This comes from my own published work called "The Once and Future Forum: Approaching Virtual Reconstructions as an Archaeological Tool"

According to tradition, the Temple of Castor and Pollux was vowed in 499 BC by the Roman dictator Postumius when the brothers, Castor and Pollux, appeared a few feet from its eventual foundation. Postumius’ son dedicated it fifteen years later. The day of the dedication is given as January 27th. Officially the temple was named Aedes Castoris.

L. Caecilius Metellus restored it in 117 BC. The concrete of the podium associated with this rebuilding is the earliest dated concrete known. It suggests construction with three rows of four columns preceding a cella like that of Temple A at Pyrgi. One of these columns was the focus of a scandal in 70 BC when Cicero accused his political enemy, Verres of merely re-stuccoing it when he had undertaken to replace the stone itself. Livy writes about the Transvectio Equitum on the 15th July. On this day, it was the custom in front of this temple for the Censors to review an assembled Equestrian order clad scarlet-striped trabea and crowned with olive-leaves in memory of the battle of Lake Regillus.

The temple was rebuilt 49.6 m. long and 32.1 m. wide as an octastyle by Tiberius in 6 AD and dedicated both by him and his brother. It now was peripteral with a Corinthian order and eleven columns on each long side. It is almost certain that a double row was set at each side of the shallow pronaos. The pronaos was 9.9 m deep and 15.8 m. wide. The cella that lay on the back of the podium was 19.70 m. deep and 16 m. wide. In it were kept the standards for weights and measures.

The columns were of white marble from Mount Pentelicon near Athens and 12.5 m. high and carried an entablature with a plain frieze and a richly worked modillion cornice. The Corinthian capitals have interlocking inner spirals that Mediapolis (Baalbek) imitated, and again by James Gibb at St. Martin’s in London, plus on the façade of the Cambridge Senate House. A more notable fragment than these was recovered over a drain plate nearby. It is nearly 8 feet long shaping the pediment’s southeastern angle. It includes the inclined, in addition to the horizontal cornice, with adjoining deeply under-cut Augustan moldings. This section supplies the proportions, and provides an accurate depiction of the temple’s pediment.

From the pronaos a flight of eleven small steps lead down from both ends to form the rostra in the front, 3.66 m. above the Sacra Via. The rostra accommodated a fairly large number of people within its railing. This may be one of the three rostra within the Roman Forum that the legionary catalogues describe, although there is no indication that it was ever decorated with beaks. Tiberius’ use of the rostra in front of his temple is probably not unique considering that there are references in literature that Metellus’ earlier temple had the same arrangement.

The concrete podium was very high, the floor standing about 7 m. above the Sacra Via (see figure 4.1). It contained fragments of earlier periods covered by tufa walls. These projected walls within the intercolumniations made chambers that protected the imperial ficus behind metal doors. Also within the loculi were the valuables of private individuals. At its southwest angle to the vicus Tuscus, the torn foundations of the monument have been completely surveyed, revealing the systematic treatment of wooden dowels for binding together the stone blocks (see figure 4.2).

Caligula, somehow, incorporated the temple in his Palatine palace, making an approach to the palace between the two statues, and so the Dioscuri (the name of the brothers, Castor and Pollux) became his gatekeepers. Claudius abolished this modification. Another restoration is attributed to Domitian, and the structure was temporarily called the Temple of Castor and Minerva. Huelsen attributes Trajan or Hadrian for a later restoration. Existing column remains and entablature fragments date from that period. Most of the remains are of the Augustan Period, and any later restorations must have been superficial as to leave no evidence.

Around 202, the Forma Urbis now showed the temple with a façade approached from the ground by a central stairway, now destroyed but still seen in a photograph taken in 1871. At an earlier period, the approach to the temple from the ground was not by a frontal flight at all, but by two lateral staircases, one on either side. The temple was still standing in the fourth century and was included in the legionary catalogues, but almost nothing is known of its history in the later imperial and medieval periods. By the fifteenth century, the temple only had three surviving white columns on the east side and the locals called the area Tre Colonne. Nevertheless, this central staircase, escalating from ground level, did not go back to the original layout. The temple’s identity was lost early on and by the nineteenth century it was wrongly identified as the Temple of Jupiter Stator.

Map of the Temple of Castor indicating the first and second construction by K.-A Nilson and C. Persson for I. Nielsen ActaArch (1988), tav. 1. The right side faces the Roman Forum.

Using VRML, the author rebuilt the temple based upon the measurements provided from the documents already mentioned. Jones reports that the eight front columns and eleven side columns have intercolumniation of 1.5 column diameters between each column. This means that the intercolumniation between the columns is 2.3605 meters. Logic dictates that if the builder puts the furthest front right column at an X-Y of 0, 0 then the next column to the left would be centered at 3.8933 meters (2.3605 [intercolumniation] + 1.5328 [column diameter]). The third column is 7.7866 m. from the center of the first; The fourth is 11.6799 m. distant, and so on with each additional column 3.8933 m. further than the previous column. Along the side, the outer edge of the front column to the outer back column is a length of 40.4658 m. The entire distance from the outer left column to the outer right column brings the entire width at 28.7859 m.

Steinby provides the general floor measurements as 49.6 by 32.1 m if the viewer is to trust that an accurate outline of the temple can be measured. She also states that the temple’s pronaos is 3.66 m. tall. Measuring the proportion of the pronaos’ plinth lower trim and crown to the entirety of the pronaos’ height, each measurement is approximately 0.33 m. tall while the plinth’s height is 2.33 m. With proper proportioning from Ware‘s American Vignola, the crown and lower trim can be extended appropriately.

D’Espouy’s measurements for the remaining part of the entablature transferred into metrics gives further details of the modillions, ova, raking, and coffers to offer a fine picture of the Temple of Castor. Fragments added together allow the measurements and the photographs fit together to provide a complete model (see figure 4.4).

Though the measurements can provide most of the information, there are still faults to the author’s model. First, the viewer may assume that the tympanum has no relief work, even though it is probable that the temple did have something there. The second is that the author’s reconstruction, like the CVRlab, shows no statues along the roof or front. The use of wainscoting, the proportion of the door and the door’s frame and material is conjecture and only based upon what scholars generally know about the period. Finally, the viewer must completely be versed in the history of the temple before separating fact from fiction.

Sketch found in M. Grant's Roman Forum, 82 Drawing by K.-A. Nilson and C. Persson