detecting the occurrence of dangerous gases and assuring adequate ventilation and for ascertaining the composition of atmospheres as a basis for planning safe procedures when conditions are dangerous, as when extinguishing mine fires by sealing and when conducting rescue and recovery work following explosions. Although approved portable detecting devices, such as the flame safety lamp, methane detectors, and carbon monoxide detectors, are widely used for examining mine atmospheres, all have a definite lower and upper limit to their range of detection. Moreover, they are designed primarily for the detection of a single gas constituent, whereas in many cases a complete analysis or determination of all the constituents of the atmosphere is necessary.

2

Where the information desired is outside the scope of portable detecting apparatus samples of the atmosphere are collected and later subjected to chemical analysis, usually by gas-volumetric methods. The apparatus for conducting such analyses may be of a portable type which can be taken to a place near the site of collection of the samples; or, if analyzing apparatus is not available, it is usually possible to collect the samples in bottles or tubes that can be tightly scaled and then shipped to a laboratory for analysis. The bureau maintains a laboratory at its Pittsburgh Experiment Station for analyzing samples collected and submitted by its field engineers. Also, in emergencies where there is mutual interest, this service is extended to the State departments of mines, mine operators, and others.

ACCURATE ANALYSES VERSUS APPROXIMATE ANALYSES AND QUICK RESULTS

When a high degree of accuracy is desired and there is no immediate need of results the sample should be sent to a well-equipped laboratory where the conditions under which the analytical apparatus is used permit more refinement in design and consequently greater accuracy than can be conveniently embodied in a portable apparatus. However, when moderate accuracy is satisfactory and compatible with the problem at hand the use of portable apparatus is advised, as it not only avoids delay in obtaining results but also avoids the loss of samples or contamination by leakage before analysis is made. A central laboratory often has a large number of samples on hand awaiting analysis, and delays are unavoidable.

2 Frequently the gas laboratory of the Pittsburgh Experiment Station receives samples in the ordinary vacuum-tube containers with the request that they be analyzed for sulphur dioxide, hydrogen sulphide, or oxides of nitrogen. With the low concentrations of these gases usually found it is not possible to make a satisfactory determination with the sample of ordinary size. Also, these gases react with dirt, moisture, etc., in the container or decompose so that by the time the sample is analyzed they have disappeared from the gas mixture. These determinations should be made in the field by direct sampling into an absorbing reagent.

The need of immediate results is very obvious in mine fires or in recovery work after explosions when the results of analysis must be known within a day or even an hour after the sample is collected, as the next move may depend entirely on the composition of the air. This need for immediate results therefore necessitates the use of a portable type of gas-analysis apparatus that can be transported to the scene of the investigation; this in turn creates the need for personnel familiar with the technique involved in collecting and analyzing mine atmospheres. The following instructions have been prepared with this in view.

The type of apparatus to which these instructions particularly apply is a portable Orsat modified to suit the needs of field use. It is not designed for the precise analysis of mine air, such as is required for most ventilation studies, but is only intended for the analysis of the samples usually taken in emergency cases (fires and recovery work) as a guide to safe and effective procedure. It can, however, be used satisfactorily to check the occurrence of methane in headings and rooms, to determine oxygen depletion and black damp, and for other uses described later. It is not accurate enough for the analysis of return air of coal mines to ascertain the quantity of methane or to determine the presence of carbon monoxide from the standpoint of health or the extinguishing of mine fires.

The apparatus and its operation are described in detail, much of which may seem unimportant to persons acquainted with the technique of gas analysis. However, it has been primarily prepared to instruct engineers, chemists, and mining men who are unacquainted with the technique and who infrequently have occasion to make analyses. On this basis the detail is thought justified. Also, the authors believed that rather than list a number of procedures specific to each of the wide variety of samples for which the apparatus might be used confusion would be avoided by giving a general procedure that would take care of the majority of cases and difficulties presented in the field. For some samples the operations and precautions as given in the general procedure could be omitted without affecting the results, and as the persons making analyses become more familiar with the technique of gas analysis the omissions and short cuts will become. apparent.

Although the chemistry and procedure involved in gas analysis are simple, practice and experience are required to attain a technique that will give satisfactory results and avoid some of the difficulties encountered, especially when dealing with a wide variety of samples. In view of this it is not merely necessary for beginners to read these instructions; they are also advised to obtain an apparatus and actually carry out the various operations while reading the text.

SAMPLING MINE GASES

The field investigator may be the collector of the samples as well as the analyst; therefore, a section of these instructions has been devoted to a description of suitable procedure for taking samples. In describing these it will not be possible to consider all of the varied situations that might arise, but the suggestions and information given may be helpful to engineers and chemists who are inexperienced in mine-gas work.

In determining the composition of mine gases by analysis the collection of a representative sample and the preservation of that sample until it can be analyzed are the most important and most difficult parts of the procedure. Analytical work has been standardized for the most part, and it is fairly well known that by following a certain sequence of manipulations and measurements carefully and precisely a result within certain limits of error will be obtained. On the other hand, the collection of the samples requires planning and ingenuity, in many cases, to insure their being representative of conditions. Improper and nonrepresentative sampling gives erroneous or irrelevant results. Erroneous results lead to incorrect interpretations or conclusions that may cause either a false sense of security or undue anxiety, and they may be used as a basis for unsafe or unnecessary recommendations.

Many suggestions for sampling have been devised, and some of those which have been found most practical and efficient by the bureau have been described in a previous publication3 dealing with the analysis of mine gases.

However, it was not possible to give in Bulletin 197 many of the apparently minor but essential details which have been found by experience to be necessary in collecting samples. In view of this the following instructions, which deal almost entirely with a specific field of work, have been prepared to supplement the information given in that bulletin.

TYPES OF GAS SAMPLES

Gas samples may be divided into two general classes, which characterize the rate at which the samples are taken. These classes are termed "grab" (sometimes called "spot," "snap," or "instantaneous") samples and "continuous" samples.

Grab samples are taken over a short duration of time, say, a few seconds to a minute or two, and represent the gas at some particular place and instant. These constitute the large majority of samples. taken in mines.

Burrell, G. A., and Seibert, F. M., The Sampling and Examination of Mine Gases and Natural Gas: Bull. 197, Bureau of Mines, revised in 1926 by G, W, Jones, 108 pp.

Continuous samples are collected over a period of longer duration than grab samples and are usually taken where the stream of gas to be sampled is of varying composition; that is, homogeneous at a particular instant and varying throughout a comparatively long period, but not varying as regards stratification in a cross section. A sample taken continuously throughout the period will give the average composition. There is seldom occasion for taking continuous samples in mines, even when information regarding the average condition is desired; a number of grab samples taken at predetermined intervals will generally be more informative than one continuous sample. The principle of slow sampling in taking continuous samples, however, may be applied to the collection of gas from feeders or other similar places where the supply of gas is not large enough to fill the container instantaneously.

A grab sample may be taken from one point in a cross section of the mine, or the opening to the container may be moved during the sampling so that gas from several points may enter and thus form a composite sample representing the cross section. It is often thought that the latter method may be used to obtain an average sample from a stratified atmosphere. If the air is moving and only slight stratification is anticipated, it is good practice to move the opening of the tube vertically across the place of sampling when the gas is entering the tube. In still air, however, where marked stratification may occur, it is not possible to obtain an average sample by this method unless special facilities are used to control the rate of movement of the opening and the rate of entrance of the gas. The use of such facilities is not ordinarily practicable for routine sampling in mines; also, a number of grab samples taken from various points in the cross section will give more information.

SAMPLING EQUIPMENT

TYPES OF SAMPLE CONTAINERS

The Bureau of Mines recommends the use of the vacuum-tube type of sample container shown in Figure 1 for all samples of mine air taken for shipment to its gas laboratory at the Pittsburgh Experiment Station. This type of container has been found most suitable for effective sealing by field men, durability for shipping, and preserving the samples where considerable delay may occur between the sampling and the analysis, as in shipment to a distant laboratory. In field work where analyses are made within a relatively short period after sampling and it is not necessary to make shipment other types of containers can be used. They are often more available than vacuum tubes, are refillable, and may be used many times, which makes their use less expensive than vacuum tubes unless a laboratory

or glass-blowing shop equipped to make, repair, and reevacuate the latter is available.

VACUUM TUBES

The vacuum-tube gas-sample containers shown in Figure 1 and used by the bureau for sampling mine gases are glass bulbs from which

99.97 per cent or more of the air has been removed by a vacuum pump and the necks of which have been hermetically sealed by heating and drawing during the final stage of evacuation. This results in an inclosed space which has practically no internal pressure but which