It is known that the primitive temple of Hera, at Olympia, had, originally, wooden posts, which were replaced from time to time during historic times, as they rotted, by stone columns. Nevertheless, the extreme taper and squat proportions of early Doric columns seem to demand an origin of masonry, and espe cially a masonry made of small stones. The origin would thus appear to be triple ; the shaft, from early rubble construction; the capital borrowed and adapted from Aegean sources; while the entablature is an interpretation in stone of traditional Dorian wood construction. The earliest example known of the Greek Doric order is that of the temple at Corinth, probably dedicated to Apollo, which must be as early as the 6th century, and may go back to the 7th. Other early examples exist at Segesta, Selinus, Girgenti and Paestum, which are all of uncertain date, but un doubtedly prior to the Persian wars. Two of those at Selinus may go back to the 7th century B.C. The latest ancient example is that of the Agora gate at Athens (12 B.C.-A.D. 2). During these 700 years the basic elements of the Doric order did not change; the development, which was great, was only in the gradual refine ment of every feature, and a continual experimentation in the exact treatment of each form, to give the desired result. Columns became taller and more slender, and the entasis (q.v.), or curved taper, more and more delicate. The ovolo of the echinus changed from the obese projections of Corinth and Paestum to the ex tremely refined and subtle curves of the Periclean period. The entablature, which in early examples had been almost half the height of the column, was gradually reduced in size, till in the Parthenon it is approximately one-fifth. The Greek Doric order was probably, in all cases, richly decorated in colour, so that its present appearance of austerity and over-restraint is illusory.
The Roman Doric order has plainly a double origin. The dif ferences between it and the Greek Doric are not due entirely to Roman inability to appreciate the subtleties of Greek work, but merely to the fact that the Etruscans, and perhaps the north Italians, generally, had, at a very early period, developed a column and entablature of their own, with a long, slender, wooden column, having an ovolo echinus similar to the Aegean, and an entablature of wood, sometimes decorated with terra-cotta appli qués. It is this Etruscan column and entablature which Vitruvius endeavoured to describe, and which was misunderstood by the Renaissance, so that the name Tuscan came to be applied to a simplified Doric. It was also the origin of the unfluted, Roman Doric column, with its simple, quarter round echinus. To this the Romans applied an entablature embodying certain Greek features. In Roman architecture, the use of the Doric was re served for small scale columns, as in many of the house court yards at Pompeii (the forum colonnade at Pompeii was originally Greek), and to engaged or attached columns between arches.
When used on a larger scale it was frequently much modified. Thus, in the temple of Hercules at Cori (attributed to Sulla, c. 8o B.c.), a base has been added, consisting of a single torus, but without a plinth. The entablature is extremely delicate in pro portion, and the capital profile approximates that of the Greek Doric. A triglyph occurs on the corner, as in Greek work. In the theatre of Marcellus, at Rome (completed 13 B.c.), there is no base, and in the Colosseum (A.D. 8o) there are no triglyphs. The Renaissance codifiers, however, took these two as the most typical, and their order is a sort of average between them. Another Doric order, coming originally from Albano, has no base except a fillet with apophyge, and the mouldings of the capital are richly carved. Other Roman Doric orders of extreme richness, in which the echinus of the capital is formed by a cyma recta instead of an ovolo are those from a temple on the Aventine, probably of the 2nd century A.D., and one from the baths of Diocletian (A.D. 305).
The Ionic order had manifestly an Asiatic origin. Its capital is a development from stele (q.v.) capitals of Phoenicia and the eastern end of the Mediterranean, which were themselves based originally on the tri-lobed lotus. A famous example of the primi tive type was found at Neiandreia, and another in Messa, in Lesbos. Excavations on the Acropolis at Athens, and elsewhere, have revealed many examples of the intermediate stages between the flaring volutes and awkward proportions of the early type and the refined perfection of the developed Ionic of the temple of Nike Apteros (probably between 44o and 410 B.c.), or the Erechtheum (407 B.c.). The characteristic features of the Greek Ionic order as found in Greece itself are the bold size and exquisite curvature of the volutes, the remarkable perfection of carved ornament that decorates the whole, and the variety of the types of base found.
The treatment of the order at the corners of a portico was difficult. The corner capital was formed with volutes on two adjacent, rather than two opposite, faces; where they met, the volutes were curved out at an angle together under the corner of the entablature. This created a new difficulty on the opposite corner, inside the colonnade, as it brought two half volutes to gether in an awkward way. It was this difficulty that led to a development of a variation of the Ionic capital with four faces the same and angle volutes. The most beautiful Greek example is that of the temple of Apollo at Bassae, designed by Ictinus, the architect of the Parthenon (probably c. 425 B.c.). In this the columns of the interior of the cella are connected by short walls to the exterior wall, and evidently, in order to have them present the same decorative face toward the entrance as toward the narrow nave between them, Ictinus adopted a four-sided Ionic capital.