Identification: Aedes Castoris Rodolfo Lanciani's photograph from 1882 reveals that the two missing fragments belonged to the lower left side of fr. 18a.

As discussed by Eva M. Steinby, who discovered the photo and thus the existence of the fragments, the pieces thus expose more of the Temple of Castor in the Roman forum and the label that inscribed it. The discovery is significant, as the pieces confirm the identification of the temple in fr. 18a as the Temple of Castor, a point that had been disputed by E. Rodríguez-Almeida (AG 1980, p. 98). The lost fragments clearly show that the inscription read [C]ASTORIS (Steinby 1989, pp. 24-25, fig. 3 or LTUR III, fig. 120).

Livy (2.20.12, 2.42.5) and Dionysius of Halicarnassus (Ant. rom. 6.13) reveal the early history of the Temple of Castor in the Forum: Vowed in 499 or 496 BCE by the dictator Postumus, the temple was dedicated in 484 BCE by his son in the very spot where Castor and Pollux had appeared after the battle of Lake Regillus (LTUR I, p. 242). The building was restored in 117 BCE and in 74 BCE, it was rebuilt by Tiberius, it was converted into a vestibule for Caligula's palace, and Claudius also restored it. It served as a meeting place for the senate and as a bank; rooms in the foundation functioned as shops; and the platform in front was for a long period the main tribunal in the Forum for judicial and legislative gatherings (LTUR I, pp. 242-243). Excavations have demonstrated that the Temple of Castor was an octastyle peripteral temple with eleven columns along the sides. Smaller columns lined the inside walls of the cella. This plan is confirmed by frs. 18a, 18b (now lost) and 18c, generally believed to represent the rest of the temple (LTUR I, p. 245).

Steinby, however, has expressed strong doubts that fragment 18a (and thus presumably frs. 18fg as well) belonged to the Severan Marble Plan at all, pointing among other things to the discrepancies between the rendering of the temple in fr. 18a and in frs. 18bc. Close analysis of Stanford's color photograph of fr. 18a as well as the Renaissance drawings of fr. 18bc suggests that it does belong.
First, on 18a and 18fg, the E wall of the podium is rendered with a double line, while the corresponding W podium wall in frs. 18bc is drawn with a single line. A close look at the color photo demonstrates, however, that the square columns were left incomplete on their E side; the engraver never carved the dashed line that would have completed the E side of the colonnade. Since this particular engraver has rendered other sections of this temple in a sloppy fashion (see for example the missing sides of the two square columns in the NW corner of the podium in fr. 18bc), it is not unlikely that the double line simply was a mistake. He was supposed to carve a dashed line to complete the colonnade, but instead he carved a solid line that looks like part of the wall of the podium. Another example of a sloppy carving that influences the representation of the colonnade along the side of a temple can be seen in fr. 273b.
Second, Steinby points out that on fr. 18a, the short sides of the footprints of the outer columns are on the east and west, while on frs. 18bc they are on the north and south (Steinby 1989, pp. 26-27). A close look at the color photo, however, shows that had the engraver completed the E side of the square columns where he was supposed to, the columns would have been almost perfectly square.
Third, Steinby notes that on fr. 18a the outer columns wrap around the SE corner of the podium--standard for a peripteral temple--but on drawings of the lost fr. 18b, there appear to be no columns along the back, suggesting that the depicted temple was sine postico. However, examination of the Renaissance drawing from which our knowledge of fr. 18b is drawn (reproduced at PM 1960, pl. 9) does not allow for certainty on this point. In general, the Renaissance drawings treat surface incisions on the fragments with a high degree of accuracy but the outlines of the fragments are sloppy. Judging by comparable fragments, it is very possible that fragment 18b in actuality broke off just beyond the final column depicted, where the podium line abruptly ends; then the outline of the fragment was cursorily drawn some way beyond, giving the erroneous impression that more of the fragment survived and that there were no additional columns where expected. (Note that the drawing of the fragment in PM 1960, pl. 21, exacerbates this distance between the end of the incisions and the end of the fragment.)
E. Rodríguez-Almeida argues that fr. 18a (and thus presumably fr. 18fg) must belong to the Severan marble plan, because it is of Proconnesian marble like the rest of the FUR fragments, a type of marble not commonly used in Rome before the Antonine period (see Rodríguez-Almeida 1995-96a for the full discussion).