It is, indeed, possible that the present structure may have had other capabilities, such as being connected with the May year, the equinoxes or the winter solstice; but it is with its uses at the summer solstice alone that we now deal. There is a difference in treatment between the observations required for Stonehenge and those which are available for Greek or Egyptian solar temples. In the case of the latter, the effect of the precession of the equinoxes upon the stars, which as warning clock stars were almost invariably connected with those temples, offers the best measure of the dates of foundation; but in Britain, owing to the brightness of the dawn at the summer solstice, such a star could not have been employed, so that we can rely only on the secular change of the obliquity as affecting the azimuth of the point of sunrise. This requires the measurements to be taken with very great precision, and as the azimuth of the place of sunrise varies with the latitude, and as a datum point on the horizon in a known position was also required, Colonel Johnston, R.E., the Director-General of the Ordnance Survey, was asked for and obligingly supplied the following particulars: { Lat. 51° 10′ 42″ Centre of stone circle, Stonehenge { Long. W. 1 49 99 { Lat. 51° 3′ 52″ Centre of spire, Salisbury Cathedral { Long. 1 47 45 The real point was to determine the direction of the so-called avenue. Measurements taken from the line of the bottom of the ditch assisted materially those taken from the crown of the bank itself. With this help and by using the southern bank and ditch whenever it admitted of recognition, a fair estimate of the central line could be arrived at. To verify this, two pegs were placed at points 140 feet apart along the line near the commencement of the avenue, and four others at distances averaging 100 feet apart nearer the further recognisable extremity, and their directions were measured with the theodolite, independently by two observers, the reference point being Salisbury Spire, of which the exact bearing had been communicated by Colonel Johnston. This bearing was also measured locally by observations of the Sun and of Polaris, the mean of which differed by less than 20″ from the Ordnance value. The resulting observations gave for the axis of the avenue nearest the commencement an azimuth of 49° 38′ 48″, and for that of the more distant part 49° 32′ 54″. The mean of these two lines drawn from the central interval of the great trilithon, already referred to, passes between two of the sarsens of the exterior circle, which have an opening of about 4 feet, within a few inches of their middle point, the deviation being northwards. This may be considered to prove the close coincidence of the original axis of the temple with the direction of the avenue.