Fourth in invocations came Quirinus, and fourth in order of precedence was his flamen. But of Quirinus I need say little; there is, on the whole, a consensus of opinion that he was a form of Mars belonging to the community settled on the hill that still bears his name. The most convincing proof of his identity with Mars (though identity is doubtless too strong a word) lies in the well-known fact that there were twelve Salii Collini, _i.e._ belonging to the Collis Quirinalis, occupied with the cult of Quirinus, answering to the twelve Salii Palatini of the cult of Mars. "Quid de ancilibus vestris," Camillus says in Livy's glowing rhetoric, "Mars Gradive (the particular cult-title of the warlike Mars), tuque Quirine pater?"[277] Now the Quirinal was, of course, _within_ the walls, and the Romans who identified the two deities noted this point of contrast with the Mars-cult; for Servius writes, "Quirinus est Mars qui praeest paci et _intra civitatem_ colitur, nam belli Mars _extra civitatem_ templum habet." In keeping with this is the use of the word Quirites of the Romans in their civil capacity; but unluckily we are altogether uncertain as to the etymology and history of both Quirites and Quirinus.[278] And as Quirinus never became, like Mars, an important property of the Roman people, but was speedily obscured and only revived by the legend of late origin which identified him with Romulus, he is not of importance for my subject, and I may leave him to etymologists and speculators. There is one other deity of whom I might naturally be expected to say something; I mean Juno. But our familiarity with Juno in Roman literature must not be allowed to lead us into believing too rashly that she was one of those great _numina_ of the early Roman State with whom I have just been dealing. She had no special festival in the calendar;[279] her connection with the Kalends she shared, as we have seen, with Janus. She had no special priest of her own; for in spite of all assertions that the flaminica Dialis was attached to her cult, I am convinced that I was right some years ago in maintaining that this is an error, though a natural one.[280] It cannot be proved that she had any ancient temple in the city; for the oldest known to us as strictly indigenous, that of Juno Moneta on the arx, was not dedicated till 344 _B.C._, and we do not know that there was an older altar on the same spot.[281] Assuredly Rome was not in early times a great centre of the Juno cult, as were some of the cities in her neighbourhood, _e.g._ Lanuvium, Falerii, and Veii;[282] and the gradual establishment of her position as a truly Roman goddess may be explained by her appearance in the trias of deities in the Capitoline temple at the end of the regal period, and by the removal to Rome of Juno Regina of Veii still later, after the destruction of that city.