The provision of stables within Roman forts is taken for granted, but it has not yet been proved conclusively that all the horses of mounted units were stabled inside their forts. Study of many forts is considerably limited by lack of definite information as to the size and type of garrison, so that in some cases it is not possible to say whether there were any mounted troops present at all. Even when it can be established that the occupying unit was an ala or an equitate cohort, there is still uncertainty about the exact numbers of horses involved. Furthermore, there are very few Roman forts where the entire area has been excavated, and so most authors have to rely on conjecture, usually designed to prove that on grounds of spacing there would have been room for however many animals the excavator thought there should have been in the unit in question. This mostly depends in turn on the tentative identification of one or two buildings as stables, followed by the expectation, without proof, that similar blocks existed in unexcavated areas of the fort.