average is as satisfactory as need be obtained for the gas sampled. If the results do not agree within practical limits, an error has been made in the analysis, the samples were not representative of the same atmosphere, or one or both of the samples have been contaminated. This will necessitate additional analyses of one or both of the samples until agreement is obtained or the cause of nonagreement established. If the two samples are finally found to have a different composition and more samples of the same atmosphere can not be obtained, the sample which contained the lowest amount of oxygen (referring particularly to mine atmosphere) usually can be taken as more closely representative of the condition in question, although in the absence of information to the contrary even this sample may be viewed with suspicion. If, however, the analyst is unable to obtain duplicate results for the same sample that are within satisfactory agreement, he is making errors in his analytical technique or the apparatus is not functioning properly.

There are cases where single samples and even single analyses are satisfactory, such as samples of practically the same atmosphere or successive samples collected at a place where the composition is not subject to marked changes, as in mine-fire seals. In reality, these furnish checks on each other or on the previous samples and need not be analyzed in duplicate unless erratic or inconsistent results are obtained or the accuracy of the result is very important.

AVERAGING DUPLICATE ANALYSES

Whenever duplicate analyses have been made the arithmetical average of all satisfactory results for each constituent except carbon dioxide should be taken as the final result. For carbon dioxide the result obtained by the first analysis should be taken wherever it is significantly higher than the second, because that gas is very soluble in the water used to displace the sample from its container, and the second determination usually gives a lower result, due to an appreciable amount being taken up by the water during the period of standing. The sample for the first analysis should therefore be drawn into the burette and measured as rapidly as possible after the sample container is opened. If, however, the second analysis indicates a higher carbon dioxide content than the first by an amount greater than the analytical error of the apparatus, the discrepancy has been caused by improper technique, leaks, etc.

These rules for averaging results should always be followed unless there is outstanding evidence that one of the analyses has been more liable to error than the other, in which case the more dependable one should be taken alone. In no event, however, should one analysis be chosen merely because its results conform more closely to the expected result or the other discarded merely because its results do not con

form to them. The analytical procedure and calculation should always be carried out unmindful of interpretation of the results.

ACCURACY OF THE PORTABLE ORSAT AND SATISFACTORY RESULTS

The limits of accuracy of the portable Orsat apparatus when used by an analyst of average skill are 0.2 to 0.3 in the per cent found for carbon dioxide, oxygen, and carbon monoxide and 0.2 to 0.3 for methane in the range of 1 to 10 per cent, about 0.3 to 0.5 in the range of 10 to 30 per cent, and 1 in the range of 50 to 75 per cent. All results of duplicate samples that fall within these limits are as accurate as can be expected and are satisfactory for the intended purpose of the portable Orsat. Frequently, for inexplicable reasons, wider variations will be obtained, but whether or not the results will be unsatisfactory depends on their particular use. For example, an error of 3 when the methane content is 35 to 50 per cent is not as significant as 0.3 when it is near the inflammable limit for checking flame safety-lamp determinations or similar work with comparatively low amounts; or an error of 1 to 2 in the oxygen percentage is usually not as significant in the results of analysis of a sample taken from a fire seal when the oxygen content has fallen below 6 per cent (regardless of the methane content) as immediately after sealing when the content is 10 to 15 per cent and enough methane is present to form an explosive or potentially explosive mixture.

RECORDS AND DATA

The analyst is usually held responsible for analytical results, and for this reason he should keep a permanent record of the history of the sample and the original analytical data as recorded at the time of making the analysis. Only copies of these data should be given as reports. All data should be kept in a legible form in a permanent leaf notebook or card system and not on tablets or scraps of paper.

SOURCES OF ERROR

The principal sources of error in the analysis of mine gases are as follows:

1. Errors and omissions in general technique and procedure of analysis.

2. Faulty and careless technique in measuring gas volumes or bringing solution to marks in pipette stems.

3. Carelessness in purging apparatus or in avoiding contamination of sample during analysis.

4. Insufficient time of contact with absorbing solutions.

5. Insufficient oxygen present for combustion of methane.

6. Leaky stopcocks or apparatus connections.

7. Errors in reading gas volumes.

8. Errors in arithmetic in calculating amounts of constituent present or percentage values.

9. Worn-out absorbing solutions, especially cuprous chloride.

10. Temperature of platinum coil too low to burn methane completely in the usual procedure of analysis.

11. Traces of absorption solutions in the manifold.

12. Alkaline condition of displacing liquid in the burette or combustion pipette.

13. Nonrepresentative sample.

14. Contamination of sample before analysis.

CARE AND STORAGE OF APPARATUS

Frequently the Orsat gas apparatus and sampling equipment for mine-fire gases have not been kept in a usable condition. When need for their immediate use arises, especially in mine fires and explosions, they are either beyond repair or more than a reasonable amount of time is required for putting them in shape for use. It is therefore advisable to add the following instructions and suggestions regarding care and storage of apparatus:

RUBBER CONNECTIONS

Every two months replace all short soft-rubber connections used to hold glass tubing together, as these connections will deteriorate if left longer and may crack or tear and give trouble through leaks. After several months they will adhere tightly to the glass and be difficult to remove. The rubber connections should be renewed every two or three weeks when the apparatus is in constant use. Never replace the same connection after it has been removed for cleaning, renewing solutions, etc.

In renewing and removing the connections apparatus breakage can be avoided by not trying to slip the connections from the stems but by shaving down the side, using a sharp knife, and then peeling off the tubing. Never use gasoline to loosen adhering pieces of rubber, as some of the gasoline might accidentally get into the pipettes or manifold and thus cause erroneous results for methane.

SOLUTIONS

Pour out all solutions and wash, drain, and replace all pipettes. Although the solutions may not be worn out, it is not good practice to save old solutions, as their quality is always uncertain. Solutions that have been but slightly used should be preserved in magnesium citrate bottles.

STOPCOCKS

Before putting the apparatus in storage remove stopcocks and wipe the grease from the barrels and plugs. Clean any accumulation of grease from the bore of the plug and from the barrel outlets. A piece of small copper wire (18 to 20 B. & S. gage) with a short L bend at one end is convenient for cleaning the outlets. After

flushing the manifold and barrels with slightly acidic water, wipe the latter dry, insert a thin strip of paper about one-fourth inch wide the entire length of the barrel and over the outlet leading to the pipette, grease the plug lightly, and place it in the barrel. This will prevent the plug from sticking or freezing (described later) and will also seal the opening to the absorbing solution. Secure the plugs against dropping out and breaking by tying them with wrapping twine or looping a rubber band around the constriction below the handle and passing it over the small end of the barrel. Another convenient method for securing the plugs is to wrap one end of a 3-inch length of about 18 to 20 B. & S. gage copper wire around the constriction at the handle and the other end around one of the

branches of the stopcock. As stopcock plugs are not interchangeable, it is a good plan to remove, clean, and replace only one stopcock at a time, otherwise they might become mixed.

If a stopcock becomes stuck, warm the barrel rather quickly by placing it in boiling water for a few seconds and then, before it cools appreciably, hold the stopcock in one hand and attempt to turn the plug or tap it gently with a piece of wood. The stopcock may be tapped. on the back if the plug sticks through the barrel or on the handle if it is not the hollow-tube type. Repeat the heating from. time to time. If this does not remove the plug, the device shown in Figure 17 (which consists of a block of hardwood, such as maple,

bored to receive the handles of the cocks and to support the barrel on the rim so that the plug may be forced out by closing a vise as shown in Figure 18) is often successful. Similar devices may be purchased from laboratory-supply houses.

The method for using the device as shown is to select a hole just slightly larger than the diameter of the plug so that the barrel will fit over the rim, place the bottom of the block against one jaw of an ordinary vise, set a hardwood pin having a diameter slightly less than the small end of the plug against the end of the latter, and close the vise until pressure is exerted on the plug. Avoid strain and allow

FIGURE 18.-Method of operating stopcock stem remover

free movement of extending arms or attachments to the plug outlets. Continue tightening until the plug is loosened or breaks, because if it will not become loosened in this manner it is almost useless to try other means.

STORAGE

Store the apparatus in an upright position. The other items to be considered in storing it are: (a) Minimize danger of breakage through rough handling; and (b) keep it away from excessive heat, as the latter causes rapid deterioration of the rubber connections and air bags.