We find, beginning at the south, a large temple with a long line of sphinxes, the temple of Mut (X) facing the large temple of Amen-Rā (K). To the north of the latter is another temple system (A and B and C), also with an avenue of sphinxes. On the east side of K another temple (O) is only slightly indicated.

PLAN OF THE TEMPLES AT KARNAK (FROM LEPSIUS), FOLLOWING THEIR ORIENTATIONS.

To the south of the large temple K is another one—that of Khons (T), also with its sphinxes. Connected with K are two other temples, L, nearly, and M, exactly, at right angles to it. There is also such a rectangular temple (Y) added to the temple of Mut. I also call attention to the temples V and W, chiefly to point out that when I went over the ground with M. Bouriant it seemed to us as if the temple V faced S.E. and not N.W. as indicated by Lepsius. Very few traces of the temple are left.

Since the labours of the French and Prussian Governments gave full records of Karnak a memoir on the temples has been published by Mariette, which gives us not only plans, but precious information relating to the periods at which, and the kings by whom, the various parts of the temples were constructed or modified. No doubt those which are still traceable form only a very small portion of those which once existed; but however that may be, I have now only to call attention to some among them.

I have previously shown that the magnificent work of Mariette has supplied us with building dates for the solar temple to which reference has been made; so that we have, with more or less accuracy, the sequence of the various parts of the completed building.

If we consider the plan without any reference to the building dates at all, the idea that the smaller temples were built for observations of stars seems to be entirely discountenanced. The temple L, for instance, instead of having a clear horizon, is blocked by the very solid wall (2) and its accompanying columns; the temple M, instead of having a clear horizon, is absolutely blocked by two of the line of pillars (1) very carefully built in front of it. But if we consult Mariette, we find in both cases that the wall was built long after one temple, and the pillars were built long after the other.

This result is satisfactory, inasmuch as it indicates that a natural objection to the orientation hypothesis is invalid. But can we strengthen it by supporting Mariette's statement as to the dates?

Mariette states that the temple M was built by Rameses III., a king of the twentieth dynasty. With this datum, we consider the orientation of the temple. The problem is one of this kind:—Taking the Egyptologist's date for Rameses III. at 1200 B.C., and taking the amplitude of the temple as 63½° N. of E., was there, when that temple was built, any star opposite to it, any star to which it accurately pointed? We can translate the amplitude of that temple into the declination of a star, making a slight correction for the stated conditions of observation in Egypt, which would make the apparent amplitude less than the true one, because the star would appear to rise more to the south. In the absence of precise information, we are justified in taking the mean of the values referred to by Biot—that is, an apparent amplitude due to a stratum of haze 1½° high, especially as the temple looked away from the Nile.

Searching the astronomical tables, we find that there was a star visible along the temple axis. The star was γ Draconis.