152 THE ANALYST. A R.\PID ACCURATE METHOD FOR THE VOLUMETRIC ESTIMATIOK OF CARBON DIOXIDE. BY THOMAS MACARA. THIS process is based on the fact that BaCO, behaves like an alkali towards methyl orange. The apparatus employed consists essentially of : 1. A decomposing flask, A (see sketch), fitted with a stoppered funnel, C, and an exit tube. 2. An absorbing flaek, B, fitted with an inlet tube, D, and a safety funnel, E, which also acts as an exit tube. Tlw necessary solutious are : ((7) A cold saturated solution of Ua(HO),. ( b ) or i\ HC1. (c) Dilute HC1 (1 : 5 ) . ( d ) Indicators, phenolphthalein and methyl orange. Method of I-roccditre.-A known quantity (equivalent to not more than 2.5 grammes CaCO,) of the sample ia introduced into A, and H,O, free from COB, added until only sufficient space is left to prevent the contents boiling over after the addition of the acid.(If the air-space left in the flask is considerable, much longer boiling is required to drive out the last traces of CO,.) A considerable excess of Ba(HO), solution is placed in flask B, and to this 2 or 3 drops of phenolphthalein arc3 added, a little of the mixture being used to seal the safety funnel 13.THE ANALYST. 153 A and B are now connected, and an excess of acid (either HCl or H2S04) is cautiously run into A through C, and the contents of A gradually raised to boiling- point and kept at that temperature until nothing but steam appears to be passing into B. During the evolution of gas B must be kept vigorously shaken in order to assist the absorption of the gas.If the shaking has been sufficiently vigorous all the CO, will be found in B, none or only traces being found in E. As soon as all the CO, has passed over, the flasks are disconnected. An ordinary small funnel is now attached to inlet tube D, and dilute HC1 run into B until all the free Ba(HO), has been neu- tralized. Two or 3 C.C. of Ba(HO), solution are then added to make the contents of B alkaline, and the flask is well shaken to insure the absorption of any liberated GO,. The cork is removed, and the contents of E are rinsed into B, while any BaCO, adhering to the tubes is rubbed off and added to the bulk. The excess of the hydrate is now neutralized with 4 HC1 (-& where special accuracy is required), and 2 or 3 drops of methylorange added.The BaCO, is then titrated with standard acid. If, as sometimes happens when the flask has not been sufficiently shaken during n F absorption, a few scales of BaCO, remain undissolved, the contents of the flask are heated and the titration finished by adding the acid drop by drop till the solution remains pink after boiling for a short time. The usual precautions are necessary when H,S, SO,, Ci, Br, and oxides of nitrogen are present--e.g., CuSO, should be added to the flask before the acid when sulphides, chromate when sulphites are present. In the latter case H,SO, should, of course, be used, as HC1 is apt to give off Cl when boiled even with moderately dilute solutions of chromate, etc. As to the accuracy of the process, this is only limited by the delicacy of the indicators, and even this can in most cases be increased considerably by using larger quantities of the sample. In this way the merest traces of GO,-e.g., as in such substances as borax, calcined magnesia, and lime, eto.-can be accurately determined. It is only necessary to use a Iarge decomposing flask, with, say, 100 grarnmes of the sample, and a, small absorption flask, with only a small quantity of Ba(HO),, and using Fg acid for the titration. The whole process need not take more than fifteen to twenty minutes once the apparatus has been fitted up. A saturated solution of Ba(HO), keeps for a very considerable time in a stoppered Winchester quart bottle, no extra precautions being necessary ; any GO, absorbed falls to the bottom, and the clear Ba(HO), may be readily poured or drawn off.