566 THE COLLIERY GUARDIAN. September 11, 1914 that which has the faster rate in general evolves the larger quantity of heat. This fact at once suggests a possible error in Lamplough and Hill’s results. If the measurement of the oxygen absorbed was wrong by a fairly constant factor, the effect would show most in those experiments in which the smallest quantity of oxygen was absorbed. A glance at the figure of the apparatus used suggests one such error. The reaction flask was connected to a bottle containing oxygen by a glass tube which allowed free communication between the two. The oxygen percentage in the reaction flask fell considerably during the reaction, even to as low as four, yet no allowance was made for diffusion of gases back from the reaction flask into the large bottle. That such diffusion took place cannot be questioned. In two cases evolution of carbon dioxide was sufficient to raise the pressure above atmospheric, so that gas undoubtedly did pass from the reaction flask into the reservoir, in which a small change in the oxygen percentage would have a very large effect on many of the results. A further error has been discovered. In determining the total volume occupied by oxygen in the reaction flask, Lamplough and Hill determined the density of the coal by immersion, and from this calculated the space that the weight of coal used would occupy in the reaction flask, assuming that the difference between this volume and the volume of the flask was the volume of oxygen contained by the latter. This assumption is not, however, borne out by experiment. Many tests were made, both with oxygen and with nitrogen, and it is found that considerably more oxygen or nitrogen will enter the flask than calculated by Lamplough and Hill’s method. For example, a flask of a volume of 450 cu. cm. was filled with 260 grammes of coal dust, and evacuated completely. The density of the coal determined in the ordinary way was 1-3, hence the volume available for gas would be, according to Lamplough and Hill:— 450 — = 250 cubic centimetres. Experiment showed that 355 cu. cm. of nitrogen would enter the flask before the pressure reached atmo- spheric. A similar test was made with oxygen. In this case a correction has to be applied for the oxygen absorbed during the measurement, but this may be determined very closely by experiment. Almost the same result was obtained as with nitrogen, 360 cu. cm. of oxygen entering the flask. The coal, therefore, behaved as if it occupied a space of about 100 cu. cm., instead of the 200 calculated from the density measure- ments in the ordinary way. The coal for this purpose behaves exactly as if it had a density of about 2-6. The reaction flask used by Lamplough and Hill has a capacity of 464 cu. cm., and contained about 300 grammes of coal; their calculation would therefore give them a “ dead space ” of 464 — ??? = 233 cu. cm., whereas rd experiment shows that the effective “ dead space ” is 464 — = 349 cubic centimetres. This difference will affect materially the calculation of the quantity of oxygen absorbed, for examination of the formula given in the paper shows that a considerable factor in it is based on the different percentages of oxygen in the reaction flask alone, at the beginning and end of the experiment. No details are given by Which the whole of the results may be corrected, but there are data for three cases, namely :—(1) For “ pyrites from coal ”; (2) for top softs rich in pyrites; and (3) for top softs. A recalculation based on the above gives :— Lamplough Recal- and Hill. culation. Pyrites from coal................... 3’3 ... 2’3 Top softs, first part .............. 3’2 ... 2’5 Top softs, containing 43% of pyrites 3’6 ... 2’7 The original results are considerably reduced, but, until the effect of the diffusion previously mentioned is known, complete corrections cannot be applied. It is sufficiently clear, however, that the importance of several factors has been under-estimated by Lamplough and Hill, and that a revision of their results would give a figure approximating more closely to that now found. Experiments are now being made to investigate the peculiarity of “ dead space ” found and described in this paper. It is so far clear only that the gas is not absorbed or adsorbed in the ordinary sense, for it still obeys Boyle’s law. It would seem not impossible that the coal is in reality porous, and that gas permeates the pores in a way that a liquid cannot. Density deter- minations by immersion would then give an incorrect result. Some such explanation must be adduced to account for the large quantities of hydrocarbons which can be held by coal. The results obtained in the present section of the paper confirm the fact that a considerable quantity of heat is evolved by the absorption of oxygen by coal. The actual heating effect is discussed in Part IV. Part IY.—The Influence of Temperature. The experiments described in the present section of the paper are a continuation of those described in Part I., in which the rate at which coal dust absorbs oxygen was studied at temperatures up to 60 degs. Cent. Experiments at higher temperatures are now described. The method adopted for determining the rate of absorp- tion was essentially the same as that described in the first paper, except that a different thermostat had to be employed to suit the higher temperatures. A very con- venient thermostat for high temperatures may be made from a glass beaker, filled with about 4 litres of a suit- able oil, on the outside of which a coil of wire is wound, through which a current passes. By putting a suitable resistance in series with the coil, and arranging that a mercury regulator in the bath shall cut this resistance out or put it in, as the temperature falls or rises, a very efficient regulation can be obtained. The bath is fitted with a stirrer driven by an electric motor. A tempera- ture of 160 degs. Cent, can be maintained constant to 0-5 deg. Cent, over an indefinite period. Table HI.—Rate of Absorption of Oxyqen bj Barnsley Hards. After Hours. 3