But if the first day of the perihelion overwhelms us with difficulty, the second day will prove disastrous to the common theory. For the extremity of the tail farthest from the sun will be required to pass with lateral motion from its perpendicular 100,000,000 miles, so that it may be in advance of the nucleus and again rest on its orbit. This orbit is an impassable line, and therefore instantly arrests the prodigious lateral velocity of the tail. That impassable orbital line is to it as solid and inflexible as a wall of adamant. The motion so instantly arrested would be disastrous to any tail, whether composed of gas, meteorites, or electricity, whatever that may be. Having shown that the common theory of comets is filled with insuperable difficulties, I will again call attention to a theory proposed about eighteen months ago in the SCIENTIFIC AMERICAN. According to this theory, a comet consists of a nucleus and an atmosphere, for the most part invisible, surrounding it on all sides to an extent at least equal to the length of the tail. The rays of the sun in passing through or near the nucleus are so modified as to become visible in their further progress through the cometic atmosphere, while all the rest remain invisible. What we call the tail is merely a radius of the cometic atmosphere made visible, and as the comet moves through space, only different portions of the atmosphere come in sight, in obedience to the ordinary laws of light. There is no difficulty in accounting for the rise and fall of the tail at perihelion, nor for the tail preceding the nucleus afterward. The spherical theory accounts easily for the different forms of tail seen in different comets. The sword shaped tails, at variance with the common theory, can be accounted for by supposing a slight difference in density or material in the cometic atmosphere, which will deflect the light as seen. The comet of 1823, which cannot be explained on the common theory, is very easily explained on the spherical. That comet showed two tails, apparently of equal length, which moved opposite to each other, and perpendicularly to the orbit of the nucleus, and showing no signs of repulsive force from the sun. On the spherical theory it is only necessary to suppose such an arrangement of the nucleus as would reflect the rays of the sun laterally; a slight modification of the nucleus would give not only two but any number of tails pointing in different directions. It may be objected to the spherical theory that a tail 50,000,000 miles long would call for a sphere 100,000,000 miles in diameter, and that would be too vast for our solar system. But it is claimed for this sphere that it consists of the same material as the so-called tail, and that it has the same capability of moving among planets without manifest disturbance to either.