3. The tribes of Sparta were again divided into obæ, which are also called phratriæ.(313) The term _phratria_ (φρατριὰ) signified among the Greeks an union of houses, whether founded upon the ties of actual relationship, or formed for political purposes, and according to some fixed rule, for the convenience of public regulations. Thus the word _oba_ comprehends houses (γένη, _gentes_), which were either really founded on descent from the same stock, or had united themselves in ancient times for civil and religious purposes, and afterwards continued to exist as political bodies under certain regulations.(314) The Spartan obæ appear to have likewise been local divisions, since the name ὠβὰ, _i.e._, οἴα, signifies _single_ hamlets or districts of a town; although in the case of Sparta it is not evident what relation they bore to the five divisions of the city, of which we have spoken above.(315) It should be, moreover, observed, that this does not prevent us from supposing that, as in the parallel case of the phratriæ, the obæ contained the houses; since we may be allowed to infer with great probability, from the simple and coherent regularity of the Spartan institutions, that the tribes had taken possession of particular districts of the town, and that these were again divided into smaller partitions, according to the obæ; a conjecture which, perhaps, will be confirmed by the statement, that a place in Sparta was called Agiadæ:(316) now this was the name of one of the royal families, which, as being an oba, appears to have given its name to one district of the town. The obæ were thirty in number;(317) that is, there were ten of the Hyllean, ten of the Dymanatan, ten of the Pamphylian tribe. Of the Hyllean, two must have belonged to the royal families of the Heraclidæ. For since the councillors, together with the kings, amounted to thirty, and as this number doubtless depended upon and proceeded from that of the obæ, it follows that the two royal families, although springing from one stock, must nevertheless have been separated into two different obæ, of which they were in a manner the representatives. And if we proceed to conclude in this manner, we shall be obliged, since there were Heraclidæ, exclusive of the kings, in the gerusia,(318) to suppose that there were, besides these, other Heraclide obæ in Sparta; although I am not of opinion that all the Hyllean houses derived themselves from Hercules, and were considered as Heraclidæ.