[16] Giving this cloud the average velocity of thirty miles per hour, its altitude was determined by the sextant at twelve miles, and we think under-estimated. While measuring, the author's attention was drawn to the fact, that although it appeared equally dense above and below, yet its middle part was the brightest, and as there was only a faint glimmer of twilight in the N.-W., he concluded that the cloud was self-luminous; for when the smallest stars were visible, it glowed about as bright as the milky-way in Sagittarius. Occasionally the whole cloud was lit up internally by the lightning, and about this time it sent off three rays: one horizontally, westward, which was the faintest; one about N.-W., towards Jupiter, and the brightest of the three; and another towards the north. These were not cirrus streaks, but veritable streams of electric matter, and had a very decided rotation from left to right, and continued visible about twenty minutes, as represented above. [17] This day the central vortex passed in about latitude 47° N.--the southern margin cannot be nearer than 250 miles, throwing off the 40′ for the horizontal refraction, would give eight miles of altitude above a tangential plane. Then another seven miles, for curvature, will give an altitude of fifteen miles for the cumuli. The height of these thunder-clouds has been much under-estimated. They seem to rise in unbroken folds to a height of ten and twelve miles frequently; from the data afforded by the theory, we believe they will be found much higher sometimes--even as much as sixteen miles. [18] These parallel bands, and bands lying east and west, are frequent in fine weather between two vortices. Sailors consider them a sign of settled weather. After dark there was frequently seen along the northern horizon flashes of lightning in a perfectly clear sky. But they were both faint and low, not reaching more than 4° or 5° above the horizon. After sunset there were very distinct rays proceeding from the sun, but they were shorter than on the evening of the 3d. These are caused by the tops of the great cumuli of the storm, when sunk below the horizon, intercepting the sun's rays, which still shine on the upper atmosphere. The gradation was very marked, and accorded with the different distances of the central vortex on the 3d and 4th--although, on the 4th, the nearest distance must have been over four hundred miles to the southern boundary of the storm.