are adjustable. The pipettes are easily removed for cleaning and refilling.

7. Assembling of the apparatus in a common train to avoid entirely making and breaking connections, which is necessary when pipettes are connected in turn to the burette for absorptions and to the combustion pipette for burning.

8. A second three-way stopcock, c, sealed to the burette, so that a gas sample can be taken into the burette at the top without disconnecting the burette from the train of pipettes.

The burette of the apparatus is graduated in tenths of a cubic centimeter from the upper stopcock down to the 100-c. c. mark, so that the quantity of gas measured is the quantity of gas analyzed.

REASONS FOR DISCARDING THE EXPLOSION PIPETTE

The explosion method of determining the combustible constituents was discarded because of the following objections:

Many gas mixtures analyzed at the bureau's laboratory do not contain explosive gases in large enough quantity to produce an explosive mixture with air; consequently, if the explosion method is used, hydrogen or electrolytic gas (hydrogen and oxygen made by the electrolysis of water) must be added to make the mixture explosive. This procedure introduces an additional factor in the preparation and measurement of the exciting gas. Furthermore, when the explosion method is employed small quantities of gas mixtures that contain a large proportion of methane or of the higher-paraffin hydrocarbons have to be used. This requirement applies especially to natural gas, in which the paraffin hydrocarbons predominate; to many special mixtures prepared in the laboratory; and also to mine atmospheres containing a large proportion of methane. For instance, when methane-air mixtures are ignited by an electric spark they have explosive limits approximately as follows: Low, 5.5 per cent methane; high, 14.5 per cent methane; and most violent, 9.5 per cent methane. To remain within the explosive limits of the mixtures and to avoid too violent an explosion, only a small quantity of gas can be taken for an explosion, and any error of manipulation or error inherent to any type of apparatus is multiplied 8 or 10 times when the calculations of results is made to a percentage basis.

The use of the explosion method also demands exceptionally tight connections on the explosion pipette to resist the pressure of the explosion. With the slow-combustion pipette there need not be heavy pressure at any time on connections and faulty analyses caused by loss of the sample through leaks are largely avoided.

n

a

DESCRIPTION OF APPARATUS

Many modifications of the original Orsat apparatus for gas analysis are used at present; each has particular merits and is adapted for some special work. The number and kind of pipettes used depend upon the composition of gas to be analyzed. If ethane is to be determined, a copper oxide tube is required (fig. 10) and

FIGURE 11.-Orsat apparatus for complete analysis of gases which do not contain ethane or other higher saturated hydrocarbons. For explanation, see text

the cuprous chloride pipettes are omitted, as carbon monoxide is determined by combustion along with the hydrogen. If ethane is absent, as in producer and blast-furnace gases, two cuprous chloride pipettes may replace the copper oxide tube (fig. 11), and the methane and hydrogen are determined from the combustion data by

L

means of the formulas given under "calculations for determining combustibles," page 26.

The Orsat apparatus 21 is used in the Bureau of Mines gas laboratory at Pittsburgh for complete and partial analyses of gases. A complete analysis of carbon dioxide, illuminants, oxygen, carbon monoxide, hydrogen, methane, and ethane can be made without connecting or disconnecting any parts. Because all of the constituents are removed before the paraffin hydrocarbons are burned, larger amounts of these constituents may be taken for combustion with oxygen and more accurate determinations of methane and ethane may be made. The same argument holds for hydrogen and carbon monoxide, which are determined by fractional combustion with copper oxide at 300° C. Mercury is used in the burette as the confining liquid.

The apparatus arranged for complete gas analysis is shown in Plate V and Figure 10. It consists essentially of a 100-c. c. waterjacketed burette, g, with a Pettersson compensating tube, h, for correcting for changes of temperature and pressure during the analysis; four pipettes, p, q, r, t, and a copper oxide tube, i, with an electric heater, 7, for the fractional combustion of hydrogen and carbon monoxide. The manometer on the compensator is filled with mercury or a 1 to 3 glycerin-water solution to the mark f.

BURETTE

Figure 12 gives details of the burette, which is graduated to 0.1 c. c., every tenth mark extending half-way around the burette. The even cubic centimeters are etched on the burette and the 100-c. c. mark is placed at the bottom. In preparing the apparatus for use. proceed as follows: Fill the simple absorption pipettes p, q, and r with pieces of hollow glass tubes as usual. Pipette p contains caustic solution for the removal of carbon dioxide; q contains fuming sulphuric acid for removal of illuminants; and r contains alkaline pyrogallate solution for oxygen. Place enough reagent in the pipettes so that when they are drawn to the marks o, o', and o", enough reagent remains in the rear branch of the pipette to form a seal. Attach rubber bags to the rear branches of pipettes q and r, to prevent access of air to the reagents. Inflate each bag slightly before attaching it to the pipette.

21 Burrell, G. A., and Oberfell, G. G., The use of copper oxide for fractionation, combustion of hydrogen, and carbon monoxide in gas mixtures: Jour. Ind. Eng. Chem., col. 8, 1916, p. 228. Jones, G. W., and Neumeister, F. R., An improved Orsat apparatus for gas analysis: Chem. and Met. Eng., vol. 21, 1919, p. 734.