This search for natural protection for the inmobile artillery must have been at the time a constant of its use in the field. Thus, we have to speak rather of an immobile field artillery in Greek times. Another use was the deployment of ship artillery, where feasible, but which should not be discussed here.
Source: La artillería en las batallas campales en el mundo grecorromano
Initially, the situation changed little in Roman times. Artillery was used in the field, but usually only in conjunction with fortified position and ancient 'trench warfare', a speciality of Roman warfare:
QUOTE
Before the introduction of the carroballistae by the Roman army about AD 100, the occasions of catapults’ employment in the battlefield were rather rare. The machines were deployed on the beforehand-prepared position, in addition usually fortified by temporary ramparts and defensive works (Livy, XXXII, 5, 13; 10, 11; XLIV, 35, 9; 22; Caes. BG, II, 8) or placed on hills and heights (Polyaen, II, 38, 2; Caes. BG, VIII, 14; Tac. Hist., III, 23). Besides, Romans have used artillery either for the defence of their field-camp (Caes. BG, VII, 41; Tac. Hist., III, 29) or for the attack of the enemy’s field-works, especially when it was passive (Tac. Ann., II, 20). Static warfare of Caesar and Pompey at Dirrhachium can be served as a spectacular example of such sort of operations (Caes. BС, III, 45, 51, 56, 63). Attempts to station machines on open position, in front of the battle line or at its intervals, likely didn’t lead to a desirable result (Plb., XI, 12, 4).
The real field artillery was created only when the all-metal spring-frames have invented for arrow-shooters and, consequently, battle mass of these pieces of artillery was decreased, which allowed for mounting them on mobile carriage. Ballistae on military carts appeared for the first time on Trajan’s Column (AD 113); these engines have two wheels only (TC 104-5; 163-4) and most likely harnessed with two mules instead of horses. The same two-wheeled artillery cart, with mules harnessed and possibly with a base and stock of arrow-shooting ballista mounted, depicted on the Column of Marcus Aurelius (AD 192). The absence of the machine’s spring-frame can be explained by the fact that it was probably dismantled for transportation and storage.
The earliest literary evidence of employment of ballistae on military carts in the battlefields of the Bosporian-Roman wars referred to the period from the end of third century to the beginning of fourth century AD (Const. Porph. De adm. imp., 53 || 28-44, 131-135). In the battle near the Bosporian city-walls ballistarii of Chersonesos with cheirobolistras on military carts covered false retreat, probably being moved from rear side of the battle line and operated from ambushes. Apparently, the detachment of ballistarii, armed manuballistae, in armed forces of Chersonesos played a rather significant role and according the message of the author of the treatise has made the decisive contribution to a victory in battles near the Bosporus and on the Ister river. The list of materials for manufacture and repairs established to annual output by the emperor Constantine II (Const. Porph. De adm. imp., 53 || 150) specifies that these cheirobolistras most likely were torsion engines with all-metal spring-frame. The list included sinews (neuron), probably intended for manufacture of spring-cord, hemp (kannabon), either for the cord or for the bowstrings, iron (sideron), probably for manufacture of spring-frames, olive oil (elaion), either to lubricate sinews (Philon, Bel., W61) or to preserve iron parts of the machines from rust.
The best evidence for the battlefield use of ballistae contained in late-Roman (latter half of fourth century) treatise of Flavius Vegetius Renatus Epitoma Rei Militari. There he sometimes mentions hand-arrow-shooters – manuballistae (Veg. Epit., II, 15; IV, 22) – and men who worked this weapons – manuballistarii (Veg. Epit., III, 14; IV, 21). The main battle line of the antiqua legio, according to Vegetius, was formed up in six parallel lines (acies). Tragularii stood behind the first line (Veg. Epit., II, 15), operating manuballistae and arcuballistae (probably, being an ancestor of medieval crossbow). Some of the carroballistae and large quantity of manuballistarii were in the fifth line, mixed with an light infantry and all kinds of throwers (Veg. Epit., III, 14), while the large carroballistae with the largest range took position behind the battle line (Veg. Epit., III, 24). In the days of Vegetius for travelling of artillery carts, side by side with the mules, the horses were already used (Veg. Epit., III, 24): carroballistas aliquanto maiores … superpositas curriculis cum binis equis vel mulis post aciem convenit ordinari … Probably, it was associated either with necessity to increase mobility of units of ballistarii or with enlargement of engines itself.
By the way, two carroballistae, depicted on Trajan's Column (TC 104-5) most likely shooting, were also placed behind the battle line. According to Ps.-Mauricius, ballistae mounted on carriages were distributed evenly all over the infantry battle line, however most of these pieces of artillery were placed on wings (Mauric. Strat., XII, B, 18), that quite corresponds to the position of machines in the battle order of Arrianus (Arr. Alan., 19).
Late-Roman anonymous treatise De rebus bellicis contains one more mention of mobile engine. Unlike the previous ones, ballista quadrirotis, harnessed with two armoured horses, they had a four-wheel carriage.
Source: Ildar
The real field artillery was created only when the all-metal spring-frames have invented for arrow-shooters and, consequently, battle mass of these pieces of artillery was decreased, which allowed for mounting them on mobile carriage. Ballistae on military carts appeared for the first time on Trajan’s Column (AD 113); these engines have two wheels only (TC 104-5; 163-4) and most likely harnessed with two mules instead of horses. The same two-wheeled artillery cart, with mules harnessed and possibly with a base and stock of arrow-shooting ballista mounted, depicted on the Column of Marcus Aurelius (AD 192). The absence of the machine’s spring-frame can be explained by the fact that it was probably dismantled for transportation and storage.
The earliest literary evidence of employment of ballistae on military carts in the battlefields of the Bosporian-Roman wars referred to the period from the end of third century to the beginning of fourth century AD (Const. Porph. De adm. imp., 53 || 28-44, 131-135). In the battle near the Bosporian city-walls ballistarii of Chersonesos with cheirobolistras on military carts covered false retreat, probably being moved from rear side of the battle line and operated from ambushes. Apparently, the detachment of ballistarii, armed manuballistae, in armed forces of Chersonesos played a rather significant role and according the message of the author of the treatise has made the decisive contribution to a victory in battles near the Bosporus and on the Ister river. The list of materials for manufacture and repairs established to annual output by the emperor Constantine II (Const. Porph. De adm. imp., 53 || 150) specifies that these cheirobolistras most likely were torsion engines with all-metal spring-frame. The list included sinews (neuron), probably intended for manufacture of spring-cord, hemp (kannabon), either for the cord or for the bowstrings, iron (sideron), probably for manufacture of spring-frames, olive oil (elaion), either to lubricate sinews (Philon, Bel., W61) or to preserve iron parts of the machines from rust.
The best evidence for the battlefield use of ballistae contained in late-Roman (latter half of fourth century) treatise of Flavius Vegetius Renatus Epitoma Rei Militari. There he sometimes mentions hand-arrow-shooters – manuballistae (Veg. Epit., II, 15; IV, 22) – and men who worked this weapons – manuballistarii (Veg. Epit., III, 14; IV, 21). The main battle line of the antiqua legio, according to Vegetius, was formed up in six parallel lines (acies). Tragularii stood behind the first line (Veg. Epit., II, 15), operating manuballistae and arcuballistae (probably, being an ancestor of medieval crossbow). Some of the carroballistae and large quantity of manuballistarii were in the fifth line, mixed with an light infantry and all kinds of throwers (Veg. Epit., III, 14), while the large carroballistae with the largest range took position behind the battle line (Veg. Epit., III, 24). In the days of Vegetius for travelling of artillery carts, side by side with the mules, the horses were already used (Veg. Epit., III, 24): carroballistas aliquanto maiores … superpositas curriculis cum binis equis vel mulis post aciem convenit ordinari … Probably, it was associated either with necessity to increase mobility of units of ballistarii or with enlargement of engines itself.
By the way, two carroballistae, depicted on Trajan's Column (TC 104-5) most likely shooting, were also placed behind the battle line. According to Ps.-Mauricius, ballistae mounted on carriages were distributed evenly all over the infantry battle line, however most of these pieces of artillery were placed on wings (Mauric. Strat., XII, B, 18), that quite corresponds to the position of machines in the battle order of Arrianus (Arr. Alan., 19).
Late-Roman anonymous treatise De rebus bellicis contains one more mention of mobile engine. Unlike the previous ones, ballista quadrirotis, harnessed with two armoured horses, they had a four-wheel carriage.
Source: Ildar
Comments:
1. It is interesting to note that the deployment of mobile Roman artillery in the field, as far as the literary record and surviving contemporary depictions goes, only started with the introduction of the iron-framed arrow-shooter. This may indicate that wooden arrow-shooters were considered too heavy for carts.
I find this rather curious, since the Chinese records show the mobile deployment of cart-based bolt-shooters. Any ideas?
2. The typical Roman mobile artillery unit was the arrow-shooter, not a stone catapult. However, since the torsion type used by the Romans can basically shoot both projectiles, arrow and stones, this differentiation may be academic. On the other hand, the weight of large Roman stone catapults easily exceeded 1 ton, which made a deployment on carts a priori impossible, a military constraint though which probably was not felt by the Romans, since stone throwers are of little effect against soft targets anyway.
3. The battle formation as described by Vegetius with the missile troops standing behind and in between the infantry reminds me of similar Chinese battle orders of crossbowmen.
), these early arrow-shooters appear just to be large crossbows. Sure, their range was probably greater but surely they were slow to load and when fired could still only take out one opponent 
