June 4, 1915. THE COLLIERY GUARDIAN. 1169 mine as a representative sample of the air at the time of sampling going up the upcast. In order to take our sample, we must have a bottle ; the type of bottle recommended by Dr. Haldane is ideal. Briefly, it is a 3 or 4 ounces narrow-necked bottle, fitted with a well-fitting stopper, which admits of its being kept securely in its place by either a rubber band passing over the top of the stopper and under the bottom of the bottle, or a loop of string attached to a belt of string round the neck. The latter I prefer, for the reason that it cannot be lost. I have had at times curious bottles sent to me, bottles with necks 2 in. wide. They are troublesome to deal with. The best bottle 1 know is an English glass bottle, supplied by Janies Woolley and Sons, Manchester. The stopper of this bottle is a perfect fit, which is essential if a good sample is to be retained. The bottle used must first be pre- pared by well washing, the final wash being with distilled water, carefully dried, and the stopper greased with a little vaseline, just so much that when, the stopper is placed in the neck and pres-ed home with a slight turn a perfectly clear film without air streaks is apparent between the stopper and the ground neck. I would like to advise here that on no account during the period of sampling must the stopper be allowed to come m contact with foreign matter. A little dust or grit on the stopper frequently prevents it being pressed well home, with a consequent spoiling of sample. Dr. Haldane clearly points out in his book on Methods of Air Analysis the importance of clean dry bottles. I quote his examples :— If a bottle is wet and dirty, an appreciable amount of carbon dioxide may appear and of oxygen disappear by bacterial action. To quote an example, after four days, the carbon dioxide had increased by 0’05 per cent., and after ten days by 0*12 per cent, in a wet and dirty bottle. If, on the other hand, the bottle is wet and clean, the carbon dioxide gradually disappears, as it is absorbed by alkali dissolved out of the glass by the water. For instance, the whole of the carbon dioxide (0'03 per cent.) in a sample of pure atmospheric air had disappeared in a wet bottle in twelve days. I would like to mention here that if carbon monoxide is suspected a second bottle should be used, only the stopper, instead of being greased with vaseline, should be rendered completely airtight with a film of glycerine, this being more satisfactory for the subsequent blood test for carbon monoxide. The sample of air may now be taken in the following manner:—A short length of glass tubing, about 8 in. long and T3K in. internal diameter, is attached to a length of rubber tubing from 2 to 3 ft. long, the other end of which may conveniently have a short piece of glass tubing inserted for a mouthpiece. The long glass tube is now inserted into the neck of the sample hottie, and the latter being inverted, can be held in any desired position by means of the glass tube. The other end of the rubber tube with glass mouthpiece being placed in the mouth, several deep breaths taken through the tube draws the air to be sampled through the bottle. Care should be taken that the exhaled breath has no possible chance of contaminating the sample. The operator should, when possible, stand with his face to the air current. The glass tube is removed from the bottle whilst inhalation is still in progress, thus avoiding all chance of breath entering the bottle through the tube. The stopper should be immediately placed in the bottle, before the position of the bottle is altered, and adjusted so that the film of vaseline between the two surfaces of glass becomes perfectly clear. The rubber band or string is now placpd in position; otherwise the reduction in pressure on bringing the sample to the surface may cause the, stopper to blow. The label is fixed with description or identification number. Whether placed on the label or recorded in notebook, temperature, wet and dry bulb readings, depth and barometric readings, might always with advantage be recorded. An alternative to the above method is the use of a rubber hand pump, 20 to 25 compressions of the bulb being ample completely to wash out the Dottie with the air to be sampled. The use of such a pump enables samples of air to be drawn from behind stoppings or holes where gas may have accumulated—gas which obviously would be unpleasant to breathe. In an airway it frequently happens that the compo- sition of the gas is not uniform. For example, when a district return joins a main return stream lines may be traced. In such cases a truly representative sample is difficult to obtain. A piece of brattice vigorously used some distance in.advance is useful, or, in the hands of a capable man, a special apparatus can be used, filled with mercury, in which the flow of mercury out, and the flow of gas in, can be regulated in such a manner that the period of sampling can be extended, and thus enable an average sample to be taken in a manner similar in principle to the taking of an average anemometer reading. To sample air from behind stoppings a pipe through the stopping is necessary, and is frequently built in with the stopping. If the pressure behind the stopping is positive, the operation is simple. The end of the pipe is fitted with rubber, cork, or clay bung, with a glass tube passing through. A rubber tube can be attached to the glass tube, but it is seldom necessary. The end of the glass or rubber tube is placed inside the sample bottle and allowed to remain there until it is certain that the bottle is completely washed out with the air to be sampled. When this is the case the bottle is slowly withdrawn from the tube, for if the flow of air through the tube be at all sluggish it is probable that air may be drawn into the sample to make good-the displacement of the withdrawn tube. The stopper is placed in the bottle immediately the tube has left the neck. If instead of a positive pressure the pressure is a negative one, unless the stopping is a good one, the sample would not be of much value, being contaminated with air -leaking through. If one is dealing with a tight stopping, however, a sample may be obtained as follows- A length of metal tubing of suitable bore is carefully pushed through the pipe in the stopping, care being taken to prevent, as far as possible, air passing in between the tube and the pipe. A plug of wet cotton waste is useful for this purpose. The metal tube should be long enough to pass completely through the pipe, and some distance beyond it. Lengths of iron gas piping, joined together during the operation, may be used, care being taken to keep the outside end plugged. When the tube is well through the stopping, the joint between the tube and the pipe is made tight with clay. A one-holed rubber stopper with glass tube is attached to the rubber pump before mention*-'d, and quickly inserted in the end of the tube in place of the plug. Sufficient air must now be pumped from the inside of the stopping completely io wash out the tube' and bottle—the latter being withdrawn from the tube slowly and during a compression of the pump, for the reason mentioned above, the stopper being placed in position immediately. It is always advisable to take two samples like this, the second having the air pump through the bottle for a little longer period, in order to see if the sample is satisfactory. Some years ago, I had the opportunity of taking samples of air subsequent to an explosion of considerable extent. One of the most interesting samples obtained was taken as follows:—Both upcast and downcast shafts had been sealed, but it was thought advisable to have samples taken at the top of the upcast, and for this purpose a hole was cut through the stopping, sufficiently large to admit of a pipe passing through, the pipe being about in. internal diameter. Samples were taken through this in the ordinary way. Befoie opening the pit, it was suggested that a knowledge of the state of the air at the pit bottom would be of value, and, consequently, I obtained the sample after some difficulty. I passed through the iron pipe mentioned above a long glass tube about 4 ft. long and | in. diameter. This tube was fitted at either end with a rubber stopper, through which passed fine capillary glass tubes. The whole was filled with water, and the upper capillary was covered with a rubber cap,, thus preventing the flow of water out of the tube. The whole was suspended by means of fine copper wire, which was first attached to the rubber plug and finally to the main tube, a small amount of slack being left between the plug and the tube. The plug was so arranged that it would readily bear the weight of the tube, but could be displaced by a sudden jerk, when, of course, the water slowly flowed out of the tube through the bottom capillary, drawing in air through the top. This apparatus was carefully lowered into the shaft; and after one or two failures, due to fouling wires in the shaft, a sample was taken, the tube being rapidly raised to the surface, plugs placed over the capillary tubes, and transferred straight away to a bottle over water and analysed. Under ordinary circumstances, little change would take place in the composition of the sample left in such a tube for a short period, air passing in or sample out having to diffuse through the capillary tubes. In the above case, however, the pressure of the sample would be a little greater than the barometric pressure at the surface, and consequently no air could pass in. The analysis gave the following figures :— Percent. Oxygen.......... 16*72 Nitrogen ....... 74*06 Firedamp .......... 7*83 Carbon dioxide .... 1*39 Per cent. Pure air ....... 79*89 Blackdamp ...... 12*28 Firedamp....... ... 7*83 This was afterwards shown by examination of samples taken by the first party down to be a fair representa- tion of the air at the pit bottom. In cases where carbon monoxide gas is suspected a mouse, as recommended by Dr. Haldane, may be used, although a fair amount of carbon monoxide may be present without the mouse showing much sign of distress. If the animal is allowed to breathe air containing only 0*01 or 0*02 per cent, carbon monoxide for 15 to 20 minutes, and then killed in the suspected atmosphere, its blood will give a faint indication, increasing with the increase of carbon monoxide present. Sampling Water. My concluding note will be on the methods of sampling water. It is necessary at times for samples of water to be taken, and for this purpose a bottle should be used, fitted with a well-fitting glass stopper. The best bottle is the one known as the “ Winchester quart,” holding about 2^ litres, which can he procured from most chemists. Stoneware bottles should be avoided, since they are difficult to clean and also may affect the hardness of the water. If a stoppered bottle is not available, a cork bottle has to be used. The cork, if possible, should be a new one, well washed with the water to be sampled. The well-washed bottle should be washed out several times with the water to be sampled. For ordinary analysis,-one Winchester quart bottle is sufficient. If, however, the water has to have its purity determined for drinking purpo-es, at least three should be taken. If the sample has to be taken from a tank, pool, or shallow well, great care must be taken to avoid stirring up the bottom deposit, either by contact, or by careless movement of the bottle. Floating impurities, such as oil, scum, &c., are frequently present. To avoid taking these in with the sample, the bottle may be carefully submerged with the thumb placed over the neck, removing the thumb when the’bottle is at such a depth that no possibility exists of surface matter being drawn in. A better arrangement consists of a two-holed stopper, fitted with a short glass tube through one hole and a long one through the other, the ends of both tubes terminating just through the stopper. By placing the thumb pve.r the end of the short tube before sub- merging and removing when well under the water, the latter is allowed to enter. The air leaving the bottle by the longer tube does not disturb the water by bubbling through, as in the case of the simpler method. If the water is required to be tested for its purity with regard to its potableness, the sample should be as far as possible kept from the light until analysed, which should be done without delay. The Discussion. The President (Mr. J. Gregory) said they had had a very interesting paper. He supposed most of them at some time had had to take samples of various kinds, but few of them had appreciated that the work, which one might naturally assume might be delegated to a labourer, was really a highly scientific proposition. The importance of the paper was this, that when a sample was submitted for analysis, in the majority of cases, the bulk was collected at the works or colliery, and it was not until the bulk had been reduced to a comparatively small amount that it came into the chemist’s hands. Mr. Gregory mentioned that in sampling slack the screens were the best place to get it, and he pointed out that in wagons he had examined there were several cones. Probably they had been loaded from a hopper, the slack having come down by two or three different roads. Several of the hints given were of great value. The means of taking a sample of water was very useful, because it was difficult in the case of water having a scum on it to take a sample by the ordinary means free from scum. Mr. Yates said the paper was very instructive and useful. With regard to the sampling of the slack, Mr. Gregory said it was surprising to find the different qualities of the various wagons of slack. All practical pitmen knew the reason. Both the nature of the seam they were working and the human element entered into it. In the case of the seams he was dealing with they got bad samples of slack sometimes, and they came from places where the men threw dirt out with the coal. The defect had to be lemedied in the pit. Where they had a band of soft dirt in the middle of the coal, unless that were removed in the process of working, they were bound to have an inferior quality of slack. This was a matter that gave him a great deal of concern, and he was airaid the chemist could not help them very much in this matter. All he could do was to give them analyses of the various seams they were working. Mr. E. P. Turner said he had listened with very great pleasure to the paper. In regard to the sampling of the trucks mentioned, he should like to know whether the slack came from different seams, or whether it all came from one. It seemed remarkable that they should vary so considerably. They had had many samples taken, and they had never found so much variation, though they had adopted a method of sampling similar to that which Mr. Gregory had mentioned. With reference to the sampling of coal seams, they had nearly always adopted the method Mr. Gregory described, but in the sampling of ironstone seams there was very much greater difficulty, because they could not get a uniform section right through the seam, particu- larly where they had a seam 5 or 6 feet thick. He did not know whether Mr. Gregory had had experience in sampling ironstone seams. A hearty vote of thanks was accorded to Mr. Gregory for his paper, and the discussion was adjourned. Mr. Gregory, in reply, said the sampling of iron- stone seams was certainly a difficult matter. He did not think, even after taking many samples in various parts of a single seam, that they would be justified in saying that they bad a representative sample of the whole. The only thing was to apply the results as far as practicable. That was the reason, of course, of applying the system of taking the sample in bulk. As to the samples taken from the wagons, he should think one or two seams were represented in those wagon loads. Swedish Iron Trade.—The last meeting of the Swedish Ironworks Association was held at Stockholm on April 30, and the following is a translation of the report then pre- sented :—The iron market has improved materially by degrees. Home and foreign demands for most products are good, and have admitted of the rise in price which was necessitated by the enormously increased cost of raw material, fuel, etc. The spring floods have increased the water supply, thus extending the productive capacity of the mills. At present, therefore, most of the works are fully occupied, but much rainfall will be required during the next few months if a scarcity of water is to be avoided. A normal supply of water in the lakes has not, by far, been obtained yet. The unsettled conditions produced by the world war still prevail, and cause constant surprises, which injure the iron trade to a not inconsiderable extent. Great care is still necessary for export business, especially for contracts with long delivery periods. The following is a specification of the total exports of iron and steel from Sweden during the first quarter of the current year, as compared with the same period of 1914 :— Pig iron Jan. 1 to Mar. 31. Increase or decrease 1915. 1915. 43,300 .. 1914. . 19,600 . + 23,700 Siliceous and silico- manganese iron 1,800 .. . 3,000 . — 2,200 Spiegeleisen, spongy iron, &c 400 .. . 1,200 . 800 Scrap iron 800 .. . 1,400 . — 600 Ingots 1,500 .. .. 3,000 . — 1,500 Blooms 1,200 .. .. 1,200 . 0 Bough bars 1,800 .. . 2,000 . — 200 Billets 1,700 .. . 1,500 . 4- 200 Solid tube material .. 3,200 .. . 4,500 . — 1,300 Iron rods 18,600 .. . 21,200 . — 2,600 Rod scrap..... 1,300 .. . 1,000 . -h 300 Wire rods 6,400 .. .. 8,100 . — 1,700 Plates 1,700 .. 500 . 4- 1,200 Tubes 4,800 .. 4,100 . 4- 700 Wire, drawn or cold- rolled 400 .. 300 . .. + 100 Nails 1,900 . .. 1,600 . 4- 300 Tons 90,800 ., .. 74,200 . .. 4- 16,600