few minutes; and finally, to securing these photographs with as few mistakes as possible when the precious moments actually arrive. Even then the whole of their efforts are quite likely to be rendered unavailing by a passing cloud; and bitter is the disappointment when, after travelling thousands of miles, and spending months in preparation, the whole enterprise ends in nothing owing to some caprice of the weather. [Sidenote: Corona follows spots.] Hence it will easily be imagined that our knowledge of the corona, the part of the sun which we can still only study on occasions of a total solar eclipse, advances but slowly. During the last twenty years there has been altogether scarcely half-an-hour available for this research, though it may fairly be said that the very best possible use has been made of that half-hour. And, what is of importance for our immediate purpose, it has gradually been established by comparing the photographs of one eclipse with those of another, that the corona itself undergoes distinct changes in form in the same period which governs the changes of sun-spots. When there are many sun-spots the corona spreads out in all directions from the edge of the sun's disc; when there are few sun-spots the corona extends very much further in the direction of the sun's equator, so that at sun-spot minimum there is an appearance of two huge wings. Although the evidence is necessarily collected in a scrappy manner, by this time there is sufficient to remove this relationship out of the region of mere suspicion, and to give it a well-established place in our knowledge of the sun's surroundings. [Sidenote: Corona may influence magnets.]