Previous |Contents |Next

The Central Precinct

The separateness of the area enclosed by the stylobate is clearly noticeable to those who do not have preconceived notions by the pattern of the floor tiles.

It is amazing that Dörpfeld and Dinsmoor, who have spent so much time and energy in trying to prove that the temple as a whole was planned by the Aiginetic foot of 326-328 mm., have missed the one part of its architecture that was actually planned by this unit. The floor tiles of the area enclosed by the stylobate of the colonnade, the sunken area that Penrose called the impluvium, was covered with tiles that measure 4 x 5 pieds de roi of geometric form. This unit of 326.8907 mm. was obtained by taking 11/10 of Roman foot of geometric form. The tiles measure 1,307.6 x 1,634.5 mm. Magne reports that they measure 1,300 x 1,635 mm.

The significance of these tiles was realized by the British architect Knowles, who surveyed the temple in 1846. He did not publish any report, but the report of Penrose, who surveyed the temple in about the same period of time, includes a plate (Table 3) which gives the results obtained by Knowles. The only other archeologist who paid attention to these tiles was Magne, but his survey is less accurate than that of Knowles. The only other archeologist who paid attention to these tiles was Magne, but his survey is less accurate than that of Knowles. Magnes’ data are useful mainly in correcting a few misprints in Penrose’s edition of Knowles’ results: for instance, whoever drew the plate with Knowles’ data in some cases read the number 7 instead of the number 1.

In general, Knowles’ measurements err on the side of excess. At times Penrose added up Knowles’ measurements of the single tiles in order to obtain data about stretches of the Naos, with the result that he contradicted other data of his and submitted figures that are substantially larger than those obtained by Magne. These two facts, that Penrose contradicted himself and often submitted measurements of length that are greater than those obtained by Magne, has been pointed out by Ernest Flagg in a detailed comparison of the reports of Penrose and Magne.

The width of the sunken area in principle is 30 pieds or 9,806.7 mm. According to Balanos, its width is 9,809 mm and according to Magne it is 9.81 m. This space is covered by 6 rows of tiles that are 5 pieds or 1,634.5 mm wide.

According to Knowles the 6 rows of tiles have the following width (from north to south):

64.4 English inches
= 1,635.8 mm
= 1,630.1 mm
386.1 English inches
= 9,806.9 mm.


In the case of the fourth row, Penrose published the erroneous figure of 64.7 inches. The figures published by Penrose add up to 386.7 inches = 9,822.2 mm. This may be the reason why Penrose reported the total width of 322.233 English feet = 9,824.6 mm.

According to Magne the tiles have the following width:

1,635 mm.
9,807 mm.

It seems that the width of the tiles was modified in order to achieve variatio, that is, in order to avoid monotony, just as the intercolumnia were varied for the same reason in all the colonnades of the temple. Since the pied de roi is equal to 11/10 of Roman foot, it was divided into elevenths which in turn are divided into tenths. It appears that one row of tiles was reduced by 1/110 of pied to 1631.5 mm., whereas, to compensate, the other 5 rows were increased by 2/1100 each to 1,635.0 mm.

The colonnade, too, was calculated in pieds de roi. The intercolumnium was 8 pieds or 2,615.1 mm. It was formed by two stylobate blocks of 4 pieds each. The distance between the square stylobate blocks of the corners and the axes of the nearest column is 7 pieds. The distance between the front wall and the axis of the first column on each side was also 7 pieds, or 2,288.2 mm.

Accordingly the width of the area enclosed by the colonnade was reckoned as follows:

2 corner spaces of 7 pieds 14 pieds
2 intercolumnia of 8 pieds 16 pieds
  30 pieds = 9,806.7 mm.

Penrose reports the following data about the spacing of the columns, from north to south:

2,286.3 mm.
9,807.5 mm

The two intercolumnia, which should be 8 pieds or 2,615.1 mm. in one case are increased by 1/11 to 8 1/11 pieds or 2,644.8 mm and in the other case are reduced by the same amount to 7 10/11 of pied to 2,585.4 mm.

The two corner spaces, instead were increased and decreased by 1/110 of pied:

6 109/110 pieds
= 2,285.3 mm.
7 1/110 pieds
= 2,291.2 mm.

Knowles, in reporting about the length of the stylobate blocks, submitted data that do not agree with these, if we are to assume that the axes of the columns fell exactly on the junction of the blocks. These are Knowles’ data in English inches, from north to south:

Corner space 90.2 = 2,291.1 mm.
Normal intercolumnium 102.0 = 2,590.1 mm.
Normal intercolumnium 103.7 = 2,590.8 mm.
Corner space 90.2 = 2,291.1 mm.
9806.8 mm.

It may be that the report of Knowles must be interpreted as reflecting the following data:

39.7 inches
= 1,008.3 mm.
3 9/110 pieds
= 1,007.04 mm.
50.5 inches = 1,282.7 mm.
3 102/110 pieds
= 1,283.8 mm.
45.2 inches
= 1,148.1 mm.
3 56/110 pieds
= 1,147.1 mm.
56.8 inches
= 1,442.7 mm.
4 46/110 pieds
= 1,444.2 mm.
55.1 inches
= 1,399.5 mm.
4 31/110 pieds
= 1,399.7 mm.
48.6 inches
= 1,234.4 mm.
3 85/110 pieds
= 1,233.2 mm.
51.1 inches
= 1,297.9 mm.
3 107/110 pieds
= 1,298.6 mm.
39.1 inches
= 993.1 mm.
3 4/110 pieds
= 992.6 mm.

If this interpretation is correct, the two corner spaces are both 7 pieds + 1/110, and the two normal intercolumnia are one 8 pieds - 8/110 and the other 8 pieds + 6/110. The total is exactly 30 pieds.

The figures of Knowles are confirmed by the less accurate report of Bö tticher (Akropolis, 122) which provides the following data for the stylobate blocks:


All the data confirm that the central area had a dimension that deviates at most a mm. from its theoretical value of 30 pieds = 8,806.7 mm. (Penrose measured it as 32.233 English feet = 9,824.6)

The pattern of the tiles measuring 4 x 5 pieds is closely related with the raison d’être of the Naos, that is, the famous statue of Athena cut by Pheidias. Today nothing remains of this statue and nothing remains even of the pedestal (called bathron in the inscriptions) that supported it; but on the floor of the Naos one can recognize clearly the emplacement of the pedestal because the area where it used to be was not paved, letting the poros stone substructure of the temple come to the surface.

Eight tiles were left out because of the pedestal, two tiles in the sense of the length and four tiles in the sense of the width. Hence, in principle the pedestal occupied an area of 8 x 20 pieds or 2,615.1 x 6,537.8 mm. According to Balanos the unpaved area measures 2,614 x 6, 542 mm. In the direction of the length of the Naos, the tiles that enclose the unpaved area do not stop all on the same line, since some protrude more than others. According to Penrose where opposite tiles come closest to each other they are at a distance of 2,601.5 mm.

Between the emplacement of the pedestal and the stylobate there was only one row of tiles on each side. In principle each tile should be 5 pieds or 1,634.5 mm. wide. Balanos reports a width of 1634 mm. on the north side and of 1633 mm. on the south side.

Markings on the tiles surrounding the pedestal indicate that at the foot of it there was a sill that covered part of the nearest tiles. The dimensions of the pedestal with the sill wee 4.09 x 8.04 m. according to Bötticher (loc. cit.) and 4,091.0 x 8,039.5 mm. according to Penrose. It could be that, when the builders placed a sill around the pedestal, they shifted back to the standard generally used in the planning of the temple, the trimmed lesser foot. The dimensions of the sill may have been intended to be 14¾ x 29 trimmed lesser feet or 4,092.3 x 8,046.0 mm.

The pedestal occupied almost the entire width of the area enclosed by the colonnade, leaving only a narrow passage on each side. According to Penrose, the distance between the sill of the pedestal and the stylobate was 883.9 mm. on the north side and 883.3 mm on the south side.

Previous |Contents |Next