hemisphere will remain unexplored as will take up 213 years more to examine." In these numbers Herschel was perhaps mistaken. Struve at Pulkowa found 80 nights suitable out of 120 clear nights; but Sir John Herschel's experience at the Cape of Good Hope gave him the whole or parts of 131 nights in 1836, and at least 100 in the following year. The estimate of 598 years, or rather 811, by Sir William Herschel may be set down as excessive.
Herschel does not appear to have been altogether satisfied with the position he had taken up. It was not warranted by pure and inductive science. The foundation on which alone he could build with confidence had not been laid, the distance of fixed stars and nebulæ. "To these arguments," he says, "which rest on the firm basis of a series of observation, we may add the following considerations drawn from analogy." Science demands something more trustworthy than arguments and analogy. Mathematical science is not content with probability: it demands demonstration, and this he could not give. He had a distinct idea of an ocean, we shall say, of ether, transmitting light. In that ocean are thousands of floating islands, each composed of myriads or millions of shining worlds, all communicating with each other by far-piercing sunbeams. What the telegraphic messages thus sent from sun to sun, from island to island, may be, Herschel had no means at first of knowing. He came to understand and even read some of these messages in later years. We are able to read more of them now, for they tell the sizes of suns, their rates of motion, their direction of motion, and other pieces of star history incredibly interesting to curious man. Herschel did not imagine
