322 PROCEEDINGS OF THE AMERICAN ACADEMY. 



hydrochloric acid. Ten grams of the solid left no residue on evapora- 

 tion in a platinum dish. The bromine was prepared essentially accord- 

 ing to one of the methods of Stas, and was wholly free from all other 

 substances except water. 



In order to measure out successive, equal portions of bromine for 

 parallel experiments, an almost saturated solution was kept in a covered 

 bottle, and measured out in a glass-stoppered pipette filled by a siphon, 

 so arranged that the top of the pipette communicated with the bottle.* 

 The cork, the only rubber connection in the apparatus, was protected by 

 soft paraffin, which also served when mixed with a little melted rubber f 

 to lubricate the stopcocks. From this pipette successive portions, differ- 

 ing less than 0.1 per cent in bromine-contents from one another, could be 

 delivered at will. A diagram illustrates this description. 



The mode of procedure in the quantitative study of the bromine- 

 oxalic reaction was as follows : — 



Into a glass-stoppered bottle of about 65 c.c. capacity a quantity of a 

 solution of oxalic acid was introduced, and into this was run a pipetteful 

 of bromine water. The bottle had been previously adjusted by means 

 of glass pearls, so that its remaining volume was just enough to receive 

 the solutions. The stopper was immediately closed and the bottle was 

 immersed in a thermostat whose temperature was kept constant to within 

 one-tenth of a degree. In order to prevent any leakage, the stopper 

 was lubricated with the rubber-paraffin mixture, which serves to seal 

 hermetically the reacting liquids. The rubber-paraffin mixture was 

 not attacked to any appreciable extent. It is necessary to fill the 



* When the solution is delivered from the pipette, air must of course flow in to 

 take its place. If this entering air flows through a bulb-tube containing water and 

 a drop of bromine, the atmosphere of the bottle will remain saturated with bromine 

 without loss to the solution. If this bulb-tube is not provided, the air-ingress 

 must be as distant as possible from the air-tube of the pipette, winch should almost 

 touch the liquid in the bottle. This latter arrangement was found in our experi- 

 ments to answer the requirements of the present work, for a single series may be 

 made with sufficient accuracy. After much air has been admitted, the solution 

 will naturally be somewhat weakened. In each series careful determinations of 

 the strength of the solution were made before the experimenting-bottles were 

 filled ; again after filling three bottles, and so on until all were filled, when two 

 final determinations were made. 



t This nvxture, suggested to one of us by W. Ramsay, has proved itself useful 

 in many exigencies. It is made by melting pure rubber, adding equal weights of 

 hard and of soft paraffin, and evaporating off the more volatile constituents. By 

 varying the proportions any desired consistency may be obtained. It may be filtered 

 when hot through fine cloth to remove accidental infused impurities. 



