“The carbon dioxide must have come from one of two sources, either from some carbonaceous substance contained in our phosphorus, or as the result of the action of ozone on the cork stoppers used to make connections. The use of rubber was avoided as far as possible, and every precaution was taken as in the earlier experiments on the carbon monoxide and ozone. It did not appear improbable therefore that the difficulty arose from the use of impure phosphorus. Phosphorus was, therefore, obtained from as many different sources as possible, and with each of these the above described experiment was repeated, using the same apparatus. In every case the precipitate of barium carbonate was obtained and as far as could be estimated in about the same quantity. Attempts were then made to purify the phosphorus. One specimen was placed in hot water under the receiver of an air pump and the air exhausted, for the purpose of recovering any gases which might be contained in the phosphorus. Other specimens were distilled in an atmosphere of pure hydrogen and the vapor condensed in cold water. No matter what process of purification had been adopted the phosphorus acted in the same way afterwards as before.” “We then constructed an apparatus in which the gases could at no point come in contact with cork stoppers or rubber joints. This consisted of a flask of from three to four litres capacity, provided with a doubly perforated cork stopper. Through this there passed one glass tube reaching to the bottom of the flask, and another reaching only half way. Outside the flask the shorter tube was connected with the wash bottles used to purify the air from carbon dioxide, while the longer tube was bent twice at right angles, and passed through the stopper of a U tube about 8 in. high. In the flask there were placed two or three sticks of phosphorus, each three or four inches long, and enough distilled water to somewhat more than fill the neck when the flask was inverted. The U tubes were filled with moistened asbestos which had been previously ignited. There was then added some mercury, so that when the tubes were inverted in which position the entire apparatus was placed when in use the mercury covered the corks with a layer from three quarters to an inch in thickness.