152 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL. 



tMAY, 



or rather Smpossihility, of using the circle, which would require a diameter 

 of al>out 2^ miles. 



Let it be required to construct a circular or other arc of uniform curva- 

 ture, whose length is 100 feet, and the rise at the centre is to be '25, or any 

 other small measure wliicli must not mucli exceed one font. Construct 

 »ith any axis a parabola, and set otf from the vertex A B = the proposed 

 rise, and draw Ij M at rijjht angles with A B. Now, L M will represent the 

 100 feet horizontally, and ordinates drawn to the curve perpendicular to 

 L M will determine the exact rise at as many points as may bo required, full 

 size. 



The curve on the upper step nnrtli side of the Parthenon, also approxi- 

 mates to a regular curve very cl(»sely ; its entire rise in the centre above the 

 line joining its extreme parts, is '300, which is very nearly in the proportion 

 of i, of the rise in the east front: it is exactly jjj length of the building. 

 The curve on the south side seems to have been identical with the north 

 side, but it has suffered more from the concussions which the building has 

 undergone, especially as there is a great depth required on this side of arti- 

 fiSal foundation, dntlie north side the steps rise almost immediately from 

 the solid rock. The curve on the west front is not quite so symmetrical as 

 on the other sides. It has, I believe, been affected by the lines of the old 

 building. The rise is exactly the same as the east end. 



The upper members on all four sides follow the steps, and are nearly 

 parallel, but there is a little more curvature given to the steps ; the entire 

 rise of architrave is -173 on east front, -175 on west. The levels of those 

 poitions of the entablatures which remain on the north and south sides 

 point out the directions which those lines had originally, and they were as 

 nearly as possible parallel to the line of the step, excepting that just at the 

 angle columns the step has a little the more declension. The frieze and 

 cornice are exactly parallel with toe architrave. In the temple of Theseus, 

 also, these curves prevail; on the fronts the rise is ^ part of its length, on 

 the flank, t;i^. The lines in the architrave are exactly parallel to the step. 



There is one relinement which the temple of Theseus possesses, which the 

 Parthenon is without. In addition to the cornice being raised, the inclined 

 lines of the pediment have a very slight convexity, between -02 and '03. I 

 was unable to fix more precisely the amount. I imagine that it was owing 

 to some degree of haste in wliich the Parthenon was finished, of which 

 there are several indications in the upper meraberi, which prevented this 

 t'.nal adjustment being made to its pediments ; the state of the political 

 horizon at that time making the completion of the long walls of more im- 

 mediate importance than the optical corrections of the Parthenon. On a 

 former occasion, I stated my impression that the came which led to the 

 adoption of this convexity of the horizontal line, existed in the contrast of 

 the inclined lines of the pediment. 



Mr. Ferguson has kindly favoured me with an illustration, which I will 

 read to you, from a description of the construction of an iron foundry at 

 Kasipur, near Calcutta, built in the year 1834. The foundry is covered by a 

 single roof, with principal rafters, tie-rods, and suspension bar from the 

 centre. The rise is 6 feet and the spaa 50 feet, which is exactly the same 

 pitch as the Parthenon and Propylaja. The passage is extracted fiom vol. 

 iv. of the journal of the Asiatic Society of Bengal, p. 116 : " Before closing 

 our short account of the Kasipur roof, we must notice a curious optical de- 

 ception, for which we are somewhat at a loss for a correct explanation. On 

 entering the room and looking up at the roof, it strikes every beholder that 

 the roof has somewhat sunk, and the horizontal tie-rod is about 5 or 6 inches 

 lower in the centre than near the walls. We were not convinced that 

 it was a deception until Major Hutchinson, at our request, caused an 

 actual measurement to he made by a perpendicular wooden batten from 

 an accurately adjusted level on the stone floor. It was then proved 

 that there did not exist a difference of level even to the tenth of an 

 inch." The conclusion is obvious that a straight tie-rod appeared to 

 be deflected; and I have no hesitation whatever in ascribing the cause to 

 the contrasting lines of the principal rafters. I do not think that it is ne- 

 cessary to have our eyes refined by a southern climate for the appreciation 

 of these effects. I suppose that there are very few gentlemen here who have 

 not felt the same disagreeable effect of a flat open roof with horizontal tie- 

 beams, unless, indeed, the lattsr be very much cambered. That this was 

 the view the ancients took of the matter I am convinced by these two 

 facts — 



That the great temple at Psestum has the convexity only in its fronts, and 

 not on its flanks ; and in the Propylffia at Athens, although the base on which 

 the columns of the two pediments stand is perfectly straight and level, 

 the line of the architrave was curved. For enough remains to determine this 

 in the eastern portico — the central columns are actually about "12 higher than 

 those at the angles. The base in this building is cut in two by the ascend- 

 ing roadway, so that there oould have been little or no advantage in a con- 

 vex base line. 



It will be well to remember that the temples at Athens were the result 

 of the experience of several centuries in which these refinements were gradu- 

 ally brought to perfection. The first process was probably to raise the 

 cornice under the pediments and entablature, by making the middle columns 

 a little higher than those towards the angles, as I have mentioned in the 

 case of the Propylfea. Still it is likely that to a fastidious eye the straight 

 line of the stylobate would appear weak. Tlie second method would be 

 that found in the great temple at Pajstum, in which the fronts have the con- 

 vexity in their steps, as well as their entablatures, the flanks being composed 

 aith horizontal hues. Perhaps a reasonable man should be content with 



this. I must willingly admit that I was perfectly content with the temple at 

 Pffistum; still nothing short of perfection could satisfy the refinement of 

 vision with which the Greeks alone, among the people of all time, seem to 

 have been endued ; and perhaps by looking at a temple constructed as 

 above-mentioned, anglewise, and contrasting the convexity of the corona of 

 the fronts with the straight line of that on the flanks, or more likely the 

 comparison of the two different forms of line on the stylobate, suggested 

 the possibility of improvement ; at any rate as early as the time of Pisis- 

 tratus, the Atiienians had begun to demand from their architects the per- 

 fected construction, as the foundations of the temple of Jupiter Olympus 

 testify, which we know were laid during his reign. 



I also refer to his time, the earlier temple of Minerva, which occupied 

 the site of the Parthenon at the lime of the Persian invasion, in which also 

 we find that the lines on the four sides of the building were convex. I can 

 bear witness also to convexity on all the four sides of the temple at Nerasa, 

 in the Peloponnesus, and Segeste, in Sicily. I could find no trace of con- 

 vexity at Corinth, Egina, Rbamnus, or Bass;e. 



The next subject is the inclination of the columns and the upright faces 

 of the building. 1st. The face of the steps inclines about -008. 2nd. 

 The columns iiicline, backward, a quantity, of which I obtained the follow. 

 ing results. From the average of the measurements of all the lower drums, 

 (scamillce impares, as Vitruvius calls them) '229. From plumbing the an- 

 gular columns of the east front, taking into consideration all the cracks and 

 movements which have modified its original position, I obtain two results of 

 •230 and -232. 



In the plumbing, I observed every precaution to ensure correctness, 

 using a very heavy weight and also watching for calm intervals of weather, 

 which are rare at Athens. I am disposed to think that '228, or one thou- 

 sandth part of the length of the building was the amount originally intended. 

 Those gentlemen who remember the perfection of the joints with which the 

 Parthenon is constructed, will allow that the openings between them, which 

 at present exist, are the exact records of all the settlements which the 

 building has undergone, and that by a careful examination of these, the 

 original amounts may be exactly recovered, which would be hopeless in a 

 building which had been of less highly-finished construction. 



Vitruvius directs that the columns of the pronaos and posticum should 

 be set perpendicular, and those of the peristyle should incline towards the 

 cella. In Cicero ad Verrem, we have an amusing passage, in which Cicero 

 relates one of his rascalities ; that having under his charge, as Roman 

 governor, a young Syracusan nobleman, whose property was subject to the 

 condition of repairing the temple of Castor in that city, Verres was ex- 

 ceedingly anxious to make a job of this ; so he goes to examine the temple, 

 but was much disappointed on finding it in perfect repair, when one of his 

 companions casually observed, there is nothing here to be done, unless yon 

 order the columns to he set perpendicular. Verres was evidently surprised 

 at this observation, for be knew nothing of architecture, and to his eye the 

 columns appeared angular ; but it was mentioned to him by those who 

 were around him, and no doubt familiar with the practice of the ancients, 

 /ere nultam esse columnam quce ad perpendiculum esse possit — namely, 

 that in a temple there was scarcely any column, which by the practice of the 

 Greeks, could be perpendicular. Verres was delighted at finding something 

 to set his young friend about, and said, " Oh, by all means let him set the 

 columns perpendicular;" and no doubt he took care himself to superintend 

 the payment. 



The object of this adjustment is to correct an appearance of fanlike 

 spreading from the base outward, which takes place in columns which are all 

 perpendicular. It may he owing to this cause that, in consequence of the 

 diminutiun of the columns, the spaces between them on the architrave are 

 greater than those on the ground (for the eye quite makes allowance for the 

 counterspread of the capitals). Again, owing to the greater depth oC 

 snadow behind them, the upper part of the columns will have apparently 

 more strength of light, and consequently appear greater. The effect altoge- 

 ther produced will cause the architrave, if equal, to appear longer than the 

 base, and consequently the angle columns will appear to lean outwards: this 

 is rectified by making the said angle columns lean a little towards the centre 

 of the building. It is a proof ot the wonderful judgment with which this 

 quantity was chosen, that so many diligent and accurate observers have 

 studied and drawn these temples without being aware of the fact. When 

 my attention was first called to it, I could not at all perceive it ; and I 

 greatly amused a French architect, who had been for some time at Athens, 

 by asking him which way they leaned : after some days the eye began to 

 take cognizance of it, and I could perceive which way it went. 



The impression of strength and beauty resulting therefrom is by no means 

 confined to those who are cognizant of the fact ; and I doubt not that many 

 of our earlier investigators have been astonished that the level-and-plumb 

 imita'ions of Greek architecture which we have in this country so little re- 

 called to their minds the consummate beauties of the Parthenon and other 

 Greek buildings. No doubt our climate is unsuited to the pure Greek, but 

 this is not enough to account for the falling off; it is not so in our Roman 

 and Palladian buildings. 



One peculiarity which I noticed is that the antae lean forwards to meet 

 the columns of the pronaos. This seems to have resulted from the inclina- 

 tion inwards of the columns of the peristyle ; those of the pronaos and 

 posticum being perpendicular, the effect would have been to produce a 

 strong contrast between their different positions, and the artifice must have 

 been detected: by the influence of the contiguous face of the untai, the 



