what was the function of the Roman cavalry?

[L C Cinna]

hmmm the old myth again.

Roman cavalry WAS effective and very very good compared to most of the opponents they faced.

First of all you have to make a difference between the Republican period and the Principate. The myth that Roman cavalry s*cks is based on the republican periode where they were beaten by Hannibal's (nummerically superior and heavier) cavalry.

The cavalry during the empire is very good and very effective. People often think they were nothing special because they were part of the auxiliary forces but that is NOT true. Commander of an Ala Millariae was one of the highest and most prestigious ranks to get. The Separation of legions, auxiliae and auxiliary cavalry is just a matter of better organization. The example I like to give is that nowadays tanks are not part of an infantry regiment as well but the 2 are combined when needed but if you separate them they can operate individually as well.

I copied this from another post I made, hope it helps a bit:

Here some numbers on cavalry strenght from Junckelmann:

Augustus: minimum of 10% of the army is cavalry. ca. 30.000
Hadrian/Antoninus: ca.18% min 75.000. the maximum calculation would be 90.000

For the late army in the Notitia it's harder to get real numbers but estimations are around 45.000-50.000 for the mobile field army plus 80.000 for the rest. So around 25% of the army was cavalry. So you see the numbers increased significantly.

Just to compare: Standard in Hellenistic armies was 10% or below.

Those are the numbers for the Roman cavalry units which were part of the army. Note that on several occasions the Romans hired additional units. One example is the army which was sent to crush the Jewish revolt during the reign of Nero. For example the army under Cestius Gallus in 66CE consisted of 16 legionary cohorts. 6 auxiliary cohorts, 4 Alae quingeniariae (around 2.000 heavy horsemen) PLUS
2000 horsearchers from Commagene
1.500 horsearchers from Judaea
1.200 horsearchers from Emesa
Such allied contingents hired by the Romans would not be part of the calculation done before as they were not a permanent part of the army.

So to bring this a bit further:

while we have numbers for the standing auxiliary cavalry (used in Junckelmann's calculation) which apply for the WHOLE Roman army the example of the Jewish revolt I've just given shows that cavalry contingents were attached to the army in a specific theatre according to what the Romans thought was needed. The whole point in creating legions, auxilias and cavalry units as seperate parts of the army is that you have smaller blocks which you can put together for certain operations much easier (like combining tanks, infantry a.s.o. nowadays)

What I want to say is that while in 66CE between 10 to 18 percent of the regular army was cavalry. this specific army here for example had:

8.000 legionary infantry
4.800 auxiliary infantry
2.000 heavy Roman cavalry
4.700 allied horsearchers
7.500 allied archers
1.000 allied skirmishers

21.300 infantry and 6.700 cavalry, so 28.000 men. this means approx. 24% cavalry which is indeed pretty high.

. The REGULAR cavalry forces were greatly increased over time as I showed in my previous post from 10 to 25% of the whole army. The question which is harder to answer is if the Romans really used more cavalry in battles in general.

One of the reasons for increasing the regular cavalry force is the greater number of enemies imho. In the 1st and 2nd century the Romans usually fought one greater operation at once (Invasion of Britain, Dacian war, Marcomanni war, Septimius Parthian campaign for example) and could just concentrate the needed cavalry forces on one point. For Traian's Dacian campaigns cavalry alae from all over the empire were sent to the theatre and it was no problem to have fewer forces left on the other boarders.

in the 3rd and 4th century where Rome is fighting on several fronts at once they needed more regular cavalry because they couldn't afford anymore to withdraw such huge forces from one area and leave only a limited defence there. but I don't think that the number of cavalry which was used on average during battles was really higher in the 4th century.

I used some numbers for forces put together for Roman operations where the Romans combined infantry and cavalry like they felt was needed.So this means operations where they had time to prepare their forces. (In other situations like the revolt of Boudicca the number of cavalry is sometimes much much lower but these were emergencies where there was no time to prepare a proper army.)

Here 2 battles from the 1st century:

Vespasians force according to Josephus:

Legiones V, X, XV
10 cohortes milliariae (10.000 men auxilia)
13 cohortes equitatae (600 footsoldiers and 120 cavalry each. so 7800 foot, 1560 cav.)
then he names 6 units of cavalry, he doesn't say anything about the size. the minimum would be 3.000 maximum 6.000. probably somewhere in the upper middle so around 5.000.

SO all in all Vespasians army had approx regular 7.000 cavalry
from socii:ca 11.000 light troops, mainly archers and around 4.000 cavalry

Soooo full force around 44.000 infantry and 11.000 cavalry. again 25%

Agricola's force at Mons Graupius 84CE: 8.000 heavy auxiliaries, 5.000 cavalry, number of legions is not known but he had 2 or 3. full number of all forces between 20-25.000 so again 20-25% cavalry.

And 2 battles from the 4th century:

Julian at Argentorate in 357 had around 22% cavalry in his force.

Valens at Adrianopolis in 378 had approx. 25%.


Note that the numbers of cavalry the Romans used in all of the battles I posted in this thread is significantly higher than the standard. Hellenistic armies had much fewer (Ptolemaios at Raphia had 6%, Antiochos had aroun 8%, Alexander had around 16% on his campaign). The Germanic tribes had usually very very few cavalry. So besides the Parthians and some Steppepeople with a horsebased culture the Romans actually used A LOT of cavalry in the imperial period.much more than most others at that time.