The question of the origin of karst poljes still remains open despite certain advances achieved during recent decades. Observations in the western part of the middle Taurus between Lake Beysehir and the Mediterranean coast near Manavgat, especially in the neighbourhood of Akseki, and also west and north of Antalya in the western Taurus, have shown that there the real reason for the formation of karst plains is to be found in the deposition of insoluble debris in concave relief forms within the limestone area. The source of these alluvial deposits are the sandy and marly rocks which were incorporated into the limestone massif by the folding process. The polje basins are always situated in vales which were once sections of a genuine valley network and which were formed before the fractures in the limestone had become active in the karst hydrography (as understood by Otto Lehmann). In many cases, but not always, major guiding lines of this valley mesh are tectonically determined. After genuine karst formation has set in (i. e. underground drainage), and when the drainage has ceased to run in one direction, closed circumference karst hollows with nonsoluble rocks in their catchment area function as vessels for the deposition of the waste material of these rocks. As long as the infilling continues the deposits take on a fan shape. These fans make the limestone basic layer impermeable so that the water of the streams causing the deposit disappears at the lower end of the fans into the cracks in the limestone, where then pronounced karst erosion takes place. The limestone slopes forming the frame recede with a cliff-like lower section. The fan surface, i. e. the polje surface, occupies the area where the limestone mountain frame was pushed back by corrosion, and the alluvial plain stretches nearly horizontally across the remaining limestone strata. These may dip in any direction. lt is thus that a karst pediment surface, thinly covered with alluvial material, is originated. The conical fan shape of the polje bottom can only be preserved while insoluble deposits are brought into the polje. Once an appreciable supply of such material ceases, the polje bottom gradually assumes an almost plain or gently undulating surface as is the case for instance in many of the !arge poljes of the Dinaric karst. Polje formation is thus neither linked with a karst water table nor, as a rule, with the sea level. The formation is initiated and continued by local causes, as for example the occurrence of water transportable impermeable rocks within the limestone region or at its margins, and the existence within the limestone region of local areas of deposition for the waste products of such rocks. Generally speaking polje formation cannot therefore be considered as a definite stage in a karst erosion cycle, nor can it be linked with peneplain formation as understood by W. M. Davis. It appears that the formation of enclosed poljes is a peculiarity of the periodically dry sub-tropics and the middle latitudes, whereas in the humid and intermittently humid tropics, as well as in the permanently humid sub-tropics with hot summers, formation of karst plains usually occurs in relation to the level of surface streams.