The most important points to be noticed in the behavior of large comets, as they approach the sun, is that usually the side of the coma which lies toward the sun is the scene of intense disturbance. Streams of luminous matter seem to rise continually toward the sun, attaining a certain distance from the head, when, assuming a cloud-like appearance, they seem to form an envelope around the nucleus. This envelope gradually increases its distance from the sun, growing fainter and larger, while within it the process is repeated, and a new envelope is formed. This in turn ascends from the nucleus, expanding as it does so, while within it a new envelope is formed. Meanwhile, the one first formed has grown fainter, perhaps has disappeared. But sometimes the process goes on so rapidly (a day or two sufficing for the formation of a complete new envelope) that several envelopes will be seen at the same time, the outermost faintest, the innermost most irregular in shape and most varied in brightness, while the envelope or envelopes between are the best developed and most regular. The matter raised up in these envelopes seems to have undergone a certain change of character, causing it no longer to obey the sun’s attractive influence, but to experience a strong repulsive action from him, whereby it is apparently swept away with great rapidity to form the tail. “It flows past the nucleus,” says Dr. Huggins, “on all sides, still ever expanding and shooting backward until a tail is formed in a direction opposite to the sun. This tail is usually curved, though sometimes rays or extra tails sensibly straight are also seen.” The description is, however, incomplete in one important respect. The matter raised from the nucleus to form the envelopes may be, and probably is, carried past the nucleus on all sides; but the appearance presented by the tail just behind the nucleus is not exactly in accordance with our ideas as to what should result from the flowing past “on all sides.” There is a dark space immediately behind the nucleus, that is, where the nucleus, if solid, would throw its shadow, if there were matter to receive the light all around so that the shadow could be seen. Now it may be thought at first that this corresponds exactly with what should be seen: when we look just behind the nucleus there is no light, or very little; when we look on either side of that dark space there is the luminous matter which has been driven back from the envelopes in front of the nucleus. But if the luminous matter flows past the nucleus _on all sides_, it must flow past the nucleus on the side nearest to the observer, and also on the side farthest away; and it is just where the line of the sight passes through these two regions of brightness that a dark streak is seen just behind the nucleus. Let the reader draw two concentric circles—one an inch in diameter, the other two inches—and let him then draw two parallel tangents to the inner circle on opposite sides of it.