238 Journal of Philology. * the opposite of this. I shall deal almost exclusively with the se- cond of those sources enumerated by Hermann the " testimonia litterarum." Given for instance the fact of Polychromy as borne out by traces of colour on temple and statue, my object would be to ascertain how far that fact admits of corroboration by like traces and allusions in the pages of authors. That such corroboration is needed to give to Polychromy a recognised place among the processes of ancient art, no thoughtful man will deny. Those, who have more especially addressed themselves to the consider- ation of this subject, have for the most part approached it, with theories cut and dried, and to suit those preconceived theories they have warped ambiguous texts, and made the very stones cry out in their support ! Thus it has come to pass, that while one man asserts, with all the emphasis of italics, that such and such a temple bears unequivocal traces of blue in frise, metope, or triglyph, another maintains with equal vehemence that the aforesaid frise is red as red can be : just as if the temple had been stared into blushes. To both these advocates (for advo- cates they are, not judges) of conflicting theories, it would be but reasonable to suggest the enquiry, how far these traces may be owing to the presence of certain substances in the stone, which, under the action of time or sea-air, may bring about, by a purely chemical process, effects liable to be mistaken for thoso of colour*. It is this absence of solid, well-ascertained facts, this contradiction in the accounts of travellers, which encourages me to confine myself almost exclusively, for the present, to the testimonia litterarum. As far as I know myself, I have no desire to shore up and underpin any crazy system or theory by desperate makeshifts and disingenuous arts. I shall take my statements as I find them, not packed, and would select for my motto the words of one of old : Nupr)v vUrju ttjv kcu uo kcu tw irpocrbuikeyouiva afiXufiia. &iXoTip.te<rdai napa Katpbv irpbs (piXovs Kai iv rais T)TJ(TC(Tt <pvXa<r<roiuT]v. Mtjkotc kcu napa rb iuoi boKtov iv rats ClTrj- creai tov vikclv tixKa tpiv acrnacralu^v. MiyKore rl /3Xd/3f/ rot) aXrjdfog kcu coot* reap a avrbs 018a irapaXoyicra<r6at rbv npoabuiXfyduepop ipicrcu cirap- Bfirjv. Tois rb dXrfdcs yots ^vvicrracriv act vuuaxoiur)u. Stob. Florileg. I. p. 39, Ed. Tauchn. Reserving for a future occasion such remarks as I may have
- I would not be understood, byjthis, causes have not sufficiently been taken
to deny the existence of bona fide traces into account, of colour. I merely suggest that other
