We have been dealing hitherto with matters that must be regarded as acquired scientific results—not that they will never be found to need improvement, but that further progress must be built upon them, as Einstein is built upon Newton. Science does not aim at establishing immutable truths and eternal dogmas: its aim is to approach truth by successive approximations, without claiming that at any stage final and complete accuracy has been achieved. There is a difference, however, between results which are pretty certainly in the line of advance, and speculations which may or may not prove to be well founded. Some very interesting speculations are connected with the theory of relativity, and we shall consider certain of them. But it must not be supposed that we are dealing with theories having the same solidity as those with which we have been concerned hitherto. One of the most fascinating of the speculations to which I have been alluding is the suggestion that the universe may be of finite extent. Two somewhat different finite universes have been constructed, one by Einstein, the other by De Sitter. Before considering their differences, we will discuss what they have in common. There are, to begin with, certain reasons for thinking that the total amount of matter in the universe is limited. If this were not the case, the gravitational effects of enormously distant matter would make the kind of world in which we live impossible. We must therefore suppose that there is some definite number of electrons and protons in the world: theoretically, a complete census would be possible. These are all contained within a certain finite region; whatever space lies outside that region is, so to speak, waste, like unfurnished rooms in a house too large for its inhabitants. This seems futile, but in former days no one knew of any alternative possibility. It was obviously impossible to conceive of an edge to space, and therefore, it was thought, space must be infinite.