As regards the subject of “air” in general, although a large number of isolated observations are recorded in somewhat tedious detail, no new fact of first-rate importance is apparent. The experiments are largely supplementary to those in the preceding volumes and are for the most explanatory or corroborative of them. Perhaps the most important are those dealing with “the production of nitrous air in which a candle will burn,” by which is signified the gas we now know as nitrous oxide, but which Priestley eventually termed _dephlogisticated nitrous air_. The process he employed is no longer used in the production of this gas, but it sufficed in his hands to determine its individuality without doubt. Priestley’s methods of experiment with his various “airs” were very uniform. He tried their solubility in water, their power of supporting or extinguishing flame, whether they were respirable, how they behaved with acid and alkaline air, and with nitric oxide and inflammable air, and lastly how they were affected by the electric spark. He occasionally made attempts to weigh them, but his determinations of their relative density were altogether untrustworthy. Indeed, it is evident from the terms in which he speaks of these efforts that he was conscious of their inadequacy. The result of submitting alkaline air (ammonia) to the electric spark, whereby it is resolved into nitrogen and hydrogen, surprised him not a little. “There are few experiments the _rationale_ of which I less pretend to understand than the production of genuine and permanent inflammable air from alkaline air by means of the electric spark.... One query on this subject is, whence comes the phlogiston, which is certainly a principal ingredient in the constitution of inflammable air. Alkaline air, indeed, contains phlogiston, because in the manner in which I have generally produced it, it is itself partially inflammable; but it is not nearly so much so as the inflammable air which is produced by means of it. Besides, it will appear by the following experiments that the quantity of the inflammable air far exceeds that of the alkaline.” Although Priestley clearly recognised the production of the inflammable air, “in no respect to be distinguished from that which is extracted from metals by acids,” and inferred it must come from the alkaline air (“the production having its limits”), he failed to detect the other constituent of ammonia. His determination of the actual increase in volume was inaccurate, and his attempt to explain the phenomenon wholly fallacious. At the instigation of Mr Woulfe, whose name mainly lives in connection with a useful piece of chemical apparatus, Priestley was encouraged to hope that he would