1. The length of a temple is adjusted so that its width may be half its length, and the actual cella one fourth greater in length than in width, including the wall in which the folding doors are placed. Let the remaining three parts, constituting the pronaos, extend to the antae terminating the walls, which antae ought to be of the same thickness as the columns. If the temple is to be more than twenty feet in width, let two columns be placed between the two antae, to separate the pteroma from the pronaos. The three intercolumniations between the antae and the columns should be closed by low walls made of marble or of joiner's work, with doors in them to afford passages into the pronaos.

2. If the width is to be more than forty feet, let columns be placed inside and opposite to the columns between the antae. They should have the same height as the columns in front of them, but their thickness should be proportionately reduced: thus, if the columns in front are in thickness one eighth of their height, these should be one tenth; if the former are one ninth or one tenth, these should be reduced in the same proportion. For their reduction will not be discernible, as the air has not free play about them. Still, in case they look too slender, when the outer columns have twenty or twenty-four flutes, these may have twenty-eight or thirty-two. Thus the additional number of flutes will make up proportionately for the loss in the body of the shaft, preventing it from being seen, and so in a different way the columns will be made to look equally thick.

Vitruvius' Temple Plan Compared With Actual Examples
vitruvius' temple plan compared with actual examples

3. The reason for this result is that the eye, touching thus upon a greater number of points, set closer together, has a larger compass to cover with its range of vision. For if two columns, equally thick but one unfluted and the other fluted, are measured by drawing lines round them, one line touching the body of the columns in the hollows of the channels and on the edges of the flutes, these surrounding lines, even though the columns are equally thick, will not be equal to each other, because it takes a line of greater length to compass the channels and the flutes. This being granted, it is not improper, in narrow quarters or where the space is enclosed, to use in a building columns of somewhat slender proportions, since we can help out by a duly proportionate number of flutings.

4. The walls of the cella itself should be thick in proportion to its size, provided that their antae are kept of the same thickness as the columns. If the walls are to be of masonry, let the rubble used be as small as possible; but if they are to be of dimension stone or marble, the material ought to be of a very moderate and uniform size; for the laying of the stones so as to break joints will make the whole work stronger, and their bevelled edges, standing up about the builds and beds, will give it an agreeable look, somewhat like that of a picture.