Prof. W. H. Pickering says that "the head of a comet, as far as our present knowledge is concerned, seems therefore to be merely a meteor swarm containing so much gaseous material that when electrified by its approach to the sun it will be rendered luminous" (_Harvard Annual_, vol. xxxii. part ii. p. 295) "... if the meteors and their atmospheres are sufficiently widely separated from one another, the comet may be brilliant and yet transparent at the same time." In the case of Swift's comet of 1892 some periodical differences of appearance were due, according to Prof. W. H. Pickering, to a rotation of the comet round an axis passing longitudinally through the tail, and he estimated the period of rotation at about 94 to 97 hours. He computed that in this comet the repulsive force exerted by the sun on the comet's tail was "about 39·5 times the gravitational force."[196] The comet known as 1902_b_ approached the planet Mercury within two millions of miles on November 29 of that year. Prof. O. C. Wendell, of Harvard Observatory, made some observations on the transparency of this comet. He found with the aid of a photometer and the 15-inch telescope of the observatory that in the case of two faint stars over which the comet passed on October 14, 1902, the absorption of light by the comet was insensible, and possibly did not exceed one or two hundredths of a magnitude,[197] an amount quite imperceptible to the naked eye, and shows conclusively how almost inconceivably rarefied the substance of this comet must be. The comet known as Morehouse (1908_c_) showed some curious and wonderful changes. Mr. Borelly found that five tails are visible on a photographic plate taken on October 3, 1908, and the trail of an occulted star indicates a slight absorption effect. According to M. L. Rabourdin, great changes took place from day to day, and even during the course of an hour! Similar changes were recorded by G. M. Gauthier; and Prof. Barnard, who photographed the comet on 30 nights from September 2 to October 13, states that the photographs of September 30 "are unique, whilst the transformation which took place between the taking of these and the taking of the next one on October 1 was very wonderful."[198] The spectrum showed the lines of cyanogen instead of the hydrocarbon spectrum shown by most comets. Prof. Barnard has suggested that all the phenomena of comets' tails cannot be explained by a repulsive force from the sun. Short tails issuing from the comet's nucleus at considerable angles with the main tail point to eruptive action in the comet itself. The rapid changes and distortions frequently observed in the tails of some comets suggest motion through a resisting medium; and the sudden increase of light also occasionally observed points in the same direction.[199]