Page:History of Art in Sardinia, Judæa, Syria and Asia Minor Vol 1.djvu/357

 Domestic Architecture. 327 to the east of those in the down-stream canal. The shape of the cutting at the point d evidences complete change of axis, and that the excavator, after leaving the false headings a, b, c, began to widen the tunnel on his right. The small buttress thus left reveals the method of excavation of the channel (Fig. 219, a, 6, c, d.) It may be considered certain that the place of junction was at the point e (Fig. 219) where, at the section and cross section, a sudden difference of level in the roof of the tunnel may be seen in plan (Fig. 219) at trans- verse section A B (Fig. 221). What confirms this view, says Conder, from whom we have borrowed, is that by examining the walls of the canal, it will be seen that south of the line A B the excavation points up-stream, proved by the sidings, whilst north of the same line, the set backs point down channel. It was at point e, there- fore, that the two gangs working separately at the lower and upper tunnel, were able to shake hands through the aperture made in the wall 1 m. 60 high. There was a little differ- ence of level (f of a cubit) between the twin channels at the place of junction, which had to be smoothed away before the water could flow in easily. On either side of this point, too, the roof becomes so low that, in order to proceed, it is necessary to crawl on all fours. The general impression left from examination of the tunnel is that it was the work of a people whose engineering was in its Fig. 219. — Plan lof Central Portion of Canal. Statements, 1882, p. 123. Quarterly