stars and their positions would be recorded in their millions; but, said the philosopher to whom I have referred, though you might accomplish all this, and much more in the same direction, yet there is a well-marked limit to your possible achievements; you can, he said, never expect to discover the actual chemical elements of which the heavenly bodies are composed. Nobody could dispute the reasonableness of this statement at the time he made it; indeed, it seemed to be a necessary deduction from our knowledge of the arts of chemistry, as those arts were understood before the middle of the last century. In the prosecution of his researches by the older method, the chemist could no doubt discover the different elements of which the body was formed. That is to say, his art enabled him to accomplish this task, provided one very essential and fundamental condition could be complied with. However accomplished the chemist of fifty years ago might have been, he would assuredly have thought that he was being mocked if asked to determine the composition of a body which was 93,000,000 miles away from him. The very idea of forming an analysis under such conditions would have been scouted as preposterous. He would naturally ask that a specimen of the body should be delivered into his hands, a specimen which he could take into his laboratory, pulverise in his mortars, place in his test-tubes, treat with his re-agents, or examine with his blowpipe. Only by such methods was it then thought possible to obtain an analysis and discover the elements from which any given substance was formed. For in the early part of this century the splendid method of spectrum analysis, that method which has revealed to us so many of the secrets of Nature, had not yet come into being. When that memorable event took place it was at once perceived that the spectroscope required no actual contact with the object to be tested, but only asked to receive some of the rays of light which that object dispersed when sufficiently heated. It was obvious that this new method must be capable of an enormously enlarged application. The flame producing the vapour might be at one end of the room, while the spectroscope testing the elements in that vapour might be at the other end. This new and beautiful optical instrument could analyse an object at a distance of a hundred feet. But if applicable at a distance of a hundred feet, why not at a hundred yards, or a hundred miles, or a hundred million miles? Why might the method not be used if the source of light were as far as the sun, or as far as a star, or even as far as the remotest nebula, whose faint gleam on the sky is all that the mightiest telescope can show.