Decoration. 529 Mycenian border {Fig. 234). Here, as in the Cretan sarcophagi (Figs. 167, 169). the rosettes are enframed by two of those undulating lines so much affected by the native decorator. No one will dispute the resemblance which the latter ornament bears to the corresponding lines enframing the rosettes of the Mycenian doorway (Fig. 234). These same waving lines are separated in Fig. 235. It is just possible that Oriental models may have suggested the notion of making the rosette do duty as border ; but where we 6nd it so utilized it has become so intimately allied with the surrounding forms, and in general with the system of Mycenian decoration, as to forbid us to seek here a Fm. 136. — Plaster finfimeiii. foreign element ; we hesitate to pronounce decidedly on a probable and direct borrowing. In the rosette, we have a flower detached from its stalk and i-solatcd from the plant that fed it. In such conditions it was impossible to prevent its assuming a more or less conventional shape. This applies in full to a chaplet, from a wall-painting discovered at Tiryns, composed of alternate blue and yellow leaves (Fig. 213). Nevertheless, we must remember that as far back as prehistoric Thera, the painter tried his hand at a realistic rendering of irises (Fig. 208), whilst from Tiryns comes a branch which seems to bend with the breeze (Fig. 236). The colours beheld about the forms, whether of flowers or VOL. I. M M
