(La Cite Antique) by Fustel de Coulanges (1830-1889). The title could more appropriately be The Ancient Cities of Greece, Rome, and the Hindus, since he does not really deal with the constitution of the cities of Persia or Egypt--possibly because of a lack of information through translation issues and fewer sources. I suspect he would have found substantial points of difference with his thesis in those cities.
His thesis is that the family, with its attendant ancestor worship, was the foundation of the clan and eventually the city, and that what bound all of these together was the worship of their ancestors.
The priest was the father, his successor his son, and everything--all ancestral lands, rights, authority came though the rituals about the ancestral tomb and the sacred fire. When families clustered, they retained individual worships, though the city would develop its own sacred fire. The eldest son gets it all; the rest of the family are subsidiary branches.
Without a recognized sacred fire, you have no right to property. In fact, the family owns the land, and the family is governed by the patriarch. Foreigners are suspect, without religion or rights; laws don't really apply to them.
Of course this develops over time, and one winds up with a city with a tiny aristocracy of priest/proprieters, a large number of clients of the same, and of course slaves.
That can evolve further, as with Rome, to include a pleb population as well--not clients and not slaves.
Stability gets to be a problem here, and cities begin to undergo revolutions (the poor side with kings against the aristocrat/priests), and the old religion begins to matter less. He describes the rise of Rome, attributing part of it to Roman early efforts to appropriate the religions and ethnicities of their neighbors, making alliance (and eventual domination) easier, and describes the effects of the empire effectively removing the religious/political rites that held conquered ancient cities together. Eventually the only citizenship that mattered was Roman, which became easier and easier to acquire (up to Caracalla's decree that gave it to every free man).
He ends with the rise of Christianity and the replacement of local gods with a universal one. The gods of the city vanished.
I'll assume that his historical references are all correct. The thesis is interesting, and plausible--within some limits.
As he says himself, Spain and Gaul didn't have these kinds of cities, so Rome established them after the conquest. He doesn't mention Egypt at all.
Over and over he hammers home the point that without the sacred fire and ancestor worship, you had no religion. But I don't think that's true. It doesn't quite match human experience: there were other gods than those of the hearth, even in the same land, and worshippers of one would have had some kind of common bond. The Eleusinian Mysteries is well known. (not sure how old it is, though)
Without a better background in the field (and no way to read primary sources myself) I take this as a partial description of the formation of some of the cities of antiquity.
Have a look yourself.
The Kindle copy I have looks like somebody quit checking for OCR typos somewhere around part 3.
