September 1th, 1820. 35 



And in the present case, the difference is not greater than this 

 effect alone would seem to warrant. Since the most elevated 

 mountains on the earth scarce extend above 20,000 feet, and 

 the proportional diameters of the two bodies being 1 1 and 3 

 nearly, the highest of the lunar mountains should not exceed 

 the dimensions now given, allowing for the effects of gravity 

 and pther known causes. 



In measuring lunar elevations, by the method now proposed, 

 notwithstanding its simplicity, certain precautions will be found 

 requisite. Great attention is obviously necessary in selecting 

 for observation, such projections only as rise immediately from 

 the moon's circumference, and whose whole height, conse- 

 quently, is exposed on the solar disc. This may be ascertained 

 from the manner in which the base appears to unite with the 

 periphery, for if it fall on either side, the mountain will present 

 a truncated form, with defined angles at the points of apparent 

 junction; on the contrary, where the whole elevation is pro- 

 jected from the circumference, the base seems gradually to 

 expand, and to blend imperceptibly with the general mass. 

 Such, at least, were the appearances which usually accom- 

 panied certain degrees of elevation in the instance before us, 

 except in parts where the moon's edge seemed so marked with 

 slight undulations only. In the terminal elevations of ridges, 

 from their being, for the most part, abrupt, and precipitous, 

 these characteristics were by no means so conspicuous. This 

 circumstance, however, was pretty generally observed, that where 

 a ridge of any considerable altitude terminated, that portion of 

 the circumference, which lay contiguous to its base, presented, 

 for a considerable extent, the appearance of a smooth, un- 

 broken surface, seeming to indicate that such a ridge was 

 surrounded by a plain, and that its entire height was visible. 



With regard to the position of the mountain, that is the most 

 convenient, where the direction of gravity at its summit ap- 

 proaches nearest to parallelism, with the line described by the 

 path of the moon's centre. Attention, therefore, was prin- 

 cipally directed in the present case to such inequalities as were 

 situate near the diameter lying in this path. In proportion as 

 D2 



