806 PHILOSOPHICAL TRANSACTIONS. [aNNO I687. 



and when she is in the meridian, and at her greatest northern latitude, and 

 consequently her utmost elevation in our horizon. It is also well known, that 

 when she is in this situation, and viewed by the naked eye, she appears to be 

 about a foot broad. But the same moon being viewed just as she rises, appears 

 to be 3 or 4 feet broad ; and yet, if with an instrument we take her diameter, 

 both in one position and the other, we find that still she is only 30 minutes. 



The celebrated Descartes attributes this appearance rather to a deceived judg- 

 ment than to any natural affection of the organ or medium of sense ; for the 

 moon, says he, being near the horizon, we have a better opportunity and ad- 

 vantage of making an estimate of her, by comparing her with the various 

 objects that effect the sight, in its way towards her ; so that though we imagine 

 she looks larger, yet it is a mere deception : for we only think so, because she 

 seems nearer the tops of the trees, or chimneys, or houses, or a space of ground, 

 to which we can compare and estimate her by ; but when we bring her to the 

 test of an instrument, then we find our estimate wrong, and our senses de- 

 ceived. These thoughts seem much below the usual accuracy of Descartes ; 

 for if it be so, we may at any time increase the apparent size of the moon, 

 though in the meridian ; for it would be only getting behind a cluster of 

 chimneys, a ridge of hills, or the tops of houses, and comparing her to them 

 in that position, as well as in the horizon : besides, if the moon be viewed just 

 as she is rising from the horizon determined by a smooth sea, and which has no 

 more variety of objects to compare her to, than the pure air ; yet she then seems 

 large, as if viewed over the rugged top of an uneven town, or rocky country. 

 Besides, all variety of adjoining objects may be taken off, by looking through 

 an empty tube, and yet the deluded imagination is not at all helped thereby. 



I come next to the solution given by the famous Thomas Hobbs, which is as 

 follows. In fig. 4, pi. 10, let the point G be the centre of the earth, and F 

 the eye on its surface ; on the same centre G, describe the two arches, the one 

 E H determining the atmosphere, and the other A D to represent that azure 

 surface in which we imagine the fixed stars ; and let FD be the horizon. Di- 

 vide the arch AD into three equal parts, by the lines BF, CF. Then it is 

 manifest that the angle A F B is greater than the angle B F C, and this again 

 greater than the angle C F D. Wherefore, says he, to make the angle C F D 

 equal, to the angle C F B, the arch C D must be greater than the arch C B ; and 

 consequently, that the moon may in the horizon appear under the same angle 

 as when elevated, she must cover a greater arch, and therefore seem greater ; 

 that is, the moon in the meridian appearing under the angle B F C, that she 

 may appear under an equal angle in the horizon, as suppose CFD, it is neces- 

 'sary that the arch CD be greater than CB; and consequently, though she 



