228 THE ANALYST. ON TEE DETERMINATION OF CARBONIC ACID I N CARBONATES. By G. W. WIGNER, F.C.S. Read before the Society of Public Analysts, on 19th Noumber, 1879. THE results obtained by the determination of carbonic acid in earthy, metallic, or other insoluble carbonates by decomposition with dilute acid in any of the ordinary forms of carbonic acid apparatus, are not of a satisfactory character.There is great difficulty in ensuring the complete decomposition of the carbonates, and almost equal difficulty in driviug off the last traces of carbonic acid gas without incurring a loss of the acid used for decomposition, and thereby an apparent increase in the proportion of carbonic acid present. Even in the case of the alkaline carbonates, the results obtained in this way are less accurate than could be wished ; but when carbonate of lead, or carbonate of baryta, or, still worse, mixtures of the two are treated, it will be found that the results obtained in the ordinary way will not only be somewhat irregular, but will, in most cases, show an error equal to as much as 2 or 3 per cent.of the total carbonic acid present. In the case of white lead, where the proportion of carbonic acid present is really the standard by which to judge of its suitability for use as a paint, an error of this kind is a very serious one.The only satisfactory process by which carbonic acid can be estimated in such samples as these, is by measuring the volume of the gas evolved on treating the sample with dilute acid, and ascertaining that this gas is entirely carbonic acid.For some years I have used a process of this kind, and after making some hundreds of determinations by it, I am convinced that, with proper care, the error in the determinations of carbonic acid need in no case exceed *lo or 005 of the total amount present. This greater accuracy is moreover attained without the use of any complicated apparatus, excepting a suitable gas measuring apparatus.Its attainment depends entirely on carrying out the decomposition of the carbonates in a, paxtiztl vacuum, so that the liberation of the carbonic acid proceeds rapidly and freely at a temperature considerably below the ordinary boiling point of the solution. By this means the time necessary for the decomposition is greatly shortened, and the risk of the evolution of any other gases than carbonic acid is also decreased.The apparatus required is as follows :-A test tube about 8 inches by 1, fitted with a good india-rubber cork having two holes through i t ; one hole is provided with a thistle funnel with stopcock, and the other hole with a glass tube of small bore also furnished with a stopcock. This test tube constitutes the decomposition apparaturr ; the thistle funnel is for the introduction of the acid by which the decomposition is effected, and the small tube for carrying the evolved gases into the measuring ctppa;raltua.THE ANALYST.229 In addition, I use a small test tube about 1* in. long and 2 in. diameter to contain the sample to be analysed. The following is the mode in which I treat the sample :-A small portion, generally in the case of white lead about 30 or 40 grains, is weighed into the small test tube; B small quantity of hot recently boiled distilled water is put into the larger test tube, and the small tube dropped into it.The quantity of water should not be more than sufficient to fill one inch of the larger tube. The decomposition tube is then securely corked, the stopcock leading to the gas apparatus closed, and the stopcock on the thistle €unnel opened.The water in the decomposition tube is now rapidly and strongly boiled; the steam, which is liberated, expels all air from the tube through the thistle funnel. After the boiling has continued for a few seconds, the thistle funnel is filled with distilled water, recently boiled but cold ; this will prove at once whether the whole of the air has been expelled, since in that case bubbles will no longer pass through the water in the funnel.The lamp having been removed, the stopcock on the thistle funnel is closed as rapidly as possible, and then the small amount of water which remains in it allowed to pass into the decomposition tube by cautiously opening the stopcock.By this means the air has been entirely driven out from the apparatus, which is ready for the decomposition of the carbonate in a vacuum. The delivery tube has, of course, been coupled up to the gas apparatus previously, and the stopcock in that being opened, the mercury in the measuring tube is allowed to fall, and so a vacuum prepared in the measuring tube for the reception of the gas as fast as it is liberated.I prefer to decompose the carbonates with hot dilute nitric or hydrochloric acid; this acid should have been boiled for half-an-hour or more, to expel all traces of dis- solved gases, and must then be allowed to pass in cautiously in small quantities at a time from the thistle funnel into the tube. The gas, as fast as it is evolved, will pass over into the measuring tube, and in from l a to 2 minutes decomposition will be completed.The decomposition tube is then filled to within one inch of the top with hot freshly boiled distilled water, and the contents of the tube boiled in the partial vacuum which still exists. By this means the last traces of gas are driven out and collect above the surface of the liquor.This remaining space is then filled up by allowing boiling distilled water to pass in through the thistle funnel until the tube is absolutely full to the stopcock. Thus the apparatus was, when the analysis was started, devoid of air, and full only of watery vapour ; and, when the analysis is complete, it contains only water and the solution of the substance also free from air.Some little difficulty will be experienced at first in avoiding the adding of the acid too rapidly, as the carbonic acid is liberated somewhat violently in the vacuum, but beyond this there will be no practical difficulty in working the process. As an illustration of the accuracy which may be attained with reasonable care, I may mention that 11 consecutive analyses of calc spar gave a minimum result of 43.96 per cent. and a maximum of 43.99 per cent., and that the 11 determinations were completed within 2+ hours. In the case of white lead and carbonate of baryta, the determinations generally agree within 003 per cent.230 THE ANALYST. I need not say anything as to the measuring of the gas, since, of course, the ordinary corrections for temperature, barometric pressure, and watery vapour, must be made in every case.