In conducting sieges the Romans showed their practical genius combined with ruthless thoroughness. If a place could not be overcome by the initial assaults or the inhabitants persuaded to surrender, it was practice of the Roman army to surround the whole area with a defensive wall and ditch and spread their units around these fortifications. This assured no supplies and reinforcements got to the besieged as well as guarding against any sorties of an attempt o breaking out.
There are several examples of efforts being made to cut off the water supply. Caesar was able to take Uxellodunum by concentrating on this target. First he stationed archers who maintained a steady fire on the water carriers who went to draw from the river which ran round the foot of the hill on which the citadel stood. the besieged then had to rely entirely on a spring at the foot of their wall. But Caesar's engineers were able to undermine the spring and to draw the water off at a lower level, thus forcing the town to surrender.
Siege weapons were varied and ingenious inventions, their main object being to effect an entrance through the gates or walls. Gateways were usually the most heavily defended positions, so that it was often better to select a point along the walls. First, however, the ditches had to be filled with hard packed material to allow the heavy machinery to approach the foot of the wall. But the soldiers manning the wall would try to prevent this by firing their missiles at the working party. to counteract this the attackers were provided with protective screens (musculi) which were lined with iron plates or hides. The musculi provided some protection but not hardly enough. So constant fire had to be directed against the men on the wall to harass them. This was managed by bringing up stout timber towers, higher than the wall, so that men on their tops could pick off the defenders.
The Siege Tower
The ram was a heavy iron head in the shape of a ram's head fixed to a massive beam which was constantly slung against a wall or gate until it was breached. There was also a beam with an iron hook which was inserted into a hole in the wall made by the ram and with which stones would be dragged out. Further there was a smaller iron point (terebus) used for dislodging individual stones. The beam and frame from which it was swung were enclosed in a very strong shed covered with hides or iron plates, mounted on wheels. This was called a tortoise (testudo arietaria), since it resembled this creature with its heavy shell and head that moved in and out.
Under the protection of the towers , most likely in protective sheds, gangs of men worked at the foot of the wall, making holes through it, or digging down to get underneath it. Excavating galleries under the defences was common practice. the purpose was to weaken walls or towers at the foundations so that they collapsed. this was of course much more difficult to do without the enemy becoming aware of it.
At the siege of Marseille the defenders countered attempts to tunnel under their walls by digging a large basin inside the walls which they filled with water. When the mines approached the basin, the water flowed out, flooding them and causing them to collapse.
The only defence against the Roman's massive siege engines was to destroy them either by fire missiles, or by sorties made by a small, desperate body of men who'd try to set fire to them or turn them over.
The Roman army used several types of powerful siege weapons for discharging missiles, the largest was the onager (the wild ass, because of the way it kicked out when it fired). Or so it was called from the late third century AD onwards.
When being moved with a legion it would be on a waggon in it's dismantled state, pulled by oxen.
Apparently there was an earlier version of this catapult, known as the scorpion (scorpio), although this was a considerably smaller less powerful machine. Onagri were used in sieges to batter down walls, as well as by defenders to smash siege towers and siege works. This explains their use as defensive batteries in cities and fortesses of the late empire.
The stones they hurled naturally were also effective when used against the densely packed lines of enemy infantry.
Another infamous catapult of the Roman army was the ballista. In essence it was a large crossbow, which could fire either arrows or stone balls.
Various shapes and sizes of the ballista were around.
Firstly, there was the large basic ballista, most likely used as a siege engine to fire stones, before the introduction of the onager-type catapults. It woudl have a practical range of about 300 metres and would be operated by about 10 men.
There were more nimble, smaller sizes, including one dubbed the scorpion (scorpio), which would fire large arrow bolts. Also there was the carro-ballista which essentially was a scorpion-sized ballista mounted on wheels or a cart, which could therefore rapidly be moved from one place to another, - no doubt ideal for a battle field.
The most likely use for the bolt-firing scorpio and carro-ballista would be on the flanks of the infantry. Used in much the same way as modern machine guns, they could fire across the heads of
their own troops into the enemy.
The large bolts varied in length and size and were equipped with various types of iron head, from simple sharp tips to crested blades.
When on the march these mid-range catapults would be loaded on waggons and then drawn along by mules.
Other, more strange versions of the ballista existed.
The manu-ballista, a small crossbow based on the same principle of the ballista, could be held by one man. No doubt it could be seen as the forerunner of the hand-held medieval crossbow.
Further there has also been some research done into the existence of the self-loading, serial-fire ballista. Legionaries either side would continuously keep turning cranks which turned a chain, which operated the various mechanisms to load and fire the catapult. All that was needed was for another soldier to keep feeding in more arrows.
The estimates regarding the numbers of these machines which a legion would have to draw on are wide-ranging.
One one hand it is said that, each legion had ten onagri, one for each cohort. Apart from this each century was also allocated a ballista (most likely of the scorpion or carro-ballista variety).
However, other estimates suggest that these engines were anything but widespread and that Rome relied for more on the ability of its soldiery to decide matters. And when used by legions on campaign, the catapults had simply been borrowed from forts and city defences. Hence there would be no regular spread of such machines across the troops.
It is hence hard to establish how widespread the use of these machines truly was.
One term causing confusion with these catapults is the 'scorpion' catapult (scorpio). This derives from the fact that the name had two different uses.
Essentially the catapults used by the Romans were largely Greek inventions. And one of the Greek ballista type catapults at first appeared to be called 'scorpion'.
However, also the smaller version of the 'onager' was given that name, most likely as the throwing arm, reminded of the stinging tail of a scorpion. Naturally, this causes some degree of confusion.