The Pyramids and Temples of Gizeh by Sir Flinders Petrie - Outside of Third pyramid

The Pyramids and Temples of Gizeh by W. M. Flinders Petrie

Chap. 10. Outside of Third pyramid Pages 110 - 116

79 . p 110. The Third Pyramid has never been quite finished. Its granite casing blocks are left in probably the same condition that they were sent from Assonan, with their outer faces in the rough,* but smoothly dressed down on the joint surfaces.

* One writer has described them as "rusticated," as if the roughness was a prepared feature; and another has attributed all the rounded irregularity of the stones to their weathering away since tbey were built. To say nothing, however, of innumerable cut holes in the outer surface, left for lifting the blocks, no weathering would add to a stone a part above its original face (see Plate. xii.). I had the pleasure of showing these details to an Engineer officer, experienced in Indian granite works, and he perfectly concurred as to the method of leaving an excess of stone on the face, to prevent injury to the block in transit.

These surfaces of contact are flat dressed, with a slightly projecting line of polished stone round the edges of the faces ; thus the stones would be in contact along the outer surfaces of the joints, though there was cement between the stones on the face in general. This line of polish is well defined on its outer edge, the stone being quite rough outside it, and sinking away sharply from it. This is important for estimating the intended plane of finish. The limestone casing which covered the upper part of the Pyramid was finished off like that of the other Pyramids; as may be seen by the worked faces, in the heaps of chips left by the Arab destroyers. But the pavement seems most probably never to have been placed around the Pyramid Perring found nothing but a substructure of rough megalithic blocks, with wide joints, and concluded that it was to be covered with finer work. On uncovering the granite casing, not only did I find no paving there, but the casing foot is quite in the rough, so that no pavement could be fitted to it; and none underlies it, as the granite rests on rough limestone.
The question then is, whether the casing was to be finished like that of the Second Pyramid, with a vertical foot, and the square-cut paving fitted against it; or whether any other plan was to be followed? The evidence seems rather in favour of a formation like that of the Second Pyramid. First, the lower course is thicker than any other, being 4 to 8 inches thicker than those p 111 just above it. Secondly, the evidence of the stones in the rough shows that their slope could not be continued down to their base at the N.N.E. the face of the bottom course is somewhat smoothed, though not finally dressed, and it ends with a rounded fall at 8 inches above the rough pavement, the granite descending also 9 inches below the limestone. At the E.N.E. the line of finish of the side joints runs straight at 51 down the side of the block, but ends abruptly at an irregular outer surface some inches above the base. Similar rough terminations of the intended slope are seen at the E.S.E. and S.S.E. the abrupt end being 6 to 18 inches above the limestone outside of the casing. Thus, in four out of five places where the casing foot is known, it is certain that the finished surface was not intended to run down in a slope to the rough limestone outside of it. It is most likely, therefore, that the face was intended, when finished, to end in a vertical foot and this would be covered by the pavement to be afterwards added.

What, then, is to be reckoned as the size of the true base of the Pyramid? Not the present edges of the granite, for they are utterly rough. And not the ends of the fine-dressed edges of the joints, for they end at various levels. But looking to the fact that all the courses of granite are intended to be equal, and a rather short two cubits each, it seems most suitable to take a mean of all the granite courses (since the upper are not thinner than the lower ones),and reckon the intended base of the Pyramid at one mean course height (40.3 1.5 below the first joint. Remembering also that the Second Pyramid courses average two cubits each near the base, and the bottom course was just two cubits above the pavement.

80 . At this level, then, the various data of the intended surfaces give the following size for the base, reducing those data that are on higher levels by the angle 51 0' :












Diff from Mean



+ 4.2




+ 16' 48"

+12' 23"

+ 12' 57"


+ 14 03"

Diff from Mean

+ 2 45"

1' 40"

1' 06"


1' 50"

The N. end of the W. side could not be reached, after several attempts ; and hence the lack of knowing the length of the N. or azimuth of the W. side.

The above results are from the best data of each part, but there are other points which are useful as checks. The actual points used are: at N.N.W,, finished line on top of 4th course at N.N.E., close joint of face on p 112 1st course; at E.N.E., line of finish along the side of casing, at adopted base level; at E.S.E., foot of rough casing, which is further in than even the joint surface requires; at S.S.E., finished line on top of 3rd course at S.S.W., close joint on face of 1st course; at W.S.W., finished line on top of 4th course. The check measures are the following three : at the N. side, courses by the entrance projected down at 51 0', fall .8 beyond the side stated; on the E. end of the S. side the rough foot of the casing which was to be dressed down projects 4.3 beyond the side stated and at the W. end of the S. side the line of finish on the 4th course, projected at 51 0', falls .9 beyond the side stated.

It must be remembered that if any different base level should be supposed to have been intended, it will make no difference in the above azimuths, nor in the differences between the sides.

81 . For the angle of the Pyramid, the data are rather divergent and not only do different methods vary in result, but the measures of similar stones vary far beyond the errors of measuring the angle or judging of the surfaces.

By 7 measures on finished granite, in situ.

By 1st and 4th courses, in situ. at S.S.W.

By 6 single blocks of granite, shifted

By 9 pieces of limestone casing (brought to England)

50 57' 28"

50 42' 07"

51 00' 09"

51 58' 15"

Considering the various sources of error : that the dressed granite in situ. is very irregular; that the 1st course joint at S.S.W. may easily be estimated too far out; and that we have no guarantee in the moved granite blocks, or the limestone from the upper part, that the courses were horizontal (on the contrary, one granite block has two different joint surfaces, 1 40' different) ; the best conclusion seems to be 51 0' 10'. But from a consideration of the granite courses (see below), the angle would be 51 10' 30" 1' 20"; and this might well be adopted, as being close to the very uncertain result from the measured angles.

Hence the height of the Pyramid would have been 2564 15; or 2580.8 2.0 by the granite courses.

82 . The courses were measured to rather more than halfway up the N.E. corner, beside measures of the lower courses taken at each corner. The series at the N.E. is as follows, reducing to the base level above adopted. 40.3 below the first joint :





























































p 113





























































The full height or the bottom course is : N.E., 45.5 S.E., 55.3 ; S.W. 43.7. The granite is here marked as ceasing at 645.2, i.e., including the lower 16 courses. The reasons for this are (1) the highest remaining fragments of granite (mere back ends of casing stones) are at the same level on each of the sides: hence, the granite must have come as high, and probably did not go higher, as all the pieces are on the same course; (2) there is a thicker course next over this, as if some great change took place there, and a fresh start was made; the 17th course is thicker than any other course of the whole Pyramid, and is followed by a course thinner than any that underlie it; (3) Diodorus states that the casing was of black stone up to the 15th course, and like the other Pyramids above that level. Now, by the stumps of the stones the granite must have come to the 16th, and probably the lowest course was covered with sand in his day; but it is unlikely (unless we credit him with loose errors like modern guide books *) that the casing went much higher. Hence the strong suggestions (1) and (2) are confirmed by (3), and may well be accepted.

* In one of the most scientific of guide books it is said that the Third Pyramid cannot be ascended (it is easier than the Great Pyramid) and that it was "covered with slabs of polished. granite, and the upper part with rough stones !" or, making matters worse still, "in the case of the Third Pyramid the whole. surface was to be, as it were,veneered with slabs. of granite !" showing that the writer had never realised the proportions of a casing stone. But descriptions of the Pyramids are usually replete with extraordinary mistakes "granite" for "limestone", "height" for "width," &c

This being settled, it is worth notice that the granite just covered one quarter of the height of the Pyramid, the total height being 4 x 641 4 Conversely this may be taken as giving a determination of the original total height, perhaps more accurately than by the varying angles of the Casing, thus :

6452 ( 5 (?) for uncertainty of paving) x 4 = 2580 2.

And this yields an angle of 51 10' 30" 1' 20".

The mean planes of the edges of the core masonry are far more irregular than those of the Great Pyramid. At the base level adopted they are 4,082 on E., 4,077 on S., and 4,109 on W. averaging 4,089 apart; and their mean distance p 114 from the casing plane varies from 14 to 46, averaging 33. The core has no uniform skew to the casing, as in the Great Pyramid. The thickness of the top of separate granite casing stones is from 35 to 46, averaging 41.

The Casing, though partly attacked in the 12th and 13th centuries, does not seem to have been removed in the time of Belon (1548), or of Villamont (1589), who describe it as perfect, and without steps.

83 . The peribolus walls around the Third Pyramid (see P1. v.) are all built of unhewn stone, neatly laid with mud mortar, like the walls of the barrack galleries of the Second Pyramid. They are, however, irregular in their position, some being nearly square and parallel with the Pyrarnid, and the others on the South being very different. They were all fixed in the survey by triangulation, with more accuracy than the wall-surface can be defined.

The N. wall joins the portion of a wall, S. of the Second Pyramid, by an elbow, and runs thence westwards at 14' from true W. Beyond its corner, where it turns to the S., a fainter enclosure wall begins, running due W. The spaces along these walls are proportional to each other; from the corner of the small enclosure.

E, to that of the larger, D ... 6275 4

or by the Southern side, F to G ... 6196 4

Corner, D, to junction of Second Pyramid peribolus, B ... 7689 5

Peribolus B to junction of cross-wall A ... 7813 5

The mean of these is 1,553, which is perhaps 75 cubits of 20.71 inches.

Not only does the peribolus of the Second Pyramid appear to be thus connected in its position, but the wall at the head of the galleries, if prolonged, would pass but 29 inches within the W. side of the Third Pyramid and therefore those seem to be intended for the same line. And this connection is confirmed by the equality of the two divisions of this line :

Outside of last gallery to S. side of Third Pyramid peribolus ... 3,258
or inside of last gallery to S. side of Third Pyramid peribolus ... 3,304
S. side of peribolus to N.W. corner of Third Pyramid ... 3,309
Also, W. side of gallery wall, C, to E. side Second Pyramid peribolus ... 3308

The mean of the latter three is 3,307, which is, perhaps, 160 cubits of 20.67 inches. The length of the W. enclosure 9,599 N. to S; is subdivided by a very faint ridge, in which no wall could be found. But this ridge runs straight towards the centre of the Pyramid, and it appears to be roughly about the breadth of the Pyramid, or 200 cubits, from the S. wall. Referring to the Pyramid side produced out westwards, as being the best-defined line of division for this, the N. side of the Pyramid is (3,309 + 92 (?) ) = 3,401 from the outside of the N. peribolus ; and as this is intentionally in line with the N. wall of the enclosure, therefore the S. wall of that enclosure is (9,599 - 3,401) = 6,198 S. p 115 of the N face of the Pyramid by intention - ie., as laid off by the builders.*

* To understand a scheme it is necessary to take measurements, as far as possible, in the same order that the builders took them - i.e., including their mistakes in each step of the laying out; and so see, not what errors there are from a mathematically rigid plan, but what errors there are in each part as it was planned.

Now this is exactly the breadth of this enclosure (6,196 to 6,275), and is equal to 300 cubits of 20.66. Hence the design of this W. enclosure is a square of 300 Cubits, W. of the peribolus wall and S. of North face of the Pyramid, while its N. wall is advanced to the line of the N. peribolus wall. The W. wall of the enclosure is nearly straight, two points fixed on it lying 3 and 12 inches outside a line joining the corners.

The total length of the W. wall of the peribolus, 14,049 from corner to corner (D to J), does not seem to have any simple relation to other parts and the only connection observed in it is that the distance of the S.W. corner (J) from the S. wall of the enclosure (F) is equal to the distance of the W. peribolus from the side of the Pyramid. The measures are :

S.W. corner (J) to S. wall of enclosure (F) ... 4,450
W. wall peribolus, from W. side Pyramid ... 4,450
or N.W. corner penbolus to line of galleries (C) ... 4,451
S.W. corner (J) to branch wall (at K) ... 8,897 = 2 x 4,448

This length is necessarily (375 - 160) cubits by the previous relations and the mean 4,450, equals 215 cubits of 20.70.

So it would seem that these walls had not been planned all in one design, but added on by different schemes ; referring more or less to the Pyrarnid, and using round numbers of cubits in general, but getting more complex quantities by addition or subtraction of simple lengths. The irregularity of the S. peribolus wall exactly agrees to this view, as it is impossible to suppose its skew and bowing line to have been laid out along with the very regular lines of the other parts.

The end of the S, wall runs through the side of a large mound, and disappears, so that it could not be exactly fixed. The end of the branch wall likewise runs through the side of a mound, and then ceases. Those mounds would have been cut through, had time allowed. The temple on the E. side of this Pyramid appears to have been the most perfect that was visible at Gizeh in 1755 ; and Fourmont mentions four pillars as then standing in it. It has now lost all its casing (used by the Mamelukes for houses at Gizeh), and merely the core blocks remain, weathered away in some parts so as to have fallen over. The marks where the walls have been cut, to fit in the backs of lining blocks, show that it was cased (probably with granite) like the temple of the Second Pyramid and the Granite Temple.

The causeway is just the width of the entrance passage walls; it is built of p 116 large blocks, and raised, probably, 20 or 30 feet above the plain, though the sides are now much hidden by sand. It ran down the hill for about 800 feet from the tempIe ; but it had no connection with the other causeway, situated half a mile further E. in the plain below, though they are often confounded together. The lower causeway is not in the line of the upper, nor parallel to it; and it only ran up to the quarries in the limestone hill, which is such a striking feature in the neighbourhood.

There was a considerable village of Graeco-Roman age around the Third Pyramid. A great amount of crude brick and pottery lies on the S.E.; crude brick is also found on the causeway, and is mentioned by Vyse as found on the pavement at the N. side.