Schwarzschild, as already stated, calculated that small spherules reflecting all the incident light and of the specific gravity of water would be repelled by the sun with a force that might balance ten times their weight. For a spherule absorbing all the light falling upon it this figure would be reduced to five times the weight. The small particles of comets which, according to spectroscopic observations, probably consist of hydrocarbons are not perfectly absorbing, but they permit certain rays of the sun to pass. A closer calculation shows that in this case forces of about 3.3 times the gravity would result. Larger spherules yield smaller values. Bredichin’s second and third classes would thus be well adapted to meet the requirements which the radiation pressure demands. It is more difficult to explain how such great forces of repulsion as those of the first group of Bredichin or of the peculiar comets of Swift and of Roerdam can occur. When a particle or drop of some hydrocarbon is exposed to powerful radiation, it may finally become so intensely heated that it will be carbonized. It will yield a spongy coal, because gases (chiefly hydrogen) will escape during the carbonization, and the particles of coal will resemble the little grains of coal-dust which fall from the smokestacks of our steamboats, and which afterwards float on the surface of the water. It is quite conceivable that such spherules of coal (consisting probably of so-called marguerites, felted or pearly structures resembling chains of bacilli) may have a specific gravity of 0.1, if we make allowance for the gases they include (compare page 106.) A light-absorbing drop of this density of 0.1 might, in the most favorable case, experience a repulsion forty times as strong as the gravitation of the sun. In this manner we can picture to ourselves the possibility of the greatest observed forces of repulsion. The spectra of comets confirm in every respect the conclusions to which the theory of the radiation pressure leads up. They display a faint, continuous spectrum which is probably due to sunlight reflected by the small particles. Besides this, we observe, as already mentioned, a spectrum of gaseous hydrocarbons and cyanogen. These band spectra are due to electric discharges; for they are observed in comets whose distance from the sun is so great that they cannot appear luminous owing to their own high temperatures. In the tail of Swift’s comet banded spectra have been observed in portions which were about five million kilometres from the nucleus. The electric discharges must chiefly be emitted from the outer parts of the tails, where, according to the laws of static electricity, the electric forces would be strongest. For this reason the larger tails of comets look as if they were enveloped in cloaks of light of a more intense luminosity.