Diffusion of Gases.—If two flasks, each filled with a different gas at a given pressure and temperature, be placed in communication with each other, the gases will be found to interdiffuse. The rate of interdiffusion is shown by theory to be nearly in inverse proportion to the square root of the product of the densities of the two gases, and the experimental results are in accordance with the theory.
Effusion of Gases.—This is exhibited in the passage of a gas into vacuum under constant pressure through a small opening in a very thin plate otherwise impervious to it. The work done in the passage of a given volume of the gas is proportional to the pressure, and the equivalent kinetic energy is proportional to the product of the density and the square of the speed of effusion. Hence the speed for a given pressure varies inversely as the square root of the density. Graham showed that this result of theory is closely realised by experiment. He showed, further, that when a discrepancy exists, it is due to the finite thickness of the plate.
Transpiration of Gases is the term for the passage of gases under pressure through a fine capillary tube. This subject was also investigated by Graham, who found that the rate of passage is not affected by the material of the tube. This seems to indicate that the tube becomes coated internally with a thin film of gas, so that the opposition to the flow of gas is due to Viscosity (q.v.).
The rates at which different gases pass through fine unglazed earthenware are inversely as the square roots of their densities. Hence we have a means of separating gases the densities of which are different (see ATMOLYSIS). If the septum be made of caoutchouc, which is not porous, the passage of gases still occurs. The gas seems to combine with the matter of the septum on the oneside, to diffuse through it, and finally to be given off on the other side. The passage of some gases, such as carbonic oxide, through hot cast-iron is analogous.