is the same as that for sampling moving air. However, when taking samples of these gases near the place of blasting and when there has not been enough time or cause for mixing the atmosphere may be markedly stratified or heterogeneous, and a representative or average sample will be difficult to obtain. The method of dividing a cross section into imaginary rectangles, described under Procedure for Collecting Gas Samples, is not adequate, because the contaminating gas may be in pockets and may be stratified in a given section; the former is especially true if there are a number of intermittent shots. A detailed description of procedures for obtaining average samples under such circumstances has been described recently in Bureau of Mines Bulletin 287.*

PROCEDURES FOR COLLECTING GAS SAMPLES

If the composition at localized places is desired, as is usually the case in sampling still or slowly moving air in mines, the procedure consists merely of holding the tip of the container at the desired place while filling. Care should be taken to avoid disturbance of the air and contamination during both sampling and sealing.

If an average analysis of an atmosphere that is not of uniform composition and is practically still air is desired, it is advisable to take several samples from a cross section. In doing this, divide the cross section as nearly as can be judged by the eye into equal rectangles, one of which represents each of the anticipated zones of variable composition, and take a sample of gas in the center of each. Usually four samples are enough to give an average of the conditions. In collecting these samples, take care to avoid contamination by the breath and to avoid undue disturbance of the air.

In collecting samples from blowers and feeders, from leaks in plugged and sealed gas wells, and from similar places where gas issues, a satisfactory procedure for obtaining a sample of highly concentrated gas (diluted as little as possible with air) is to place a tin or glass funnel, tin can, bucket, or small keg, with the bottom removed, over the place of emission and provide for collecting the gas from a hole in the top. The receptacle may be sealed with mud around the bottom. Before sampling, it should be purged by allowing concentrated gas to flow in and remove the original air. The time necessary for this operation can be judged from the rate of flow of gas and the size of the receptacle. The sample may be taken by any of the methods previously described. Where the flow of gas and the receptacle are small a type of container that can be slowly filled as gas is emitted should be used.

Sometimes, however, the volume of gas liberated from blowers, crevices, etc., is large enough or enough gas is already at hand to per

Gardner, E. D., Howell, S. P., and Jones, G. W., Gases from Blasting Tunnels and Metal-Mine Drifts: Bull. 287, Bureau of Mines, 1927, 96 pp.

mit direct sampling-that is, without the use of a funnel or collecting device by merely holding the tip in the stream and breaking it according to the manner described later for filling vacuum containers.

FILLING VACUUM TUBES

In taking samples where there is an appreciable air movement, stand facing the air and hold the bulb at arm's length (avoiding contamination of the sampling zone by the collector's breath) as near as possible to the point to be sampled. Break off the end of the capillary tube at the file mark or scratch. This may be done by holding the shoulder of the tube firmly in one hand and with the other attempting to bend the capillary tube at the mark. As the fractured edges are often sharp it is well to use a glove, piece of cloth, or handkerchief for protection of the thumb and fingers used to grip and bend the tip. A suitable breaking device is a 4 by 3 inch pipe nipple or a small piece of hardwood with a hole slightly larger than the capillary neck, bored in the end. Insert the neck nearly to the scratch and press downward or attempt to bend the tube; this will break the neck. Common pliers, the head of a cabinet-lock key, or similar devices may be used. Do not strike the tube against the roof, ribs, or timbering or with any object, as this is unnecessary and frequently breaks the tube at other than the desired place, making it difficult to seal the tube if not entirely spoiling it. If the scratch mark does not appear on the neck, a suitable mark may be made with a common file.

When the end is broken a hissing noise will be heard as the air rushes in. Often a grain of glass will lodge in the capillary opening and may slow down the entrance of gas and consequently the time of collection to such an extent that the collector will unknowingly seal the container before it has been filled with gas. The end should always be watched for such stoppages, the tip being kept in the zone of air being sampled, and any obstruction should be removed. Observation should be made to ascertain whether breaking the tip has severed the capillary opening. The bureau's gas laboratory has received vacuum tubes which the collector thought were filled with gas but which really contained their original vacuum.

SEALING VACUUM TUBES

After breaking the tip, hold the sample container in the place of sampling for about five seconds after the hissing noise has stopped, then remove the tube and seal it by filling the neck with the special sealing wax furnished with the bottles. The preparation of this wax is described in Bureau of Mines Bulletin 197 already cited. Do not use chewing gum, paraffins, or tar, as they do not adhere tightly to the glass. The wax is usually supplied in the form of wax-filled

metallic cartridges. The vacuum tube should be sealed by working the cap over the end of the tube with a twisting motion so that the wax is forced into the capillary opening but not so that it enters the sample container. It is important not to force the wax into the container. Finally, force the cap on as tightly as possible and make a tight joint between the cartridge and the shoulder of the tube by smoothing the wax that exuded from the cartridge. (See fig. 1.) If a small piece of wax has been used instead of a wax-filled cartridge, work it well into the capillary tube and finish the sealing by modeling a small egg-shaped lump over the tip; cover it with a small piece of thin paper to keep it

Gas should be discharged
outside sampling zone

tainers shown in Figure 2 and ordinary FIGURE 5.-Suggestions for filling container from accessible places by bottles are filled by

gas displacement

gas displacement. Figure 5 suggests how the containers may be filled. In each instance the gas is removed from the container, thus allowing it to fill with the desired atmosphere. In sampling stratified atmospheres it is good practice to discharge the exit gas at a level below that of the container for methane and at a level above for black damp. About 10 times the volume of the sample container should be passed through for taking a satisfactory sample. On that basis it will require 50 squeezes of the ordinary 50 c. c. displacement bulb to take a 250 c. c. sample.

After the required amount of air has been passed through, the tube is held in the sampling position until the bulb arrangement is removed and the container closed.

FILLING BY LIQUID DISPLACEMENT

Containers of the types shown in Figure 2 and ordinary bottles can be filled by liquid displacement. Salt solutions, water saturated with the gas to be sampled, and mercury are sometimes used instead of ordinary water with this method to obviate solubility errors when gases containing large amounts of carbon dioxide and other constituents that are appreciably soluble in water are sampled. However, from a practical standpoint this precaution is not necessary for ordinary samples of mine gases. Oxygen, carbon monoxide, hydrogen, methane, and nitrogen are not soluble enough in water to cause appreciable error. Carbon dioxide is more soluble, and where the amount of that constituent is of unusual importance ordinary water can not be used; in these cases, mercury, salt solutions, or water saturated with the gas to be sampled should be used for the liquiddisplacement method.

Briefly, water may be satisfactorily used for taking samples of ordinary mine air to determine the oxygen, methane, hydrogen, carbon monoxide, and nitrogen content. The results for carbon dioxide will be slightly low but will not ordinarily vitiate the interpretation of the results as a whole, in relation to mine-fire and explosion hazards.

The only advantage of the water-displacement method for filling containers in accessible places is convenience and time for filling.

The method for filling containers in accessible places by water displacement consists of filling them with water, taking them to the place of sampling, draining or pouring out the water, and closing. The container is held in the zone of the gas to be sampled. Facing the current, holding the container at arm's length, and other precautions previously given for filling vacuum tubes obviously apply to sampling by water displacement.

SAMPLING FROM INACCESSIBLE PLACES

In mines the only inaccessible places usually sampled are behind fire or gas seals. The place of sampling and mode of collection depend on the particular place and facilities for sampling and the ingenuity of the collector.

CHOICE OF PLACE OF SAMPLING

There is not much choice in the place of sampling in mine fires which have only one or two seals. However, if there are more than this number of seals, the choice of the place of sampling is usually a significant factor in obtaining samples that are most representative of the conditions around the fire. The most important items to be considered in this connection are (1) distance of place of sampling from fire and probable course of gases coming from the fire; (2) changes in

barometric or fan pressures that would affect the extent or direction of air change in the sealed area; (3) occurrence of water seals or falls or previously constructed temporary seals that would isolate the place of sampling from the fire area; and (4) position of sampling in the seal, mainly with respect to elevation. These items are considered in more detail under their respective headings.

PLACE OF SAMPLING WITH RESPECT TO FIRE

Samples should be obtained from the seal nearest the supposed course of the gases coming from the fire. This is not always, however, the seal nearest the fire. It is governed more by the position of the seal with respect to differences in ventilation pressures that might cause a slow but gradual air change over the fire, the direction being in general from the place of higher to that of lower pressure. This is a very important consideration in choosing the place or places for sampling.

CHANGES IN BAROMETRIC OR FAN PRESSURES

When the place of sampling has been selected its suitability should be observed each time thereafter that a sample is collected. Frequently a seal breathes; that is, it exhales or gives off gas part of the time and inhales or takes in air at other times. A sample collected when the seal is taking in air is worthless, because the air immediately behind the seal will be contaminated to an unknown degree with normal mine air. In these instances other seals (if present) should be examined as a change in pressures may have occurred that would cause others to be exhaling at this time. If all seals are taking in air, due possibly to increased barometric pressure, cooling of the area, or both, the sampling should be postponed until the conditions are more favorable for obtaining a representative sample. If these observations regarding the seal are not made, erratic results are often obtained which may lead to wrong conclusions or be inexplicable and cause undue anxiety. The pressure conditions on the seal are sometimes difficult to ascertain. A U gage filled with water will indicate appreciable pressure differences. In lieu of this, a film of water or saliva may be placed over the end of the outlet tube and observed for the direction of breaking or bulging of the film. Another method is to sift a little fine dust across the opening in the seal and by use of a flash light or electric cap lamp observe the motion, if any, in the suspended dust.

OCCURRENCE OF WATER SEALS, FALLS, OR PREVIOUSLY CONSTRUCTED TEMPORARY SEALS BETWEEN PLACE OF SAMPLING AND FIRE

Occasionally water seals form in dips, or falls occur after sealing, which isolate the chosen place of sampling from the fire. The possibility of these occurrences should be considered. Frequently one or more temporary seals of wood or canvas are present either imme