The results of the present paper show that the orientation of Roman towns in Italy is not random: it
comprises two “families”, one lying in the sector within ten degrees SE, the other near the winter
solstice sunrise. Orientation of some towns to the sunrise in dates corresponding to important
festivities of the Roman calendar, in particular Terminalia, looks also probable.
The existence of astronomical orientations confirms statements made by many Roman writers
themselves, and raises the problem of the symbolic meaning of the castrum layout.
Sin duda se trata de un estudio muy interesante, aunque todavía está por ver su relevancia.
Más info en el paper del autor (pdf).