[16] In Plate II. of this work the chariot group is represented as facing transversely, in the Frontispiece and Plate III. as facing longitudinally to the building. It may be as well to mention here that I have introduced several such discrepancies into the plates, which are neither oversights nor errors. This is one; another is that, in Plate II., the lions at the angles of the pyramid are omitted, but inserted in the other three plates: a cymatium has been introduced as crowning the order of the base in one plate, and another moulding substituted in the others. The Monte Cavallo groups have been introduced in Plates I. and III. and omitted elsewhere. The object of these alterations is that, as these are mere suggestions, they are offered as such in order that the reader may exercise his own judgment regarding them. The dimensions, and all those parts which are certain, are repeated throughout; but, unless some further discoveries are made, there must always be some details which must be left to the taste or the knowledge of the restorer. [17] There is a discrepancy of three inches in this dimension, which must be apportioned somewhere. I fancy it is to be found in the cymatium gutter, but this could only be ascertained from a thorough re-examination of the fragments themselves. [18] See page 11 ante. [19] The mode of lighting Greek temples and Greek buildings generally has never fully been investigated by architects. I read a short paper on the subject at the Royal Institute of British Architects on the 18th of November last; and though that is an amplification of my remarks in the True Principles of Beauty in Art some fourteen years ago, it is far from exhausting the subject. But it is enough to prove that the mode of introducing light was as perfect and as beautiful as every other part and every other contrivance of Greek architecture. [20] Plates I. II. and III. [21] These stairs, indicated by dotted lines in the plan (Plate I.) being on one side, clearly indicate that the sepulchre was not symmetrically placed to occupy the centre of the building. Curiously enough, the Tomb at Mylassa (Woodcut No. 3) has a doorway placed unsymmetrically, for no reason that can be guessed, unless it were in imitation of its celebrated prototype. What also is curious is that at Mylassa a pillar stands directly over the centre of the doorway leading into the principal chamber of the tomb, exactly as occurs at Halicarnassus, and that chamber has a flat stone-roof, as here suggested, for the Mausoleum. [22] The ease with which the Knights got access to this tomb would entirely contradict the supposition of its being walled up, if it was the Tomb of Mausolus they reached. It may have been that of the Queen.