[887] This was the general opinion of antiquity, and was reproduced by Eratosthenes, Hipparchus, Ptolemy, and others; in short, by all the Alexandrine school. At the time of Eratosthenes, the obliquity of the ecliptic was 23° 45’ 17”. Syene was therefore 20’ 6” from being exactly under the tropic; for 24° 5’ 23” (latitude of Syene)—23° 45’ 17” = 20’ 6”. This would be the distance of the centre of the sun from the zenith of Syene; whence it follows that the northern limb of the sun was about 5’ from it. In the time of Strabo, the obliquity was only 23° 42’ 22”; the difference between the zenith of Syene and the northern limb of the sun was about 8’. Lastly, about 140 of the vulgar era, the obliquity was reduced to 23° 41’ 7”. Syene was then 24’ 16” from the tropic, and its zenith was about 10’ from the northern limb of the sun; when the shadows of gnomons of any tolerable size must have been perceptible, and Syene could not have been any longer considered as lying under the tropic. As regards the well which served to ascertain the instant of the solstice, Pliny and Arrian both mention it. The formation of it no doubt belonged to a very remote period. In the time of Strabo, the rays of the sun could not have reached entirely to the bottom, but the shadow was so small that it was not sufficient to shake the ancient opinion. In fact, the angle being about 8’, and supposing the depth to have been 50 feet, the northern side would have projected a shadow of about 18 lines; the rest would have remained in full light, and the reflexion would have caused the whole circumference of the well to appear illuminated. _Letronne._ [888] Kramer considers the passage between brackets to be an interpolation, as the same sense is conveyed in the passage which immediately follows. [889] The number here given is nearly twice too great. Kramer quotes G. Parthey (de Philis insula) for correcting the error to 50 stadia, and for perceiving that it arose from the very frequent substitution in manuscripts of the letter P (100) for N (50). [890] Unhewn stones, with a head of Mercury upon them. [891] Herod. ii. 28, who, however, seems to doubt the veracity of his informant. [892] Above, § 8. [893] B. C. 28. [894] B. xvi. c. 4, § 23. [895] The modern hamlet of Dakkeh occupies a portion of the site of ancient Pselchis. [896] Called Primis by Ptolemy and Pliny. It is placed by the former beyond Napata, and just above Meroë. Hence it is identified with Ibrim.