But this is not all. In the avenue, but not in the centre of its width, there is a stone called the "Friar's Heel," so located in relation to the horizon that, according to Mr. Flinders Petrie,[15] who has made careful measurements of the whole structure, it aligned the coming sunrise from a point behind the naos or trilithon. The horizon is invisible at the entrance of the circle, the peak of the heel rising far above it; from behind the circles the peak is below the horizon. Now, from considerations which I shall state at length further on, Mr. Petrie concludes that Stonehenge existed 2000 B. C. It must not be forgotten that structures more or less similar to Stonehenge are found along a line from the east on both sides of the Mediterranean.[16] [Illustration: STONEHENGE, FROM THE NORTH.] It will be seen that the use of the marking stone to indicate the direction in which the sun will rise answers exactly the same purpose as the long avenue of majestic columns and pylons in the Egyptian temples. In both cases we had a means of determining the commencement and the succession of years. [Illustration: STONEHENGE RESTORED.] Hence, just as surely as the temple of Karnak once pointed to the sun _setting_ at the summer solstice, the temple at Stonehenge pointed nearly to the sun _rising_ at the summer solstice. Stonehenge, there is little doubt, was so constructed that at sunrise at the same solstice the shadow of one stone fell exactly on the stone in the centre; that observation indicated to the priests that the New Year had begun, and possibly also fires were lighted to flash the news through the country. And in this way it is possible that we have the ultimate origin of the midsummer fires, which have been referred to by so many authors.[17] We have thus considered solstitial temples scattered widely over the earth's surface far from the Nile Valley. We may now return to the equinoctial temples which can still be traced to the N.E. of that valley--the chief ones being those, remains of which still exist at Jerusalem, Baalbek, and Palmyra, where stone was available for the temple builders. These temples were apparently as perfectly squared to the equinox as the Pyramids at Gîzeh. I will take the temple of Jerusalem first, as its history is more complete than that of the others. We learn from the works of Josephus that as early as Solomon's time the temple at Jerusalem was oriented to the east with care;[18] in other words, the temple at Jerusalem was parallel to the temple of Isis at the Pyramids; it was open to the east, closed absolutely to the west. In plan, as we shall see, it was very like an Egyptian temple, the light from the sun at the equinox being free to come along an open passage, and to get at last into the Holy of Holies. We find that the direction of the axis of the temple shows the existence of a cult connected with the possibility of seeing the sun rise at either the spring or the autumn equinox.