If we refer to the plan of the temple, the point of importance to us in our present inquiry has relation to the circumstances connected with the buildings of the temple itself. We have in the outer court to the north-west certain pillars which were built by one of the Ethiopian kings. These I mark =1=, =1= (see page 118). There is the temple =M=, built by Rameses III, according to Mariette. There are walls with columns, marked =2=, =2=, built by the twenty-second dynasty, north and south of this outer court; and then there is the temple =L= in the outer court, supposed to have been built by Seti II. The western part of the temple, therefore, is of no high antiquity. To find this we have to go some 200 yards to the south-east. Near the central portion of the temple (marked =4=) there are traces of the twelfth or possibly the eleventh dynasty. What existed then might have been a shrine with nothing to the north-west or south-east of it. This seems almost to have been its condition at the time of Thothmes III. even. According to an inscription quoted by Brugsch,[36] "The king (Thothmes III.) found it in the form of a brick building, in a very dilapidated condition, being a work of his predecessors. The king with his own hand performed the solemn laying of the foundation stone for this monument." From this point, indeed, the temple seems to have extended in both directions--that is, north-west and south-east--the sanctuary being thrown back to the eastward and pylons added to the westward. It follows from the above very brief sketch that the original orientation of the original shrine is to be gathered from the walls towards the centre of the present ruins. Let us agree to this. The Egyptologist already gives us eleventh-dynasty time, say 2500 B.C. for a part of the existing temple. Let us now pass to the astronomical problem. Lepsius and others have measured the amplitude of this part of the temple. It is given as 20° or 20° 30′ N. of W. When there I measured the height of the opposite hills (near the tombs of the kings) roughly at 2½°. If we, therefore, deal with the amplitude, considering the height of the hills as 2½°, we find that, as the horizon was above the sea horizon and the sun travels down an inclined path from south to north, it would meet the hill sooner than the sea horizon; the apparent amplitude would, therefore, be less than the true one, so that we get an amplitude of 25° instead of 26°, and if we correct that for refraction we get 25½°. Let us take the lower amplitudes. We can construct the following table:-- With present obliquity 23° 30′ we have at Thebes, lat. 25° 40′ amplitude on horizon (sun's centre) 26° Corrected for refraction 26° 30′ The amplitude behind hill, 2½° high, will be 25° Making correction for refraction 25° 5′ So that, taking the lowest amplitude, the temple axis points almost 1° too much to the north.