Troy. 167 direction, remains of houses more or less roughly built, fragments of prehistoric vases, weapons, and stone implements, heaped up in prodigious masses, forming beds that did not always retain their regular course, but sank into one another in a very perplex- ing sort of way. To have uncovered the walls along the whole line would have entailed removing the accumulated masses of sherds as they were uncovered one after another ; but adoption of a system involving complete destruction of all the upper layers along with the buildings they enclosed, would not only have been a most serious undertaking but a very undesirable one, in that future explorers would for ever have been deprived of the means of comparison, always useful, nay, sometimes quite indis- pensable for a full comprehension of ulterior discoveries. Plans and tracings, however carefully made, wear a very different aspect on paper from the objects themselves as seen in situ. Besides, so slow a method would have ill suited Schliemann's sanguine temperament, eager to reach the mound's centre and extrude its secret therefrom. In 1872 he cut a trench cir. twelve metres broad, and reached the virgin rock sixteen metres from the summit. This enormous breach, with direction from north to south, is still extant ; like a main thoroughfare, it traverses the hill from end to end, whence branch off minor streets, later trenches that run out at a higher level in narrower and shallower channels (Fig. 38 ; trench intervening between X and z). In the accumulations of materials which he displaced, Schlie- mann at that time thought to recognize seven periods, answering to what he called seven superimposed cities buried within the artificial mound.^ These arbitrary subdivisions were not corro- borated by later researches carried on by trained archaeologists and architects. In their opinion the strata, which on account of the character of the objects found in them admit of being singled out, are reducible to four. Resting on the rock itself is the earliest or first settlement, and over it what Schliemann ^ Schliemann's writings published before 1880, except that they enable us to trace the modification which his ideas underwent, have lost their interest for the general reader. They are moreover set at naught by his liios^ in which he has set forth the conclusions to which he was led by his critics and his own excavations of 1882, when Dorpfeld was already with him. Hence in this book reference is made throughout to liioSy whilst correcting and completing the evidence it affords with the help of the researches undertaken in 1889 and 1890.
