panied at a short distance by a second and fainter nebulosity, which disappeared before the principal nebulosity was lost to view. It is to be regretted that this object remained visible for so short a time as a fortnight, and that our knowledge of it depends on the authority of only one observer.[1] The Comet of 1881 (ii.), according to the testimony of 2 observers, threw off a fragment which became virtually an independent comet, and lasted as such for some days, when all trace of it was lost[2]; but a still more interesting case is that of Brooks's Comet of 1889 (v.) described in some detail in a later chapter.[3]
Another very striking instance is afforded by Swift's Comet of 1899 (i.) which was carefully studied by Bredichin. It was discovered on March 13, 1899, by L. Swift, and passed its perihelion about a month later. Before this it had a stellar nucleus of the 10th mag., a coma 7' in diameter, and a small tail. After perihelion it became visible to the naked eye and brightened up to the 3rd mag. with a tail several degrees long. On May 7 the nucleus was observed at the Lick Observatory to be double, and the 2 portions gradually separated until on May 21 they were 29' apart. The fainter portion was followed till June, until it was too faint to be seen. The comet eventually assumed what is known as the "scymitar" form, and showed indications of twisting which suggested the idea of rotation or oscillation about the line drawn from the Sun to the comet. The tail was of type I of the Bredichin types, with the exception of a faint stream which was of type III.[4] There was no difficulty in tracing on photographs the outward and vibratory motion of the material of the tail, and Bredichin says that the extremity of the tail, as seen on May 19, was formed of matter which had left the head 4 days earlier. He does not hesitate to say that the partition of the tail was caused by the disturbing influence of the Sun, and that both nuclei were moving in hyperbolic paths, the smaller nucleus
