The phenomena produced by the pencil of rays passing through the system are viewed by a strongly magnifying cylindrical eyepiece, 7, because the dark interferometer bands, being only a few hundredths of a millimeter apart, could not otherwise be distinguished. The band image of the upper half of the rays remains constant during all the operations. If the two chambers e and ƒ are filled with the same gas at the same temperature and pressure, and the compensator plates are perfectly parallel, the interference phenomena produced can not be distinguished from those produced by the upper half of the rays. If, however, pure dry air freed from carbon dioxide is introduced into chamber f, and pure dry air containing carbon dioxide is placed in chamber e, the lower band image is shifted laterally. Manipulation of screw k brings the bands back to their original position, the zero position on the screw. The difference of the readings on the graduated screw before and after the gas mixture is introduced is a measure of the refractive difference of the two gases. It must be remembered that the gases in the two chambers must have the same temperature and pressure and must be dry.

CALIBRATION OF INSTRUMENT

The authors have calibrated the interferometer by comparing the indications of this instrument with determinations by an exact analytical method. When one constituent of a gas mixture is determined the absolute refraction coefficients should differ by 0.0001 to possess an accuracy of about 0.02 per cent. The greater the difference of absolute refraction coefficients the greater is the accuracy of the method. Consequently, if carbon monoxide, nitrogen, or oxygen dilutes the air, the interferometer does not give accurate results, because the absolute refraction coefficients of those gases differ only in the fifth decimal place. If air always had to be used as a gas of comparison a serious limitation would be placed upon the interferometer. Fortunately this is not so. Pure air is used only when a constituent that contaminates pure air is to be determined. The general rule can be stated that the standard gas, or that used in the chamber as the gas of comparison, is the same as that in the chamber containing the gas to be tested, except that the constituent to be determined is removed from the standard gas by some suitable means. The difference in refraction is then due solely to the constituent the determination of which is sought.

RESULTS FROM USE OF THE INSTRUMENT

The authors have utilized the interferometer in analyzing mixtures of air and carbon dioxide, air and methane, and air and Pitts

burgh natural gas; and they have also determined the effect of lowering the oxygen content in mixtures of air and carbon dioxide and of air and methane.

2

Results of the analysis of mixtures of CO2 and air by means of the interferometer and by gas analysis (Pl. III) are shown below. With the eudiometer the results are accurate to within 0.02 per cent.

Results of analyses of mixtures of carbon dioxide and air

• Difference between zero and final readings in drum divisions on the micrometer screw.

The results of analyses of mixtures of methane and air follow. Such proportions of methane as frequently occur in mine air were used. All eudiometric analyses were made by means of the modified Haldane gas-analysis apparatus shown.

In making tests with the interferometer, the standard gas was passed into the comparison chamber at atmospheric temperature and pressure. The gas to be tested was then passed into the test chamber. When readings were taken the flow of gas to be tested was interrupted, and its pressure brought to atmospheric by opening one of the rubber connecting tubes to the air. When readings were promptly taken, diffusion of outside air back into the gas chambers was not appreciable. Soda lime, calcium chloride, and sulphuric acid were used where needed to purify the gases.

Results of analyses of mixtures of methane and air

Difference between zero and final readings in drum divisions on the micrometer screw.

EFFECT OF LOWERING OXYGEN CONTENT OF MIXTURES ANALYZED

As oxygen and nitrogen are not present in some mixtures in the proportions found in atmospheric air, some experiments were made to ascertain the effect due to less oxygen and more nitrogen than are present in air. The results are tabulated below. Of the two tabulations, the first shows the effect of decreasing the oxygen content of mixtures of CO2 and air and the second the effect of decreasing the oxygen content of mixtures of methane and air.

2

Effect of lowering the oxygen content of mixtures of carbon dioxide and air

Therefore, for every 1 per cent decrease in oxygen and a corresponding increase in nitrogen, the CO2 determined by the interferometer will appear about 0.15 per cent too high if normal air is used as the gas of comparison.

Effect of lowering the oxygen content of mixtures of methane and air

On the basis of the results presented above the following statements may be made: For every 1 per cent decrease in oxygen and a

corresponding increase in nitrogen the proportion of CO2 determined by the interferometer will appear about 0.15 too high if normal air is used as the gas of comparison, and for every 1 per cent decrease in oxygen and a corresponding increase of nitrogen the proportion of methane as indicated by the interferometer will appear about 0.16 per cent too high if normal air is used as the gas of comparison.

For application of the interferometer to other gases not mentioned is given a table of values of absolute indexes of refraction of the commonly occurring gases: 41

Values of absolute indexes of refraction (0° C. and 760 mm. Hg pressure)

2

The interferometer can be successfully used for technical gas analysis, especially for these tests and determinations: Tests of mine air for CO, and CH,; tests of the purity of manufactured hydrogen, chlorine, and other gases; determination of CO, in flue gases and breathing air; determination of sulphur dioxide in gases in connection with sulphuric acid manufacture; and others. Haber 42 has used the interferometer in determining the ammonia content of gases in the production of synthetic ammonia.

Mohr 43 has used the interferometer for analyzing flue gases. In Mohr's method the gas, after going through a drying tube, was passed through the interferometer and through a tube containing soda lime to absorb the CO2, then the gas was again passed into the interferometer. After the absorption of the CO2 the reading, of course, gave a different value than that obtained before absorption.

The difference between the two values when referred to prior calibration for CO2 gave the percentage content of CO2. The flue gas freed from CO2 (no account being taken of the small quantities of

2

41 Seibert, F. M., and Harpster, W. C., Use of the interferometer in gas analysis: Tech. Paper 185, Bureau of Mines, 1918, 18 pp.

42 Haber, F., and Le Rossignol, R., The production of synthetic ammonia: Jour. Ind. Eng. Chem., vol. 5, 1913, pp. 328-330.

43 3 Mohr, O., Die Verwendung des zeisschen Interferometers zur technische Rauchgasanalyse: Ztsch. angew. Chem., Jahrg. 25, June 24, 1912, pp. 1313-1317.

H2, CH4, and CO) does not have the same refractive exponent; consequently, Mohr calibrated the instrument with varying proportions of carbon dioxide, nitrogen, and oxygen. The oxygen content was estimated from a table prepared from the indications of the interferometer but as the sum (CO2+02) is not 20.9 per cent, and varies considerably with the fuel used, the results become only approximate, as to insure accuracy too many varying factors must be known.

The instrument can be used to determine methane in mine air if a suitable gas of comparison is employed; in other words, for comparison, the same mine air must be used as that whose methane content is to be determined, except that the methane in the gas (air) used for comparison must be removed by burning it and absorbing the carbon dioxide produced. However, if the mine air does not differ appreciably from atmospheric air, the latter can be used in the comparison chamber, but if the oxygen content of the mine air is much lower than that of atmospheric air, the proportion of methane will appear too great, as will be seen from the results in the table headed Effect of lowering oxygen content of mixture of methane and air.” As the oxygen content of mine air is seldom much lower than that of atmospheric air, the interferometer becomes generally applicable to the determination of CH, and CO2 in mine air.

66

Küppers has used the portable interferometer to determine methane in normal mine air, and his results agree rather closely with those obtained by ordinary analytical methods.

APPARATUS FOR ANALYZING NATURAL GAS

With a view to gathering data on the composition and fuel value of natural gas, the Bureau of Mines has been collecting samples of gas from different gas and petroleum fields. The natural gas piped to Pittsburgh is used at the Pittsburgh experiment station in making tests of explosives, flame safety lamps, electric mine motors, and so on, consequently the bureau has studied the composition of this gas to determine its explosive limits.45 Analyses of natural gas from the southern California oil fields have also been reported.46 Additional samples have been collected in West Virginia, Alabama, Nevada, Louisiana, Utah, Washington, Oklahoma, Kansas, Pennsylvania, Tennessee, Ohio, Kentucky, and Oregon.

44 Küppers, A., Die Bestimmung des Methangehaltes der Wetterproben mit Hilfe des tragbaren Interferometers: Glückauf, Jahrg. 49, Jan. 11, 1913, pp. 47-50.

45 Hall, Clarence, Snelling, W. O., and Howell, S. P., Investigations of explosives used in coal mines, with a chapter on the natural gas used at Pittsburgh, by G. A. Burrell, and an introduction by Charles E. Munroe: Bull. 15, Bureau of Mines, 1912, pp. 63–77. 46 Allen, I. C., and Jacobs, W. A., Physical and chemical properties of the petroleums of the San Joaquin Valley, Calif., with a chapter on analyses of natural gas from the southern California oil fields, by G. A. Burrell: Bull. 19, Bureau of Mines, 1911, p. 56.