1852.] OF THE ROYAL INSTITUTION. 127 
throwing aside all advantage to be derived from them, but by con- 
tinually improving them by the application of a scientific and re- 
fined criticism. 
The early temple consisted of an oblong cella, the shrine of the 
god, surrounded by a Peristyle, or colonnade, usually leaving an 
ambulatory all round the cella. This composition was covered by a 
single roof from end to end, with a triangular pediment, or flat gable 
end, at each extremity. In the latest and most complete examples 
we find the simple general form preserved, but combined with a 
highly organized construction, The most important peculiarities 
are these. The contours of all the columns are delicate curves. 
The axes of the columns, with scarcely an exception, inclined inwards 
towards the temple, and the horizontal lines have a slight curvature. 
The process of refinement would seem to be as follows : — 
The analogy of nature in all bodies which are employed to exert 
force or pressure at a distance from their attachment to a fixed 
point, would soon suggest the practice of making the columns 
taper. The earliest builders would indeed find this done to their 
hands in the trunks of trees, which found a place in the supports of 
the hut, more or less the archetype of the Greek temple. It 
would then be observed that a column built with straight sides, 
i.e. tapering as the frustum of a cone, would appear attenuated or 
drawn in in the middle — owing to the eye not judging of the form 
absolutely as an instrument, but bringing into the calculation cer- 
tain feelings derived from contrast, which make the apparent differ 
from the real form. 
The causes which lead to this apparent attenuation seem to be 
chiefly these : — 
There is some reason for imagining that a frustum of a very 
sharp cone will under any circumstances appear formed of concave 
instead of straight sides, but this effect, if it take place at all, will be 
very slight: other causes however affecting such a frustum when 
used as a column are more influential. 
The eye will rest longer upon the top and bottom of the shafts 
than on any intermediate points, and this relative importance pos- 
sessed by those situations will make the intermediate diameters 
appear too small, when compared with the two extremities of the 
shaft. 
This effect will be further increased by inequalities of chiaro- 
oscuro. For the relative excess of light on the upper part, owing 
to the deep back ground of shade under the Portico; and the actual 
excess on the base owing to the greater amount of reflected light, 
will throw the balance of light in favour of the extremities, and 
according to a principle familiar to observers of Nature the brighter 
parts will appear broader on that account. Among the most obvious 
illustrations of this phenomenon are the apparent disc of a fixed star ; 
the magnified appearance of a distant candle; the sudden ap- 
pearance of dilation in the disc of the moon when issuing from a 
fleecy cloud, or its opposite appearance of stooping on entering it, 
so poetically described by Milton. From all these, and possibly 
