in it, and rapidly increased in amount until all the nitrogen of the filtrate was in this combination. If nitrification is due to simple oxidation, it is difficult to see why it was so slow in commencing; but, if it is due to an organism which required time to develop in the artificial soil, the delay is at once explained.
Sewage was passed through the soil in this way for four months, with complete oxidation of its nitrogen. As soon, however, as vapor of chloroform, which is known to be inimical to the action of organized ferments, was caused to penetrate the soil, nitrification ceased, and did not recommence after the chloroform was withdrawn. After the sewage had passed unchanged for seven weeks, a small amount of turbid washings of a soil known to nitrify with ease was poured upon the top of the soil. After eight days (i. e., exactly the time required for the liquid to traverse the column of soil), nitrates reappeared in the strata, and continued to be formed as long as the experiment was continued. All these facts point plainly to an organism as the cause of nitrification. It developed in the soil during the first twenty days of the experiment from germs introduced by air or sewage; it was killed by the chloroform-vapor, and reintroduced in the soil-washing.
In 1878 appeared the results of experiments made by Warrington[1] in the Rothamsted Laboratory, which fully confirmed those of Schloesing and Müntz. He first showed that a very considerable nitrification took place in a good garden-soil when a current of air was aspirated through the moist soil, but that hardly any formation of nitrates took place when this air contained vapors of chloroform or carbon disulphide, while vapor of carbolic acid seemed to produce the same effect so far as it was brought in contact with the soil. Thus far the results were simply confirmatory of those of Schloesing and Müntz. Further experiments, however, developed the important fact that nitrification could be brought about in dilute solutions of ammonium salts, by seeding them with a small amount either of a nitrifying soil or of a similar solution which had undergone nitrification. The first experiments were made with the dilute solutions employed in the determination of ammonia by Messler's method, with the addition of small quantities of tartrate and phosphate of potassium, and precipitated carbonate of calcium. The solutions used in later experiments had the following composition per litre:
| Ammonium chloride | 80 | milligrammes. |
| Sodium potassium tartrate | 80 | " |
| Potassium phosphate | 40 | " |
| Magnesium sulphate | 20 | " |
Precipitated calcium carbonate was added to supply the necessary base. By this discovery the way was opened for the easy and fruitful study of the process and of the conditions affecting it.
Since the publication of Warrington's paper, a large amount of
- ↑ "Transactions of the Chemical Society," 1878, p. 44.
