24 THE ANALYST. ESTIMATION OF NITROGEN IN COMMERCIAL SUBSTANCES CONTAINING NITRATES. IN the Chenliker Zeitzcng, Dr. Paul JVagner, detaiIs a very simple apparatus for this purpose, which he has found very useful in the analysis of Chili saltpotre and nitrated manures. The actual principle of liberating nitric oxide by means of the action of a ferrous salt is, of course, not new, but the arrangement of the apparatus is so simple and inexpensive that we quote it for the benefit of any of our readers disposed to try the process and report their success or otherwise in its application. The apparatus is simply a flask of 200c.c. capacity, fitted with an india-rubber cork and two glass tubes. One of these is an ordinary bent tube serving for the delivery of the nitric oxide formed, while the other is a funnel tube fitted with a glass stopcock and having the lower end somewhat narrow and above the liquid in the flask.A solution of ferrous chloride containing 200 grams of metallic iron per litrs is employed and 400c.c. of this solution are placed in the flask. The air is expelled by boiling and then 1Oc.c. of a solution containing 33 grams of pure sodium nitrate per litre are placed in the funnel tube, and allowed gradually to drop into the solution in the flask. And nitric oxide which is formed is collected in a graduated tube holding 1OOc.c. When the Bodic nitrate has nearly all passed into the flask, the tube is filled with HC1 of 20 per cent. strength. This is allowed to pass into the flask. The funnel is once more filled with HCl, and when this has also passed into- the flask, the gas tube is removed and put aside. Without interrupting the boiling, 1Oc.c.of the solution to be tested are now poured into the funnel. This solution must be of such a strength that 1Oc.c. will evolve between 50 and 1OOc.c. of gas. Tho operation is conducted as before. The final rinsing with HCl leaves the apparatus ready for another estimation, andin this way six or seven estimations can be made before the ferrous salt is used up. It is advisable, finally, to repeat the analysis on the pure saltpetre. The tubes containing the nitric oxide are now adjusted in water in the usual way, and the volumes read off. As they are all under the same temperature and pressure no correction is necessary. Supposing the pure saltpetre gives 9Oc,c, of NO, then we know that 9Oc.c. represent -33 grammes of pure NaNO, and 1c.c. = 90366 grammes of NaNO, = *000604 N. The cdculation [is very simple.