In his palace decorations the Assyrian artist set himself to free his figures from all unnecessary surroundings and to simplify his theme as much as he could. But we must make a distinction between those reliefs that may be called historical, such as the pictures of battles and sieges, and those in which the king is shown in the accomplishment of some duty belonging to his position, and part of his daily or periodical routine. It is to the latter class that the most carefully-executed works belong. In these no particular locality is specified; like that of the Panathenaic procession, it is left undetermined, and the mind of the spectator is silently invited to fill it in for himself. Those who frequented the palace were accustomed to see the king upon his throne, or traversing the wide quadrangle, or pouring libations on the altar that stood in front of the temple; so that they had no difficulty in imagining all that the sculptor had left unsaid. In the hunting pictures the same method was followed with but little modification. A flat surface suggesting the unbroken expanse of the desert, was the only indication of a _locus in quo_. [Illustration: FIG. 113.--Assurnazirpal offering a libation. Height 7 feet 8 inches. British Museum. Drawn by Saint-Elme Gautier.] It would have been difficult, or rather impossible, to adhere to such a rule in those reliefs in which the actual incidents of military expeditions were retraced. In them the sculptor thought it necessary to insert such details as would permit the various episodes commemorated to be identified. One of the simplest means of insuring the desired result was to render not only buildings, such as castles and fortified towns, but also the natural features of the scene, with the greatest possible truth. This the Assyrian artist did, as a rule, with excellent judgment. Thus, if an action or campaign had been fought in a mountainous country, he made use of a kind of lattice-work or reticulation, which every spectator thoroughly understood (see Vol. I. Figs. 39 and 43); if among forests, he introduced numerous trees among his figures. He made little attempt to distinguish between one kind of tree and another, but in most cases employed forms as conventional as that by which he indicated hills (Fig. 114). [Illustration: FIG. 114.--Tree on a river bank; from Layard.]