These, in later times, predominated more and more, till in the fourth and last style, the only irregularity is found in the unequal thickness of the stones of the same course, corrected sometimes by the introduction of a sloping line, or more often by a notch to let the larger stone into the course above or below. Though I believe this to have been the general progress of the art, yet you must not imagine them as distinctly characterizing different periods; on the contrary, there is hardly any considerable wall of Cyclopean masonry, which does not exhibit in different parts, two of these methods; and sometimes three are found, without any appearance that they have been restorations of different periods; we may however observe, that the second style is most common in Latium, the fourth in Tuscany; the third is perhaps about equally diffused in both countries. At all times, these blocks were used without cement, and all that I have hitherto seen, are mere terrace walls against a hill, and exhibiting in consequence one face only; but I am told of instances where both sides are seen, and that in such cases two walls are built back to back, without any attention to the regularity, or evenness of what was to be the internal part, and without any filling in. No arches, that is, no system of wedges mutually supporting each other, is to be found, though such an arrangement would seem to grow more easily out of these inclined lines, than from regular courses of stones; but where there are openings, (of which I have seen none hitherto) there is a very large stone, worked square, and laid horizontally to cover it; and in one instance, at Arpino (perhaps because the builders could not meet with a stone large enough to cover the opening,) the size of the aperture is reduced by advancing courses, into the form of a pointed arch. There is indeed a real arch at Fiesole, which by some has been supposed to be part of the Cyclopean construction, but both the arch and the fragment on which it rests are obviously of a date much posterior. There are many remains of Cyclopean walls both at Tivoli and Palestrina, and as according to Virgil, Tibur and Præneste were founded about the time that Æneas landed in Italy; this epoch has been assigned to their construction, but it must be confessed, that the argument is not altogether conclusive. It is held essential to Cyclopean walls, that there should be no cement, and _à fortiori_, no rubble-work employed in their composition; but in this neighbourhood, at what is called the villa of Brutus, which I shall shortly mention to you; there is a wall of Cyclopean masonry, resting for its whole length, and apparently backed in its whole extent by a wall of rubble. This Cyclopean wall has been faced by another of _opus reticulatum_, so common in the time of Augustus, and in that of the first emperors, and which may be seen in almost all the villas about Tivoli. It seems that the Romans did not like the appearance of these