"A Woulff's bottle, a in the annexed figure, of blown glass and holding about 1/4 litre is used as the generating vessel. One neck, about 15 mm. in internal diameter, is connected by flexible tubing with a globular vessel b, having two tubulures, and this vessel is further connected with a conical flask c, holding about 100 c.c. The other neck is provided with tubing d, serving to convey the gas to the inlet-tube, with tap e, of the 20-litre measuring vessel f, which is filled with water saturated with acetylene, and communicates through its lower tubulure with a similar large vessel g. The generating vessel a is charged with about 150 c.c. of water saturated with acetylene. The vessel f is filled up to the zero mark by raising the vessel g; the tap e is then shut, and connexion is made with the tube d. Fifty grammes (or say 2 oz.) of the pulverised carbide are then weighed into the flask c and this is connected by the flexible tubing with the vessel b. The carbide is then decomposed by bringing it in small portions at a time into the bulb b by raising the flask c, and letting it drop from b into the generating vessel a, after having opened the cock e and slightly raised the vessel f. After the last of the carbide has been introduced two hours are allowed to elapse, and the volume of gas in f is then read while the water stands at the same level in f and g, the temperature and pressure being noted simultaneously."
A second, but less commendable method of decomposing the carbide is by putting it in a dry two-necked bottle, one neck of which is connected with e, and dropping water very slowly from a tap-funnel, which enters the other neck, on to the carbide. The generating bottle should be stood in water, in order to keep it cool, and the water should be dropped in at the rate of about 50 c.c. in one hour. It will take about three hours completely to gasify the 50 grammes of carbide under these conditions. The gas is measured as before.
Cedercreutz has carried out trials to show the difference between the yields found from large and small carbide taken from the same drum. One sample consisted of the dust and smalls up to about 3/5 inch in size, while the other contained large carbide as well as the small. The latter sample was broken to the same size as the former for the analysis. Tests were made both with a large testing apparatus, such as that shown in Fig. 22, and with a small laboratory apparatus, such as that shown in Fig. 23. The dust was screened off for the tests made in the large apparatus. Two sets of testings were made on different lots of carbide, distinguished below as "A" and "B," and about 80 grammes wore taken for each determination in the laboratory apparatus, and 500 grammes in the large apparatus. The results are stated in litres (at normal temperature and pressure) per kilogramme of carbide.
___________________________________________________________________
| | | |
| | "A" | "B" |
|_____________________________________________________|______|______|
| | | |
| Lot |Litres|Litres|
| Small carbide, unscreened, in laboratory \ (1) | 276 | 267 |
| apparatus . . . . . / (2) | 273 | 270 |
| Average sample of carbide, unscreened, in \ (1) | 318 | 321 |
| laboratory apparatus . . . / (2) | 320 | 321 |
| Small carbide, dust freed, in large apparatus (1) | 288 | 274 |
| Average sample of carbide, dust freed, in \ (2) | 320 | 322 |
| large apparatus . . . . / | | |
|_____________________________________________________|______|______|
As the result of the foregoing researches Cedercreutz has recommended that in order to sample the contents of a drum, they should be tipped out, and about a kilogramme (say 2 to 3 lb.) taken at once from them with a shovel, put on an iron base and broken with a hammer to pieces of about 2/5 inch, mixed, and the 500 grammes required for the analysis in the form of testing plant which he employs taken from this sample. Obviously a larger sample can be taken in the same manner. On the other hand the British and German Associations' directions for sampling the contents of a drum, which have already been quoted, differ somewhat from the above, and must generally be followed in cases of dispute.
Cedercreutz's figures, given in the above table, show that it would be very unfair to determine the gas-making capacity of a given parcel of carbide in which the lumps happened to vary considerably in size by analysing only the smalls, results so obtained being possibly 15 per cent. too low. This is due to two causes: first, however carefully it be stored, carbide deteriorates somewhat by the attack of atmospheric moisture; and since the superficies of a lump (where the attack occurs) is larger in proportion to the weight of the lump as the lump itself is smaller, small lumps deteriorate more on keeping than large ones. The second reason, however, is more important. Not being a pure chemical substance, the commercial material calcium carbide varies in hardness; and when it is merely crushed (not reduced altogether to powder) the softer portions tend to fall into smaller fragments than the hard portions. As the hard portions are different in composition from the soft portions, if a parcel is sampled by taking only the smalls, practically that sample contains an excess of the softer part of the original material, and as such is not representative. Originally the German Acetylene Association did not lay down any rules as to the crushing of samples by the analyst, but subsequently they specified that the material should be tested in the size (or sizes) in which it was received. The British Association, on the contrary, requires the sample to be broken in small pieces. If the original sample is taken in such fashion as to include large and small lumps as accurately as possible in the same proportion as that in which they occur in the main parcel, no error will be introduced if that sample is crushed to a uniform size, and then subdivided again; but a small deficiency in gas yield will be produced, which will be in the consumer's favour. It is not altogether easy to see the advantage of the British idea of crushing the sample over the German plan of leaving it alone; because the analytical generator will easily take, or its parts could be modified to take, the largest lumps met with. If the sample is in very large masses, and is decomposed too quickly, polymerisation of gas may be set up; but on the other hand, the crushing and re-sampling will cause wastage, especially in damp weather, or when the sampling has to be done in inconvenient places. The British Association requires the test to be made on carbide parcels ranging between 1 and 2-1/2 inches or larger, because that is the "standard" size for this country, and because no guarantee is to be had or expected from the makers as to the gas-producing capacity of smaller material. Manifestly, if a consumer employs such a form of generator that he is obliged to use carbide below "standard" size, analyses may be made on his behalf in the ordinary way; but he will have no redress if the yield of acetylene is less than the normal. This may appear a defect or grievance; but since in many ways the use of small carbide (except in portable lamps) is not advantageous--either technically or pecuniarily--the rule simply amounts to an additional judicious incentive to the adoption of apparatus capable of decomposing standard-sized lumps. The German and Austrian Associations' regulations, however, provide a standard for the quality of granulated carbide.
It has been pointed out that the German Association's direction that the water used in the testing should be saturated with acetylene by a preliminary decomposition of 1/2 kilogramme of carbide is not wholly adequate, and it has been suggested that the preliminary decomposition should be carried out twice with charges of carbide, each weighing not less than 1 per cent. of the weight of water used. A further possible source of error lies in the fact that the generating water is saturated at the prevailing temperature of the room, and liberates some of its dissolved acetylene when the temperature rises during the subsequent generation of gas. This error, of course, makes the yield from the sample appear higher than it actually is. Its effects may be compensated by allowing time for the water in the generator or gasholder to cool to its original temperature before the final reading is made.
With regard to the measurement of the temperature of the evolved gas in the bell gasholder, it is usual to assume that the reading of a thermometer which passes through the crown of the gasholder suffices. If the thermometer has a very long stem, so that the bulb is at about the mid-height of the filled bell, this plan is satisfactory, but if an ordinary thermometer is used, it is better to take, as the average temperature of the gas in the holder, the mean of the readings of the thermometer in the crown, and of one dipping into the water of the holder seal.
The following table gives factors for correcting volumes of gas observed at any temperature and pressure falling within its range to the normal temperature (60° F.) and normal barometric height (30 inches). The normal volume thus found is, as already stated, not appreciably different from the volume at 15° C. and 760 mm. (the normal conditions adopted by Continental gas chemists). To use the table, find the observed temperature and the observed reading of the barometer in the border of the table, and in the space where these vertical and horizontal columns meet will be found a number by which the observed volume of gas is to be multiplied in order to find the corresponding volume under normal conditions. For intermediate temperatures, &c., the factors may be readily inferred from the table by inspection. This table must only be applied when the gas is saturated with aqueous vapour, as is ordinarily the case, and therefore a drier must not be applied to the gas before measurement.