According to Liebig, it is the atmospheric ammonia which furnishes the nitrogen of plants. The traces of nitric acid which have been detected in the air, are referable to the oxidation of the constituents of ammonia ;

and not to the direct union of the oxygen and free nitrogen of the atmosphere. Dr. Pereira states that a carbonaceous (sooty) substance, and traces of sulphates, chlorides, and calcareous matter, are the usual impurities of the first rain water of a

shower. Zimmerman found oxide of iron and chloride potassium in rain water other chemists have been able to deteet no iron in it, but have found meteoric iron

and nickel in dew. Brande detected in it, chloride of sodium, chloride of magnesium, sulphate and carbonate of magnesium, sulphate of lime, and oxide of manganese. The putrefaction to which rain water is subject, shows that some organic matter is present. The term pyrrhin (from nvqqog red) has been applied by Zimmerman to an atmospheric organic substance which reddens solutions of silver. Whenever rain water is collected near large towns, it should be boiled and strained before use, as it contains less saline impregnation than other kinds of natural waters,

it is more apt to become contaminated with lead from roofs, gutters, cisterns, and water pipes. To purify rain water and render it useful, for the delicate purposes of chemical experiment, Morveau recommends dropping into it a little barytic water

and then exposing it for some time to the atmospheric air. This combines with the carbonic acid, which being the solvent of the carbonate of lime, both it and the carbonate of baryta are precipitated as insoluble salts. Instead of exposing it to the