From the examples of the use of the ekkyklema which have just been cited we may gather some further particulars as to its character and construction. It appears that persons upon the ekkyklema could easily descend to the stage, and that persons on the stage could easily touch those on the ekkyklema. It follows that it must have been a low platform, not much above the level of the stage. As regards its length and breadth, it was evidently large enough to support several persons. At the same time it cannot have been of any very great size. Its width must have been less than the width of the doors in the background, to permit of its being rolled through them. Its depth cannot have been very great, because of the narrowness of the Greek stage. In the Acharnians, when Euripides is rolled out, he is represented as still sitting in his room upstairs. But it is unlikely, as some suppose, that in this case the platform was made taller than usual, to produce the effect of an upper story. As Euripides has to hand various articles to Dicaeopolis, who is standing on the stage, there cannot have been much difference of level between the two. The exact mechanism of the ekkyklema, however, remains uncertain. It is practically undisputed that the grooves or rails found at Eretria, running on to the later stage straight from its back-scene were intended for some such contrivance to run on.[626] On the other hand it has been argued from the use of certain words in the scholiast’s descriptions that the ekkyklema must have revolved on a pivot,[627] and it has been suggested that the mechanism was like that of which a diagram is given in the accompanying figure, where _ss_ is the stage, _ww_ the back-scene, _a_ shows the ekkyklema at rest and not in use, _b_ shows it in process of being rolled round for use, _c_ shows it after being rolled out. This, however, finds no confirmation in anything in the ruins; the straight rails at Eretria are against it, and the words referred to may be explained by the use of a windlass or similar mechanism used in rolling out the ekkyklema. Judging from the width of the rails at Eretria, the width of the ekkyklema may have been about ten feet, and the doors must therefore have been rather larger. The suggested revolving ekkyklema might afford more standing room, but there is not sufficient evidence of its existence. [Illustration: FIG. 14 A.]