Phenol-phthalein gives to alkaline solutions an intense crimson colour, but it is very sensitive to CO, and this property can be taken advantage of to make approximative determinations of carbonic acid in the air. The latest development of the method is that of Lunge and A. Zeckendorf.1 It requires simply a solution of well ignited sodic carbonate Na, CO3, 53 grms to the litre, coloured crimson by 02 grm. of phenol-phthalein and a suitable flask and an india-rubber pump delivering about 70 c.c. of air; the india-rubber pump must be provided with valves so as to admit the air in one direction and expel it in another. The

B

pump B is connected with the flask A as in the figure (Fig. 5), so that the air passes through the solution. To make a determination of CO, in the air, the pump is worked several times so as to fill the flask with the air to be examined, then 10 c.c. of the sodic carbonate solution are added and the pump slowly worked until the sodic carbonate solution is decolourized and the number of times noted. Very impure air will be decolourized by two complete emptyings of the india-rubber pump (30 per cent.); for moderately impure air 9-10 (09-10 per cent.) volumes will be required. The ordinary air of towns will take twenty-five volumes (052 per cent.) and the best country air forty volumes (038 per cent.). G. Lunge and Zeckendorf standardize their apparatus by actual experiment as against the Pettenkofer process and with a pump of 70 c.c. found the following percentages of CO, corresponded to the number of fillings of the pump. The air in the flask is not taken into account, for it is a constant quantity. If therefore the operator uses a rubber pump delivering 70 c.c. the table will be sufficiently exact, but with any other size, he must standardize it by actual experiment.

1 Zeitschr. f. angew Chemie, No. 14, 1888.

10 CC.

FIG. 5.

H

TABLE XVIII.

SHOWING THE PERCENTAGE OF CARBONIC ACID (PETTENKOFER) CORRESPONDING WITH EVERY FILLING OF THE BALL OF THE MINIMETRIC APPARATUS.

Mr. Wanklyn suggested imitating the cloud or turbidity which carbon dioxide produces in clear baryta water, by adding to another portion of baryta water, known quantities of CO, in the form of sodic carbonate. This is a simple method, which takes but little skill or practice.

The following apparatus is required :

Large clean stoppered bottles, such for instance as Winchester quarts, the capacity of which has been carefully determined by filling them with water and then measuring the water (163862 c.c. 1 cubic inch), a pair of bellows, several glass cylinders such as are in use for" nesslerising," a graduated burette with glass stopcock the graduations being in 10ths of a c.c.; a solution of carbonate 447 grams to the litre, 1 c.c. of this is equal to 1 c.c. of CO2 or by weight to 1.97 milligram; baryta water strength about 1 per cent. The air of the place to be tested having been blown in for a little while by the bellows, to replace the original air in the bottle the stopper may be inserted and the bottle conveyed to the laboratory or testing place, for it is seldom convenient to perform the testing in the room. To perform the actual estimation 100 c.c.

2

of the baryta water is poured into the bottle, the stopper inserted and the bottle shaken some minutes, the baryta now absorbs all CO2 which unites with the baryta forming an insoluble carbonate, and produces a greater or less turbidity according to the amount of CO, present; the baryta water thus made turbid is poured into a glass cylinder, and placed side by side with another cylinder containing 100 c.c. of clear baryta water; into this clear baryta water is dropped from a burette the solution of sodic carbonate until the two cylinders of liquid are equal in turbidity.

If the bottle contain 2,000 c.c. of air and the turbidity is imitated by using 2 c.c. of the sodic carbonate, this of course means that 2,000 volumes of air contain 2 c.c. of CO2-that is 1 per cent. If greater accuracy is required the air acted upon must be reduced to the standard pressure and temperature by calculation.

A similar method has been used by Dupré and Hake, the carbonic acid being absorbed by pure basic lead acetate solution, and the turbidity produced being imitated by running into a similar bulk of basic acetate solution a solution containing known quantities of CO..

2

The Author's method. The author prefers to absorb CO2 by a strong solution of caustic soda and then to estimate the CO2 absorbed either by weight or as a gas. Caustic soda is dissolved in water to something like 25 or 30 per cent. If the caustic soda is not quite pure, that is free from carbonate, most of the carbonate settles to the bottom, sodic carbonate being imperfectly soluble in strong soda lye, hence the soda must be allowed to settle before being used. 20 c.c. of this strong soda is then placed in the bottle containing the air to be tested, and having been allowed to act for some ten minutes or more, the 20 c.c. are transferred to a Schrotter's CO, apparatus (see Fig. 6) and the last traces of soda are washed out by means of some recently boiled distilled water. The flask is charged with strong sulphuric acid in a, to dry the issuing gas, and also by weaker acid in b, the whole is carefully weighed, and then the acid is allowed to fall drop by drop on to the alkali; when the reaction is finished the last traces of CO, are expelled by gently boiling the contents for about a minute. On cooling, the flask is reweighed--the loss represents the weight of CO, in the volume

a.

FIG. 6.

of air. Weight is translated into volume or vice versa by the aid of the following short table :

M

TABLE XIX.

SHOWING THE RELATION BETWEEN WEIGHT AND VOLUME OF CO2 IN

MILLIGRAMMES AND CUBIC CENTIMETRES.

It is necessary to observe that a determination is made once for all in the same manner of the amount of CO, in the soda solution and the number obtained is a constant correction; an example will make all clear.

20 grms of the soda solution put into the Schrotter were found after the addition of the acid and boiling up to have lost 12 milligrammes.

20 e.e. of the soda after acting on 2,500 e.c, of air were found to have lost 82 milligrammes, which subtracted from 12 gives 20 milligrammes derived from the air; 20 milligrammes are by the table equal to 1011 c.c. of carbonic acid, that is 2,500 volumes contain 10-11 of CO, or 4 per cent.

If it is preferred to measure the CO, as a gas, this may be done simply by putting a short test tubo charged with more than suflicient acid to noutralize the alkali, in the flask containing the soda; to the flask must be adapted by means of a perforated caoutchouc cork, a piece of glass tubing which in its turn is connected with a pioco of rubber tubing some 13 or 16 inches long and has a short angled bit of glass tubing terminating in an almost capillary orifice. All an is to be expelled from the apparatus by boiling the liquid until the steam rushes out with considerable force, the delivery tubo is then inserted under mercury in a eudiometer tube also moreury filled, the value of the divisions of

which are known in c.c. and the acid carefully spilt little by little into the soda by inclining the flask; the gas displaces the mercury in the tube, and the last traces are collected by boiling. The whole is allowed to cool and the gas measured after the usual corrections. In this case also a control experiment must be made on a known bulk of soda solution, and the amount of CO2 subtracted. This method is not really so accurate as that by weight, as determined in a Schrotter's apparatus, because of the solubility of CO, in the water condensed from the steam and floating on the top of the mercury in the eudiometer tube.

(62) Pettenkofer's Volumetric Method.

2

The method which is known as Pettenkofer's consists in ascertaining precisely the alkalinity of clear baryta water then absorbing the carbonic acid by the baryta water, allowing the insoluble carbonate to settle and again taking the alkalinity, which is of course proportionately less.

The best strength for the baryta water is 7 grms. of the crystallized hydrate to the litre; it may be stored in a bottle, stoppered by a doubly perforated cork, the one carrying a short tube loosely filled with pumice-stone moistened with potash; into the other perforation is fitted a long tube bent syphon wise, the one limb of the syphon reaching nearly to the bottom, the longer limb extending outside a few inches below the bottom and being fitted with a short piece of caoutchouc tube and clip, the bottle itself standing on a suitable shelf, and the syphon once filled and kept full the baryta can always be syphoned off clear, and since the air drawn in to replace the liquid drawn out is freed by the potash from CO2, the alkalinity of the liquid will not alter. first standardized by a solution of oxalic acid 2.8736 grms. to the litre (each c.c.1 milligramme CO2), the indicator may be either phenol-phthalein or methyl orange. The solution of oxalic acid is run from a burette into say 25 c.c. of the baryta water coloured by the indicator until by the change of colour it is seen that the exact point of neutralization is reached.

The baryta is

This number is to be taken as the standard, supposing for instance that 25 c.c. of clear baryta water took 30 c.c. of the acid solution and another 25 after acting on a given bulk of air and being