The East Under the Julio-Claudians

A Reconstruction of a Roman Legion Praetorium, 3rd Century

Typical Roman Army Fortified Camp (Vindolanda)

The World According to Herodotus -- 430 BC

Hegetor's Ram-Tortoise #9

>>>> Hegetor’s Ram - Tortoise <<<<

This machine is the subject of by far the most detailed of Athenaeus's and Vitruvius's descriptions, but there are some doubts and misunderstandings which have inspired a succession of astonishingly varied reconstructions through the years, and several uncertainties remain.  Athenaeus and Vitruvius both preserve details of a different ram - tortoise from that described by Diade's ram tortoise and borer.  This one apparently devised by an otherwise unknown engineer named Hegetor of Byzantium. In essence Hegetor's machine was a 10m-high tortoise with a central turret, but the reconstruction of this key element is controversial.  The sources give a detailed descripton  of the ramming beam itself, with rope reinforcement and rawhide covering.  It has been suggested that Hegetor worked for Demetrius Poliorcetes, who had a penchant for grandiose machinery, though the connection is more than a little tenuous.  The historian Diodorus Siculus records that, that during the siege of Salamis, Demetrius ‘constructed enormous battering rams and two ram-carrying tortoises’, and that, at Rhodes his ram-tortoises were ‘many times larger’ than the ditch-filling tortoises that preceded them.  The ramming-beams are said to have been 120 cubits (53.2m) long, the very length that Athenaeus attributes to Hegetor’s battering ram, but the practicality of such a long beam has been questioned.

Athenaeus claims that Hegetor’s 120-cubit ram was rectangular in cross-section, and tapered from a rear-end 2 ft. (59cm) ‘thick’ (by which he must mean the ‘height’ of the beam) and 1 1/4 ft. (37cm) broad to a tip 1 ft by 3/4 ft (36.9cm x 22.2cm).  Vitruvius gives a completely different set of dimensions: the length, he says, was 104 ft (30.75m), and the rear end was 1 1/4 ft by 1ft (36.9cm x 29.6cm), tapering to a 1 ft by 3/4 ft (29.6cm x 22.2cm) at the tip. (The anonymous Byzantine muddies the waters  by combining Athenaeus’s statement of length, with Vitruvius’s dimensions for the thickness of the beam.)

Shramm believed that  a 50m beam would buckle, and the ends would drag on the ground, making the whole contraption unusable.  He proposed that Athenaeus’s text should be amended to read 120 ft (35.5m), considerably shorter than 120 cubits, but still some way from Vitruvius’s figure.   (The alternative approach adopted by the Greek scholar Sir William Tarn, who postulated  that a special short cubit of around 34cm was used in Macedon, takes us even further from Vitruvius.)

A better solution, which actually goes some way towards reconciling the two sources, is to assume  that the Greek text of Athenaeus has been corrupted during transmission down through the ages, and that an original statement of ’70’ (hebdomekonta) cubits was miscopied  as ’120’ (hekatoneikosi) cubits.  A length of 70 cubits  (31m) is very close to Vitruvius’s measurement (Precisely how Diodorus came upon the measurement of 120 cubits for Demetrius’s battering rams remains unknown; perhaps both he and Athenaeus drew upon a common source, which had already been corrupted by their day.)

The beam was suspended from a rope cradle high up in the turret, and stabalized by rawhide-covered chains running around a pair of rollers.  The ramming beam was capped with an iron tip, like the beak of a warship.  Basically, this was a hollow lump of iron, designed to fit over the end of the beam, but was secured by four ten-cubit (4.4m) iron strips which trailed back along the beam like streamers and were nailed into position. (Vitruvius calls these streamers lamminae, which is the usual  term for a strip of metal, but Athenaeus calls them ‘iron spirals’, implying that they were wound around and along the beam.)  The beam was further reenforced with ropes, using a technique well known in the ancient world for bracing the hulls of ships, and then completely wrapped in rawhide, a necessary protection against fire because it was entirely exposed above the level of the tortoise.

The tortoise itself was similar in size to Diade’s model.  Athenaeus gives the dimensions as 42 cubits (18.62m) long and 28 cubits (12.42m) wide.  Vetruvius’s version, at 60 ft by 13 ft (17.7m x 3.8m), is obviously wrong, and is usually corrected by emending the manuscript 13 (XIII) to read 42 (XLII); 42ft is the equivalent of 28 cubits, and thus matches the width quoted by Athenaeus.  Vitruvius’s length of 60 ft is 3 ft short of Athenaeus’s 42 cubits, but this may be a manuscript error.  The machine ran on eight wheels, 4 1/2 cubits (1.99m) high and two cubits (0.88m) thick which according to Vitruvius, comprised three layers, each 1ft thick pegged together with dowels  and fastened with iron bands.  (Here again, Vitruvius uses the word lamminae.)  Unfortunately, as with the other eight-wheeled machine, the helepolis of Epimachus, we are not told the configuration of the wheels, but positioning them four abreast  would distribute the massive weight of the machine more evenly.    Also a machine built to Athenaeus’s dimensions and following the principles of the ditch-filling tortoise would have rested on an undercarriage some 16 cubits (7.10m) square; consequently, there would not have been space for four-in-line wheels, and they must have been arranged four abreast.

Like the ditch-filling tortoise, the ram-tortoise would have had a hipped roof meeting at the top in a transverse ridge.  The whole machine would then be boarded over and covered with a fireproof layer.  As with Diade’s ram-tortoise, this style of construction resulted in a ‘middle floor’ (mese stege, or media contabulatio) which caused much confusion amongst those trying to reconstruct the machines.  In the case of Hegetor’s tortoise, this second storey had floor space of 16 cubits square (7.10m) and headroom of 8 cubits (3.55m) up to the roof ridge.  Athenaeus says that it accommodated an artillery position (belostasia) and Vitruvius explains that scorpions  and catapults were located there.  Firstly, this contrasts with Diade’s version, where the artillery occupied a third storey turret, rising above the the middle floor; and secondly it implies that there were windows through which the catapults could fire.  There seems an altogether more practical arrangement than Diade’s rather fragile and cramped turret.

But even though Hegetor deployed the necessary supporting artillery in the middle floor, he did not entirely dispense with a central turret.  According to both Athenaeus and Vitruvius, the working of the ram somehow depended upon a frame, which rose through the middle floor to project some 4m above the roof ridge, and incorporated a crow’s nest at the top.

The potential firepower of the tortoise can be estimated by comparing the middle floor area with the space requirements of small to medium-sized catapults, but both the sloping penthouse construction and the timber uprights of the turret must be taken into account.  The first would have limited the usable area to the very middle of the floor, and the second divided this across the middle.  The rear was best reserved for ladders, allowing the crew to move about the machine.  This would leave enough space in front for three 3-span arrow throwers side-by-side.  Positioned roughly 9m above ground level, the catapults would have enjoyed a superior vantage point for targeting the average battlements. The construction of the turret is not explained and so we must resort to conjecture.  The sources mention four robust 24 cubit (10.64m) uprights , and another two 30-cubit (13.3m) uprights.  The latter pair supported a device consisting of two rollers, sitting side by side.  In the words of Vitruvius, ‘ the ropes which held back the ram were fastened around these [rollers]’.  However, this is not the ram holder itself (kriodoche, or arietaria machina); that component, as both authors explain lay somewhere in front of the double rollers.  In addition, whereas Diade’s battering ram seems to have rested upon the ram holder, Hegetor’s ram was suspended in the middle by a thick hank of ropes.

The German scholar Otto Lendle has devised the most plausible interpretation of this enigmatic structure.  The ram-holder, being the suspension point for the ramming beam, would have been centrally located in order to distribute the weight most efficiently, and Lendle fixes it between the four uprights of the turret.  At this point there would have been less than 2m clearance above the roof of the tortoise, so the suspension-tackle must have been relatively short, to prevent the ramming beam from snagging on the roof ridge.

It has been conjectured that the ropes running from the rollers were in some way instrumental in altering the height of the ram head, and indeed both Athenaeus and Vitruvius suggest that the enemy wall could be battered up to a height of 70 cubits (31m).  This is an extraordinary claim, given that the battering ram  was suspended only about 26 cubits (11.5m) above ground.  In any case, 70 cubits greatly exceeds the usual range of fortification heights: even operating horizontally, the beam would have been higher than most town walls.  Sadly, neither author gives any idea of how the battering ram was operated.  The necessary pendulum motion would have required  some means of pulling the beam backwards, and there were perhaps several ropes attached to its rear end, to be pulled by hauling crews on the ground.  Furthermore, the length of the beam’s suspension would have restricted it to short blows.  It is not clear how successful this method would be if the beam were set at any angle other than the horizontal, and it must be admitted that many aspects of Hegetor’s ram-tortoise remain a mystery.

Reference:

Duncan B. Campbell, Brian Delf (Illus..), “Greek and Roman Siege Machinery 399 BC -- AD 363,” New Vanguard-78, (Osprey Pub., 2003)

Respectfully Submitted;

Marcus Audens

  

Hegetor's Ram-Tortoise #8

This view shows the artillery loft through a stripped side panel;
 catapult to the right, and ropes / rope stanchions
to the center.  Above the artillery loft you can see
 the open plank deck used for rigging the ram beam
in readiness for use.  You can also see the ladder into
the lookout box.  Note the structure and reenforcement
of the ram beam. 

Hegetor's Ram-Tortoise #7

This view, again shows the view of the interior
first level box and flooring, the front plank shielding,
side plank shielding,  and the ladder to the
artillery loft.

Hegetor's Ram-Tortoise #6

This view shows the full length of the ram beam
the full size of the lookout box and the base as well
as the artillery deck.  Note the double-hipped
 roof planking.

Hegetor's Ram-Tortoise #5

This view is looking directly at the front of the ram-tortoise.
Note the slanting sides and the plank section directly
in front.  Of course, all sides would have been planked
over, but these openings allow one to see
 the interior more clearly.  The wooden planking would
have been boarded, padded, and underlaid with a
woven blanket of small sticks / vines, and over all that,
wet hides to minimise impact damage to the hipped roof
by rocks of fire arrows.

Hegetor's Ram-Tortoise #4

This picture is looking directly into the artillery loft.
Behind the dart catapult are the ropes and the rope
stanchions supporting the ram beam.  On the right
side you can see the protective shield for
the ladder to the lookout box.  Note the
sloping sides of the loft.

Hegetor's Ram-Totoise #3

This photo shows the three beams supporting the
outside planking, the floor of the base, and the
ladder to the artillery loft.  A man's height would
be equal to about seven rungs on the ladder.

Hegetor's Ram Tortoise #2

Ram interior showing the floor of the base and the ladder to the
upper artillery loft and lookout box.

Hegetor's Ram--Tortoise #1

Sambuca # 6

This view of the Sambuca is over the rock counterweight and up the soldier's passage
from their entry point.  Below, you can see the winch and handles which turns the
hoisting screw and lifts the Sambuca up to its attack height.

Sambuca # 5

This view is from the front of the Sambuca, looking into the inside and down the length of the shielded soldier's passageway.

Sambuca # 4 - Top View

Sambuca #3

This view shows the Sambuca lowered to its lowest height.  The front ladder is touching the ground.

Sambuca #2

Sambuca #2, which shows the Sambuca at it's height ready to assail a city wall.
This view shows the soldier's passage enclosed by planking.  On top of the planking
would be a thick layer of wicker-work topped by wetted or fresh animal hides.
A man's height would be approx. six rungs of the fron t ladder.

Sambuca

The most famous examples of the sambuca (or sambyke) were mounted 

aboard ships, where they resembled giant laddered drawbridges for thetransfer of marines onto seawalls  of coastal towns.  However, the sambuca that Biton, a writer-engineer writing at sometime between 231 and 133 BC attributes to Damios, an otherwise unknown engineer from Kolophon in present-day Turkey, is quite different.  First, it was designed for use on land, and second, it utilized an innovative vertical screw to alter the elevation of the ladder.  Biton says that the sambuca itself , a 60 ft (18 m) ladder with an assault platform at one end, and a counterweight at the other, sat on a 'trestle' (killibas); the trestle was fixed to a 27 ft. (8m) undercarriage.  Biton's ladder clearly has sidewalls, 'so that men climbing up will make the ascent confidently', and a widened jumping off area at the top.  Biton's description of the Dammios sambuca is extremely concise.  For the undercarriage he simply gives the beam dimensions of 3 ft. x 2ft. x27 ft. (0.9m x 0.6m x 8m) and notes that the wheels were 3 ft. (0.9m) high; here a rectangular under carriage with six wheels is assumed.  The main component , a 60 ft. (18m) ladder was hinged to the rear of the tresle, which supported a centrally located vertical screw; the screw's function was to raise and lower he main ladder.  It is reasonable to suppose that as much of the machine as possible was boarded in, in order to protect the crew.  Biton specifies that the trestle was 14 ft. (4.2m) high, whereas the screw was 15 ft. (4.5m) long:, consequently, when fully turned , it would project 1 ft. (0.3m) above the trestle.  Assuming a forward projection of around 12m for the ladder, with the screw fully turned, the assault platform rose almost 9m above ground level.

Maintaining the machine's stability would have been a delicate task.  Withe the main ladder in it's horizontal position, the vertical ladder at the front was perhaps to support the machine while the assault team took up position on the forward platform.  Their presence there would have severely unbalanced the machine, so the counterweight at the rear must have been intended to restore its equilibrium.  Only then would the main ladder have been elevated  and the machine rolled forward.

Reference: Duncan B. Campbell, Brian Delf (illus.), "Greek and Roman Siege Machinery, 399  BC --- AD 363," New Vanguard-78, (Osprey Pub., 2003), Pges 24 and 44. (ISBN 1-84176-605-4)      

Corvus Demolitor #3 (Top View)

Corus Demolitor #2

Corvus Demolitor #1 ("Iron Hand -- demolition raven" )

The ancient sources mention defenders using grapnels  or iron hands for ensnaring men

 and siege machinery.  Diades (a pupil of Phillip's engineer Polyidus) is credited with
the invention of one of these, the corvus demolitor or "demolition raven."
The image in the reference was imaginatively reconstructed by de Folard (authors
collection).  This device was used by attackers to pull stones off a curtain wall.   The
height of a man is about one/fifth of the length of the gripping pole (red head).  As
shown in the book's picture, it seemed to work best when there was no-one on the
defensive wall.  The movement needed to swing the "iron hand" into a solid gripping
situation and then to pull the stone off the wall seems to be more than the defensive
mantlets can screen!  Vetruvius describes the above machine built for
Alexander the Great.

Reference: Duncan B. Campbell, Brian Delf (illus.), "Greek and Roman Siege
Machinery 399 BC -- AD 363," New Vanguard-78, (Osprey Pub., 2003), Page 16
 and 34.  (ISBN 1-84176-605-4)

Ditch-Filling Roman "Tortoise." #5

This large opening allows you to see the interior from the side, and
also shows the extension for use over a ditch.

Ditch-Filling Roman "Tortoise" #4

The two openings in the planking allow you to see the interior of the "Totroise."

Ditch-Filling Roman "Tortoise" #3

The slanted side actually covers the ditch, and allows the men to fill
in the ditch while protected from materials being thrown at them from the wall.

Ditch Filling Roman Tortoise #2

The dark square is the door through which the men entered carrying
 the fill material.  The cut-away sections allow you to view
the interior of the machine.

Ditch-Filling Roman "Tortoise" Siege Machine

This instrument was designed and built to fill in ditches around a fortress or 

fortified city.  The men entered from the rear (to the right) with their baskets or 

sacks of ditch fill (sod, fachines, rock, or dirt).   The machines top and sides 

would first be planked solidly and then covered with wet and raw hides against 

any attempts to use fire against the machine.

According to Athenaeus, the ditch-filling tortoise could be rolled sideways, as 

well as backwards and forwards, probably by briefly raising each corner in turn,

and changing the orientation of the axle (see page 16 illus.).  Without 

experimentation, it is unclear how this was accomplished, but the large frame

would have allowed a dozen or more men to congregate around each wheel 

assembly  and jointly take its weight.  With the machine in position, there 

would have been ample space in the interior for men to work unhindered, 

evening out depressions in the ground.  The gap between the rafters and the 

ground would have been sufficient to allow baskets of earth and rubble to be 

brought in at the rear, from which they could be dragged forward for the task 

of ditch-filling.

The scene (Plate A) is based on the siege  of Halicarnassus in 334 BC, when 

Alexander  was obliged to fill the newly cut 13. 5m wide, 7m deep defensive 

ditch, in order to bring up heavy machinery.  The remains of the fortifications 

suggest that the curtain was a single-line, single-story affair, but not enough 

survives for an accurate picture.  Here, the reconstruction is based on the defences

of Paestum (Italy), generally thought to have been built around 330 BC.   

The approximately 9m high wall is crowned by a closed battlement with 

shuttered windows, as a defense against escalade.

Reference:  Duncan B. Campbell, Brian Delf (illus.), "Greek and Roman Siege 

Machinery, 399 BC -- AD 363," New Vanguard-78, (Osprey Pub., 2003), 

Page 44.  (ISBN 1-84176-605-4) 

"Ship's Prow Tortoise" #7

This view of the machine shows it right side up.  Note the two iron plates on the prow to deflect any flaming debris or logs.  In the vee of the photo, one can see the hinge assembly for the support brace timber. 

"Ship's Prow Tortoise" #6

This view of the Tortoise shows the machine upside down, showing the wheels, axles, and crossbrace.

"Ship's Prow Tortoise" #5

This view of the tortoise right side up, showing the wheels on the ground and the anchor timber in place.  The height of the two wings from the ground was about seven feet.

"Ship's Prow Tortoise" #4

"Ship's Prow Tortoise" #3

This view of the tortoise, lying on one side, and showing the iron plate protecting the
prow of the machine.

"Ship's Prow Tortoise" #2

This view of the machine is laying on one side showing the four transfer wheels , two axles and the center timber with held the tortoise in place against heavy materials (boulders, logs, etc.) rolled down the hill  toward the instrument.

"Ship's Prow Tortoise" (Roman Siege Machine) #1

This machine was used on slopes leading up to the walls of the city or
fortress under siege.  It's purpose was to protect those men working on
the slope area digging trenches and erecting various siege field
fortifications and other siege machines.  The extreme prow of the
machine was sheathed in iron in order to shunt aside any flammable
engines used by the defenders such as barrels of oil and round packages
of straw rolled down the hill in an effort to discourage the attackers.
This picture view of the tortoise shows it upside down.  The
Scene (Plate E) shows the 'Ship's Prow Tortoise' with associated
earthworks.  Appolodorus opens his 'Poliorketika' with the above scene
of an assault on a hill town, in which he emphasises that the
besiegers must guard against heavy objects being rolled downhill.
He lists in particular, tree trunks, round boulders, heavily laden wagons,
and barrels filled with gravel or earth.  First, he recommends the digging
of a 5 ft. (1.5m) deep ditches, running obliquely downhill; the spoil
from the ditches forms a rampart, to break the momentum of the objects,
and the ditches are intended to channel them away from the main
besieging force, waiting further downhill.  Next, he explains that
the men digging the ditches should be protected by a slanting palisade
line, boarded over and interwoven with branches to form an 'outwork'
(proteichisma).
Finally, the key element in the scheme is the 'tortoise shaped like a
ship's prow'.  Appolodorus' brief description suggests a vertical-sided,
open-topped shelter with a triangular ground plan, arranged so
that the apex, facing uphill, would deflect rolling objects to either
side.  Here it is assumed that the walls would be sufficiently high to
conceal the soldiers crowded inside.

This scene (Plate E) is based on the siege of a hilltop stronghold that
 appears on Trajan's Column and is perhaps intended to represent
the Dacian capitol, Sarmizegethusa.  It has been assumed that the
polygonal masonry of the murus  Dacicus was surrmounted by a
timber breastwork, sections of which could be easily removed to
allow heavy objects to be rolled down against Apollodorus'
ship's prow tortoises.

Reference: Duncan B. Cambell, Brian Delf (illus.), "Greek and
Roman Siege Machinery, 399 BC -- AD 393," New Vanguard-
78, (Osprey Pub., 2003), pages 45-46. (ISBN 1-84176-605-4)   

Testudo, End view, (note the side boards on right side)

Testudo, Top View, (Note the hides / skins on roof)

Roman Testudo

VINEA, mantlet: The sides were covered with skins or wisker work when required.

The TESTUDO was similar to the VINEA, but squarer and more stoutly built, with
 a sloping roof and/or a shutter on the side next to the enemy's walls: it
was covered over with hides or other non-inflammable substances.

The MUSCULUS was longer, lower, and narrower than the VINEA, forming  a
covered gallery, pushed up at right angles to the enemy's walls.

Reference:

Caesar, G. P, Goold (ed.), H. J. Edwards (trans.), "The Gallic War," (Loeb Classical Library, LCL 72), p. 631 

Ballista and Ballista Bolts #2

Above is a model of an ancient Ballista and the bolts that it fired.  The two
arms attached to the "Bow String" are powered by skiens of fiber made
up from horse hair and silk.  The skiens are twisted as the bowstring is moved
 back and the trigger releases the bolt.  This model projected a bolt about 35 feet.  

Ballista and Ballista Bolts #1