Thirdly, we have the actual remains of Greek sculpture, a constantly accumulating store, yet only an insignificant remnant of what once existed.  These works have suffered sad disfigurement.  Not one life-sized figure has reached us absolutely intact; but few have escaped serious mutilation.  Most of those found before the beginning of this century, and some of those found since, have been subjected to a process known as “restoration.”  Missing parts have been supplied, often in the most arbitrary and tasteless manner, and injured surfaces, e. g., of faces, have been polished, with irreparable damage as the result.

Again, it is important to recognize that the creations of Greek sculpture which have been preserved to us are partly original Greek works, partly copies executed in Roman times from Greek originals.  Originals, and especially important originals, are scarce.  The statues of gold and ivory have left not a vestige behind.  Those of bronze, once numbered by thousands, went long ago, with few exceptions, into the melting-pot.  Even sculptures in marble, though the material was less valuable, have been thrown into the lime-kiln or used as building stone or wantonly mutilated or ruined by neglect.  There does not exist to-day a single certified original work by any one of the six greatest sculptors of Greece, except the Hermes of Praxiteles (see page 221).  Copies are more plentiful.  As nowadays many museums and private houses have on their walls copies of paintings by the “old masters,” so, and far more usually, the public and private buildings of imperial Rome and of many of the cities under her sway were adorned with copies of famous works by the sculptors of ancient Greece.  Any piece of sculpture might thus be multiplied indefinitely; and so it happens that we often possess several copies, or even some dozens of copies, of one and the same original.  Most of the masterpieces of Greek sculpture which are known to us at all are known only in this way.

The question therefore arises, How far are these copies to be trusted?  It is impossible to answer in general terms.  The instances are very few where we possess at once the original and a copy.  The best case of the kind is afforded by Fig. 75, compared with Fig. 132.  Here the head, fore-arms, and feet of the copy are modern and consequently do not enter into consideration.  Limiting one’s attention to the antique parts of the figure, one sees that it is a tolerably close, and yet a hard and lifeless, imitation of the original.  This gives us some measure of the degree of fidelity we may expect in favorable cases.  Generally speaking, we have to form our estimate of the faithfulness of a copy by the quality of its workmanship and by a comparison of it with other copies, where such exist.  Often we find two or more copies agreeing with one another as closely as possible.  This shows—­and the conclusion is confirmed by other evidence—­that means existed in Roman times of reproducing statues