gem, and even their locations correspond fairly to those of Eratosthenes' description. The slight discrepancies relate to the arms. There are no stars actually placed on the man's left arm; but since the whole forearm is concealed by the body of the snake, the lapidary has resorted to the naďve device of placing four stars in a close-set row just under the point where the left hand grasps the snake. These stars evidently belong to the left forearm, and three are to be seen on the right forearm. The Greek text gives three to the right, four to the left; but the interchange of the words right and left is required by the subsidiary sources.
Even in the snake the engraver has come fairly near agreement with the description by Eratosthenes, who seems to have assigned twenty-three stars to it. Nineteen are visible on the stone, but abrasions near the left margin may account for the loss of one or more on the creature's tail. The location of the stars, however, does not conform strictly to Eratosthenes' description.
When this stone viewed directly, the human figure is seen in three-quarter front view, turned slightly to the observer's left, facing the serpent's head, which is bent back towards him, and it is the man's right hand that grips the serpent's neck. This position, however, cannot be reconciled with other representations of Ophiuchus, or with the most complete ancient description, that of Ptolemy, whose words make it clear that the snake's neck was held by the man's left hand, and who proceeds in his description from right to left.17
Other evidence also indicates that Ophiuchus was imagined as standing facing to the spectator's right, with his left hand holding the snake's neck.18
This orientation is also accepted by modern astronomers. For these reasons we are obliged to assume that the London haematite was meant to be viewed in an impression, which would agree with Ptolemy's orientation of the figure. This is surprising, for the stones of this type apparently did not serve merely to attest the owner's interest in astronomy, but were designed as amulets; and as we have seen, with few exceptions amulets were meant to be looked at directly.
A different representation of Ophiuchus occurs on a broken haematite in the Cabinet des Médailles (D. 352).19
The fragment is the lower half of an ellipse, which contains the figure of Ophiuchus intact; it is idle to speculate about the design that occupied the lost upper half.20
The human figure is here seen in three-quarter back view, facing the observer's left and holding the snake's neck with his left hand, while his right grasps its tail. There are no stars either on the man or on the serpent, but there is a single large star in the field at the left. This may be the device commonly employed to show
17 Claud. Ptolem. Syntaxis Mathem. (ed. Heiberg), 7, pp. 70–72.
18 Comm. in Aratum (ed. Maass), §§ 83, 86, pp. 354–355.
19 Chabouillet 2184; P. J. Mariette, Traité des pierres gravées, II, 70 (Paris, 1750). Mariette's engraving shows the scorpion intact; the damage mentioned above seems to have occurred at a later date than that of his book, or else he was a keener observer than Chabouillet, at least in this instance, and instructed his draftsman about the proper restoration.
20 For an astronomical figure combined with another design see No. 1727 in Fossing, Antique Gems in the Thorvaldsen Museum; Eros above, bandaging a lion's paw, Capricorn below.