HERODOTUS.--This far less brilliant but equally interesting object is about 23 miles in diameter, and is not so regular in shape as Aristarchus. Its W. wall rises at one point more than 4000 feet above the very dusky floor. Except on the S.W. and N.E., the border is devoid of detail. On the S.W. three little notches may be detected on its summit, which probably represent small craters, while on the opposite side, on the inner slope, a little below the crest, is a large crater, easily seen. Both the E. and W. sections of the wall are prolonged towards the S. far beyond the limits of the formation. These rocky masses, with an intermediate wall, are very conspicuous under oblique illumination, that on the S.W. being especially brilliant. On the N. there is a gap through which the well-known serpentine cleft passes on to the floor. Between the N.W. side of Herodotus and Aristarchus is a large plateau, seen to the best advantage when the morning terminator lies a little distance E. of the former. It is traversed by a T-shaped cleft which communicates with the great serpentine cleft and extends towards the S. end of Aristarchus, till it meets a second cleft (forming the upper part of the T) running from the S.E. side of this formation along the W. side of Herodotus. The great serpentine cleft, discovered by Schroter, October 7, 1787, is in many respects the most interesting object of its class. It commences at the N. end of a short wide valley, traversing mountains some distance N.E. of Herodotus, as a comparatively delicate cleft. After following a somewhat irregular course towards the N.W. for about 50 miles, and becoming gradually wider and deeper, it makes a sudden turn and runs for about 10 miles in a S.W. direction. It then changes its course as abruptly to the N.W. again for 3 or 4 miles, once more turns to the S.W., and, as a much coarser chasm, maintains this direction for about 20 miles, till it reaches the S.E. edge of a great mountain plateau N. of Aristarchus, when it swerves slightly towards the S., becoming wider and wider, up to a place a few miles N. of Herodotus, where it expands into a broad valley; and then, somewhat suddenly contracting in width, and becoming less coarse, enters the ring-plain through a gap in the N. wall, as before mentioned. I always find that portion of the valley in the neighbourhood of Herodotus more or less indistinct, though it is broad and deep. This part of it, unless it is observed at a late stage of sunrise, is obscured by the shadow of the mountains on the border of the plateau. Gruithuisen suspected a cleft crossing the region embraced by the serpentine valley, forming a connection between its coarse southern extremity and the long straight section. This has been often searched for, but never found. It may exist, nevertheless, for in many instances Gruithuisen's discoveries, though for a long time discredited, have been confirmed. The mountain plateau N. of Aristarchus deserves careful