In a forecast like the present it is impossible to avoid a certain amount of overlapping in different sections of the subject and a certain blending of topics in a single chapter. The attempt to differentiate consistently between the progress of science as science, and the concurrent advance of practical invention by which scientific discovery is turned to use would only involve needless repetition. I have already had occasion to suggest elements of material progress which presuppose the advance in pure science that would make them possible. Thus, in endeavouring to suggest what the methods of commerce and the condition of our cities are likely to be in the future it was necessary to conceive certain advances in our knowledge of what is rather clumsily called "wireless" telegraphy, and to predict the discovery of new and cheap methods of analysing water into its component gases as a source of fuel and as means for the production of electricity: and in order to avoid useless repetition it was found convenient to work out in a rough manner the various ways in which the cheap and inexhaustible supplies of hydrogen and oxygen which I have imagined discovery to have placed at the disposal of invention would be employed in the arts. Similarly, when we interrogate imagination on the subject of scientific discovery itself, we shall be forced to think chiefly of the practical results likely to be achieved by it, and indeed there would otherwise be hardly any purpose to serve by the effort. What imports the greatest amount of complexity into the subject is the difficulty of conceiving the lines upon which science is likely to travel, unless we allow ourselves to be guided by the practical requirements of the future as far as we are able to foresee them. Imagination has indeed superabundant room in which to run riot when it endeavours to give form to the probabilities of scientific discovery; and the only danger is that effort may be wasted in purely fanciful directions, if it be not pretty securely tied down by some such artificial restraint as the convention of keeping more or less strictly to the anticipation of discoveries likely to have immediate practical application.