the arch which we see in Greece; but here the strivings were rewarded with success at an early time. The attempt succeeded; the perfect arch was lighted on, and the historical architecture of Rome was developed from the principle of the arch. Thus, while Fæsulæ, Tusculum, Signia, a crowd of others have their Greek parallels, there is no Greek parallel to the _cloaca maxima_ of Rome. Then, again, as we have already hinted, these examples show that the pointed arch, simply as a constructive form, is as old as the round. Because the pointed arch happened to become the leading feature of an architectural style later than the round arch, we are apt to fancy that the form is later in its own nature, that it must have been developed out of the round, that he who built the first pointed arch must have seen round arches. Yet the pointed form is just as natural in itself, just as likely to occur to a primitive builder. Indeed we might almost say that it was more likely. The first step towards the arch would doubtless be setting two stones to lean against one another, and this would lead much more easily to the pointed arch than to the round. It so happened that the first Italian builders whose strivings after the arch were quite successful were led to the round and not to the pointed form. But had the Tusculan or the Tirynthian engineer actually reached the construction to which he came so near, an architectural style, with the pointed arch for its great constructive feature, might have arisen in Latium or Argolis a thousand years before it actually did arise under Saracenic hands. Again, in considering these matters, we must carefully keep ourselves back from any tempting ethnological theories, above all from such ethnological theories as lurk in the dangerous word Pelasgian. No one doubts the near connexion of the old Italian and the old Greek races, a connexion nearer than that of common Aryan origin. But the same kind of analogies which may be seen in their earlier buildings may be seen also in the early buildings of races which are much further apart. If Tiryns finds its best parallel at Tusculum, Mykênê finds its best parallel at New Granga. Nearly just the same strivings after the arch may be found in more than one land altogether beyond the pale of European or Aryan fellowship, as for instance in the ruined cities of Central America. The analogies in the primæval architecture of remote nations exactly answer to the analogies in their weapons, dress, and customs. They belong to the domain of Mr. Tylor.