At Heliopolis, as I have hinted, the matter is still less certain. Almost every trace of the temple has disappeared, but of remains of temenos walls in 1844, when the site was studied by Lepsius, there were plenty. At Karnak, where both temples and temenos walls remain, we can see how closely the walls reflect the orientation of the included temples, even when they seem most liable to the suggestion of symmetrophobia. I have before stated that the Egyptians have been accused of hating every regular figure, and the irregular figures at Karnak are very remarkable; in the boundary walls of the temple of Amen-Rā there are two obtuse angles; round the Mut temple we also have walls, and there again this hatred of similarity seems to come out, for we have one obtuse and one acute angle. But if we examine the thing a little carefully, we find that there is a good deal of method in this apparent irregularity. The wall of the temple of Amen-Rā is parallel to the face of the temple or at right angles to its length. One wall of Mut is perfectly parallel to the face of the temple or at right angles to the sphinxes. And the reason that we do not get right angles at one end of the wall is that the walls of the temple at Mut are parallel to the chief wall of the temple of Amen-Rā. Surely it must be that, before these walls were built, it was understood that there was a combined worship; that they stood or fell together. One thing was not attempted in one temple and another thing in another, but the worship of each was reflected in the other. If this be true, there was no hatred of symmetry, but a definite and admirable reason why these walls should be built as they were. [Illustration: THE MOUNDS AND OBELISK AT ANNU.] With the knowledge we possess of both temenos walls and temples at Karnak, and of the, I may almost say, symbolism of the former, it is fair to conclude that when temples have gone we may yet get help from the walls. The walls at Heliopolis are the most extraordinary I have met with in Egypt, as may be gathered from the accompanying reduction of Lepsius' map. The arrow in Lepsius' plan is so wrongly placed that the plan is very misleading. It follows from Captain Lyons' observations and my own that the longest mound heads 14° N. of W. to 14° S. of E. within a degree; the condition of the mounds renders more accurate measures impossible.[9] It is to be gathered from the inscriptions that the temple within these mounds, now only represented by its solitary obelisk, was styled a sanctuary or temple of the sun.[10]