The intercolumniation of this order differs from that of all the others, inasmuch as the interco]umns are determined not by the diameters of the column, but by the arrangement of the triglyphs; and the different methods, instead of being distinguished as pycnostyle, eustyle, araostyle, cfc., are com prehended under the terms monotriglyph, ditriglyph, according to the number of triglyphs over each intercolumn ; the former term designating the arrangement in which there is but one triglyph between the columns, the latter that in which there are two within the same limits. This method of disposing the columns naturally arises from the employ ment of the triglyph, for this ornament forms so conspicuous an object in the elevation, that it was necessary to make its position conform with the other principal members of the order, of which the column is the most important, so that the colonnade and entablature might present to the eye a similar arrangement. Had the triglyphs been placed in the frieze without reference to the position of the columns, the eye, after passing up the length of the column, would have been confused upon reaching the frieze, and probably would have stopped short at that member, there being nothing to carry it upward to the cornice. Did we adopt the Vitruvian theory, the position of the triglyph would readily be accounted for in another way, for it stands to reason, that the feet of the rafters would be placed immediately over their support.
It being necessary then that the triglyph should stand over the centre of each column, and the proportion of the metope or space between the triglyph being determined, we only require the height of the frieze, and the intercolumniation is fixed. The width of the metope being about equal to the height of the frieze, and the triglyphs somewhat less, it is evident, that to place a column under every triglyph, would be impracticable, without increasing the height of the frieze quite beyond proportion ; there would be scarcely room for the feet of the columns, much less for any space between. h became necessary therefore to place another triglyph in the centre of each intercolumn, with two metopes instead of one : this arrangement answered very well, and is the most frequent in Doric temples, and indeed is seldom departed from in any buildings. By inserting another triglyph, you are compelled to add another metope, and this makes the intercolumn half as wide again, which is almost too wide to suit the requirements of taste, as well as the proportions and construction of Doric buildings. There are very few instances of this arrangement, and these only in the centre intercolumn opposite the entrances, where greater space was required ; this is especially noticeable in the Propyhea, where a large space was required for the admission of chariots. It may be supposed, that the monotriglyphic method would cause the columns to appear too closely set, and this cer tainly would be the case in some instances, where the columns are less than a diameter and half apart, were it not that the shafts converge so rapidly towards the upper diam eter as to leave a space under the soffit of the architrave, even in such instances equal to more than twice the upper diam eter.
The peculiar position of the extreme triglyph has already been noticed in speaking of that member, as also the effect produced by it in lessening the extreme intercolumn by the space of half a triglyph ; but there still remains to notice another peculiarity, which was first published by Mr. Donaldson; we allude to the inward inclination of the outer columns.
" The axis of the columns of the Parthenon," says he, " both on the flanks and on the fronts, as well as those of the temple at Egina, and of Concord at Agrigentum, have a con siderable inclination inwards (a circumstance I am not aware to have been before noticed) though not to such a degree as required by Vitruvius, and not confined, as he directs, to the columns of the peristyles only." Vitruvius thus directed :— " The bases being thus completed, we are to raise the columns on them. Those of the pronaos and posticum are to be kept with axes perpendicular: the angular ones excepted, which, as well as those on the flanks right and left, are to be so placed, that their interior faces towards the cella be perpendicular. The exterior faces will diminish upwards, as above mentioned. Thus the diminution will give a pleas ing effect to the temple." Mr. Bartholomew alludes to the same circumstance thus "The ancients, knowing how much more secure were their fabrics when made to settle together and consolidate by their own gravity, set the lateral columns of their temples with their axes falling towards the cells, so that the inner faces of the shafts of the columns should be perpendicular, and the outer faces of them receding the whole quantity of columnar diminution, in order to afford to the building a more solid, pyramidal, and graceful appearance; and by this shrewd device they rendered the avenues between the side-walls and the colonnades of their temples no wider next the soffits of the architraves than down upon the pavement ; and it is not improbable, that the preservation of this symmetry led to the omission of the inner columns of the ancient Pseudo dipteral temples ; whereas the moderns, in general, not attending to his dynamic and optical nicety in architecture, so set their columns, that when we walk down a modern collonnade, we cannot divest ourselves of the idea that the axes of all the columns are falling outwards : and, indeed, accurate admeasurement would often find this to be no illu sion, since the work, not erected so as to fall together, will, in general, with the slightest inevitable settlement, expand at its upper part." It is worthy of remark, that, in many in stances, the angular columns are made somewhat thicker than the others, so as to give them an appearance of much greater strength.