and in the

second place, owing to the fact that the thermal action, which produces the union of the nitrogen and Oxygen by sparking, is reversible in its action-i. e., the same heat which produces the nitric oxide will also cause its dissociation unless it is speedily removed from the influence of the electric spark. It is this second factor which has caused efforts to use Cavendish's discovery and to

mercially practicable basis. In this process the secret and vital factor is the means adopted for the speedy removal of the nitric oxide formed by the electric spark, from the proximity of the latter, whereby the dissociation of the nitrogen and oxygen is prevented. In the course of an experiment with a continuous electric current of forty amperes, 600 volts tension, Professor Birkeland accidentally established contact between the adjacent metallic pieces of the apparatus, which resulted in the production of an intensely powerful magnetic field. As the points were brought into contact a sharp explosion was developed and a flat flame semicircular in form and about 3.9 inches in diameter was produced. He followed up this phenomenon with a further experiment in which he used a current of only two amperes taken from a continuous current dynamo and intensified to 3,000 volts. The contact pieces or electrodes were placed at right angles between the poles of a powerful electro-magnet and resting about .08 inch apart. When the arc was struck there was produced an intense spark the peculiarity of which was that it traveled backwards along each electrode until it broke. Instantly another spark followed in like manner

and so on with such tremendous frequency that the resultant effect was a brilliant half-round flame, it being possible to strike several hundred arcs per second.

With a continuous current it was observed, however, that the halfdisk of flame was always produced upon one side, so in order to obtain a more or less completely round flame Dr. Birkeland resolved to test the phenomenon with an alternating current of very high tension. The result was as anticipated, the arc being struck alternately on each side of the electrode and with such rapidity that the flame was almost circular in shape. In this experiment he also made another important discovery. Whereas with the continuous current, as the voltage was increased so was the volume of the noise caused by the sparking, but with the alternating current the sound produced was decreased, suffused, and quite different in character from that produced with the direct current, because here the number of cycles of the alternating current combined with the intensity of the electro-magnet exercised a peculiar influence upon the phenomenon. In one of the accompanying photographs is shown such a disk of flame produced in a 250-horse-power Birkeland-Eyde electric furnace, the disk being produced with an alternating current of 50 cycles per second with a tension of 5,000 volts.

For the oxidation of the atmospheric nitrogen by means of the electric arc, Messrs. Birkeland and Eyde designed a special type of furnace, which possesses several ingenious features and in which have been embodied the results of the many observations concerning the various and peculiar chemical and electrical

actions that take place in the oxidation of the nitrogen. In this furnace which is of especially stout construction to ensure durability, the electrodes are made of hollow copper tubes .59 inch in diameter, internally cooled by water, and placed two millimeters apart. The atmosphere from which the nitric oxide is to be formed is forced into the internal space of the furnace surrounding the points of the electrodes through a series of channels, and directly the oxygen and nitrogen have been united forming nitric oxide by passing through the electric flame, the gas is swept away through two other canals for the subsequent phases of the process. The stout construction of the furnace and the long life of the electrodes several hundred hours-constitute a prominent feature of the process and are of considerable economical importance, reducing to an appreciable extent the initial outlay for the apparatus and also the expense of maintenance.

When the inventors had sufficiently perfected their process to enable it to be commercially applied, a small installation was laid down in 1903 at Frognerkillens, in Norway. This was a plant of twentyfive horse-power and an alternating current of 5,000 volts was utilized. Three months later a larger plant became necessary and this was completed at Ankerlokken, the energy being increased to 150 horse-power. Developments, however, followed so rapidly that a still larger installation became requisite and this was carried out at Vasmöen, being of 1,000 horse-power, while a few months ago a fourth and still larger plant was laid down at Notodden for the commercial production of the nitrates upon an extensive scale, the energy in this instance aggregating 2,500 horse-power. These works are now in full operation and already extensive additions are in progress for the development of the industry and the increase of the output. Several eminent scientists and physicists of various nationalities have visited the Notodden plant to study and follow the various details of the process, and from the results of their observations it is generally conceded that a successful and commercially practicable solution of the prophesy made by Sir William Crookes. is in a fair way toward being fulfilled.

The electrical equipment comprises three furnaces of identical design constructed in accordance with the principles enunciated by Messrs. Birkeland and Eyde. They are the largest that have yet been constructed for this purpose, being from 500 to 700 kilowatts capacity each. The necessary energy is derived from the neighboring river Tinnfos, and is obtainable at a cost of $3.20 per kilowatt-year.

The atmospheric air is fed into the furnaces through shafts or conduits by means of fans at the speed of nearly 900 cubic feet per minute, making for the three furnaces an aggregate consumption of 2,600 cubic feet, or 75 cubic meters of air per minute. After passing through the flame of the arc, the oxidized nitrogen is gathered in a novel canal collector at a temperature ranging between 500 and 700 degrees Centigrade. As this tem

tigrade. The vapor that is produced during this operation is in itself utilized at a later point for the concentration of the solutions of nitrate of lime. From this cooler the gas is carried through additional similar apparatus wherein the temperature is still further reduced until it approximates fifty degrees Centigrade, which is the most suitable temperature for the following operations. These elaborate cooling systems are placed at the top of two large towers where the oxidation of the nitrogen is carried out. These towers are constructed of iron and are lined with a material which is impervious to the corrosive action of nitric acid. After remaining a short while in these towers the nitric oxide from the electric furnace becomes transformed into peroxide, into peroxide, which is dispatched through conduits to what are called absorption towers, in which the nitric oxide becomes absorbed in the water and forms nitric acid. At this stage the oxide of nitrogen in combination with the oxygen of the air is in

The absorption towers, which are rectangular in shape and of 1412.56 cubic feet capacity each, are ranged on either side of the central passage through the building in two parallel lines. There are four main towers built of masonry and two subsidiary towers the function of which will be described later. In the four principal towers are placed quantities of broken pieces of quartz about the size of the fist, and circulating around the interior in opposite directions are the supplies of gas and water. As the water by continuously playing upon the broken quartz keeps the latter in a constantly moistened condition, the moisture is seized by the gas and becomes charged with formed nitric acid; the other nitrous products present in the nitric oxide and the lesser oxides are re-oxidized in the tower by coming into contact with the air and in turn also yield nitric acid. This process of absorption is continued until the nitric acid solution produced in the towers has reached in course of repeated contacts between the gas and the

water a concentration of fifty per cent (i. e. 50 kilograms of nitric acid monohydrate in 100 litres of liquid), when it is collected in open vats and temporarily stored until required for the manufacture of nitrate of lime.

Although the greater proportion of the nitrous products contained in the gas are arrested during the passage of the latter through these absorption towers it is of great import from an economic point of view that only the minimum of the nitric-oxide mingled with the gas that passes from the electric furnace should elude absorption by the water and thus escape into the outer air. It is for this purpose that the two subsidiary towers. already mentioned are employed. The first of these two tanks is built of wood and is of the same dimensions as the principal absorbing tanks, only instead of being charged with pieces of quartz, a more energetic absorbing medium is utilized. This absorbent, milk of lime, is deposited upon bricks with which the tank is lined and the nitrous gas is seized by this medium, giving an extraction of nitrite and nitrate of lime. This mixture is subsequently dissolved by

the aid of nitric acid into nitrate of lime and peroxide of nitrogen gases, and then re-introduced into the principal absorbing system already mentioned where the process of absorption is repeated in the manner described.

By the time the gas has issued from the fifth tower it contains but a very small percentage of nitric oxide that has escaped, but even the greater part of this remaining percentage is reclaimed by passing the gas through the second and smaller auxiliary absorbing plant which contains live lime. After passing through this vessel the gas is permitted to escape into the outer atmosphere. So complete and thorough, however, is the system adopted for absorbing the nitric oxide in this series of tanks that the quantity of nitrogen gas oxidized in the electric furnaces which successfully escapes is insignificant. Many comparative analyses and observations have been carried out by independent and distinguished scientists to ascertain the quantity of nitric oxide that resists seizure and these have revealed the fact that ninety-five per cent of the oxide of nitrogen produced in the electric furnace is arrested and converted