Line relief whilst under attack

[Justin]

Hello everyone,

Not being well-read in ancient sources I would like to get some input
those who are.

I've always been curious as to how the Republican legion line relief
system functioned. Working on the assumption that the Hastati
initially deployed in a continuous line whilst the Principes doubled
centuries behind them, leaving wide gaps in their line for the Hastati
to pass through, the question naturally arises: how did the Hastati
retire through the gaps in the Principes line whilst under pressure
from the enemy, without the enemy troops themselves pouring through
the gaps and cutting the the fragmented Principes and exhausted
Hastati to pieces? I take it for granted that isolated blocks of
infantry are not a suitable formation for fighting a continuous enemy
line.

My initial theory is that the Hastati and the enemy they were facing
(Gauls or whoever) fought to exhaustion and then broke off a few yards
from each other for a breather. This pause in the fighting was enough
time for the Hastati to quickly retire through the gaps which the rear
Principes centuries would then fill, before the enemy had gathered
their wits enough to charge. On reflection however, the theory doesn't
hold water. The enemy, seeing the momentary weakness in the Roman
formation, would surely charge, fatigued or not, and wreck the
exchange manoeuvre.

Then I had this idea, which I would like to bounce of the group.
Infantry lines in Antiquity were organised by files and manoeuvred as
files, not as ranks. If a line turned into column (as did Scipio's
Triarii at Ilipa) the outermost files would advance then turn right or
left, followed by the others, as opposed to every man individually
turning in place and following the man to his right or left (thanks
for that input Patrick!).

With this in mind it seems natural to me that when the rear centuries
moved up to plug the gaps between the front centuries in the Principes
line, they would do it file by file and not as a single block. In this
manner the gaps would gradually narrow as the files one by one filled
them in, until the last files closed them completely. Whilst this was
going on the Hastati would gradually fall back through the narrowing
gap, but leaving always enough troops to cover the front entrance of
the gap until it had closed completely wherein the last few Hastati
would filter between the files of Principes and join the reconstituted
Hastati line to the rear.

At no point in this process would the gaps be exposed to enemy
exploitation. The process can be carried out even whilst the line is
under attack by the enemy, the Principes gradually taking over from
the Hastati whilst both maintain a continuous front.

I have included some jpegs in the Files section to illustrate how I
hypothesize this happening.

Question is: are there any references in the primary sources that
would tend to confirm this hypothesis? Plausible doesn't mean
substantiated, and I would like to know if any of the ancient authors
(or archaeological findings or anything related) confirm that I am on
the right track.

Thanks in advance.

Justin

[Mark]

Justin:

I am not an expert by any means, but as far as I know there are no
Classical sources that deal with this issue in explicit detail. As a
result a conclusive answer is likely to remain unattainable.

The principal sources are Livy and Polybius. Neither is really a
"primary" source, though Polybius is of course much closer to being
one than is Livy.

Livy says this (8.8.9-12, tr. Foster):

"If the hastati were unable to defeat the enemy, they retreated slowly
and were received into the intervals between the companies of the
principes. The principes then took up the fighting and the hastati
followed them. The triarii knelt beneath their banners, with the left
leg advanced, having their shields leaning against their shoulders and
their spears thrust into the ground and pointing obliquely upwards, as
if their battle-line were fortified with a bristling palisade. If the
principes, too, were unsuccessful in their fight, they fell back
slowly from the battle-line on the trarii. (From this arose the adage
'to have come to the triarii,' when things are going badly.) The
triarii, rising up after they had received the principes and hastati
into the intervals between their companies, would at once draw their
companies together and close the lanes, as it were; then, with no more
reserves behind to count on, they would charge the enemy in one
compact array."

And Polybius says (18.32, tr. Shuckburgh):

"The Romans do not, then, attempt to extend their front to equal that
of a phalanx, and then charge directly upon it with their whole force:
but some of their divisions are kept in reserve, while others join
battle with the enemy at close quarters."

There are also some references to relief manoeuvres in accounts of
battles from the period after the introduction of the cohort (e.g.
Caesar, Civil Wars, 1.46-46, 1.83, 3.94 and Gallic War 2.25) but again
there is not an explicit description of the movements on the ground.

In arguing for a vision of ancient infantry combat as a series of
tentative and sporadic clashes, Prof. Sabin made this observation
("The Face of Roman Battle", JRS 90 (2000): 1-17 at p. 16):

"Envisaging Roman infantry clashes as a natural stand-off punctuated
by periodic and localized charges into contact is of particular
assistance when trying to explain the operation of the Roman multiple-
line system. In a situation of pervasive mutual deterrence, it is much
more plausible to imagine forward sub-units fighting with gaps in
their line, covered at a distance by supporting sub-units whose
presence would dissuade the enemy from breaking his own line by
advancing into the gaps concerned. We know that such deterrence could
operate at the grand tactical level, since the strong Punic centre at
Ilipa watched impotently while its flanks were destroyed, rather than
advancing into the massive box formed by Scipio's Roman flanking
formations and his refused Spanish centre (Polybius 15.14). A similar
process on the much smaller frontages of maniples and cohorts seems
even more plausible, and would allow the second line to move up
smoothly into the intervals in due course, either to fight alongside
the first line or to take over the burden of the combat."

Best,
Mark

[Justin]

Thanks for your input Mark. I had not read the passage from Livy. I
notice one or two things about it:

a) the Hastati retreated 'slowly' through the intervals between the
companies of Principes. This would fit with a gradual narrowing of the
gaps by the Principes going hand in hand with a gradual withdrawal
through them by the Hastati.

b) "The principes then took up the fighting and the hastati followed
them." What exactly does that mean? If the Hastati are pulling back
then how do they follow the Principes? It seems to imply that the
Hastati added their weight to the rear of the Principes' line. Is that
plausible?

c) "The triarii, rising up after they had received the principes and

hastati into the intervals between their companies, would at once draw

their companies together and close the lanes, as it were". This could
be interpreted in the sense of the mechanic I describe above, but not
necessarily so.

I had read the passage from Prof Sabin but was not entirely convinced
by it. I feel that a small portion of a battleline, effectively 10 men
wide, that bulges in through a gap between Principes centuries, is a
very different proposition to an entire 10 000 man phalanx confronted
with a refused Spanish centre and advancing Roman wings. The 10 man
wide bulge could quite easily catch the Principes blocks on their
corners and decimate them whilst simultaneously fighting off an attack
from the second line behind them, particularly if the second line
consists of retired Hastati who have already had enough of the fight.
Of course it doesn't help not being there to see how it worked in
practice.

Cheers,

Justin

[Mark]

Livy never saw how it worked in practice, either, so it's not
surprising that his description is a bit vague and incoherent. He was
describing a long-gone tactical system, as illustrated by his
digression explaining "it has come to the triarii" ("ad triarios
redisse") - something like a present-day writer explaining "keep your
powder dry."

I don't know what evolution is contemplated by "the principes then
took up the fighting and the hastati followed them," but that's
exactly what he says: "tum principium pugna erat; hastati
sequebantur." It does seem to indicate that the hastati would rally
and reform behind the principes.

Mark

> > > Justin- Hide quoted text -
>
> - Show quoted text -

[Patrician]

It is an intriguing question.

Livy's description in the original Latin is:
"ubi his ordinibus exercitus instructus esset, hastati omnium primi
pugnam inibant. si hastati profligare hostem non possent, pede presso
eos retro cedentes in intervalla ordinum principes recipiebant. tum
principum pugna erat; hastati sequebantur. "

My best effort at this is:

ubi his ordinibus exercitus instructus esset
When the army had been drawn up in these ordines (i.e. the subunits
described in the previous section of text)

hastati omnium primi pugnam inibant
of all [the troops] the hastati began the fighting

si hastati profligare hostem non possent
if the hastati were unable to overcome the enemy

pede presso eos retro cedentes
pressed by [the enemy] foot they, retiring rearwards,

in intervalla ordinum principes recipiebant
were received through the intervals in the ordines of the principes.

I have deliberately not translated 'ordinum' and 'ordinibus'. An
'ordo' could signify a line, rank or file (usual meanings), or it
could mean the 60-man subunit that Livy has hitherto referred to by
this designation. Exactly which is meant here is of some significance
to understanding what may have happened. The generally accepted
understanding is that 'ordinum' refers to the 'ordines' or 60-man
formations, with spent hastati units retreating through 10 man-wide
gaps. It is possible that Livy, with his often loose or anachronistic
terminology, actually meant that there were no gaps between sybunits
but the individual files of hastati retired between the individual
files of principes, a much easier evolution to achieve in the face of
an enemy.

I raise the possibility because, having recently seen how Dionysius of
Halicarnassus appears to have misunderstood a term in his source and
as a result produced a somewhat unlikely description of the armament
of the principes, the possibility that Livy may have misunderstood his
source may be one we would be unwise to exclude.

If this suggestion has any merit, the implication would be that the
manipular formation was primarily for ease of manoeuvre and that on
the battlefield the hastati, principes and triarii would all form
lines without gaps between subunits but with files at the usual
interval of 3 feet per man. Retirement would pass the individual
files of the retiring line though between the files of the line
behind, making disengagement and handover easy.

Concerning "tum principium pugna erat; hastati sequebantur," this
suggests the retiring files of hastati (or what was left of them)
would form up behind the files of principes. This may have been a
measure to avoid hastati going AWOL or it may have been 'accensi
syndrome', i.e. putting rows of helmets behind those of your troops
who are doing the actual fighting. Enemies tend to look at the
numbers they face, and hastati still able to stand would be able to
contribute to the war shout and have the sun gleam off their helmets
so that the enemy still glimpsed an apparently endless succession of
Romans through the dust of the battlefield. This would be in keeping
with the psychological aspect of triarii commitment being as important
as their actual impact (Livy VIII.8.13), and fits intriguingly well
with the use of the rorarii as battlefield replacements for an
apparently combined formation of principes and hastati in Livy's one
description of how this type of legion fought (VIII.9.14). The
rorarii would be unable to 'add their strength' to the hastati and
principes (addebantque vires hastatis ac principibus) unless the
latter had been drawn up behind the former after withdrawing through
them.

Interesting discussion. Then again, I suppose that is what we are
here for!

Patrick

[ga...@berkeley.edu]

> Livy never saw how it worked in practice, either, so it's not
> surprising that his description is a bit vague and incoherent. He was
> describing a long-gone tactical system, as illustrated by his
> digression explaining "it has come to the triarii" ("ad triarios
> redisse") - something like a present-day writer explaining "keep your
> powder dry."
>

Hi all,
I hope my 2 cents here are useful, not annoying.
My understanding of the idiom ("ad triarios redisse") in common use in
Latin is "it was a close call," meaning in military terms that the enemy
fought on until the triarii defeated them. The term translates into
modern use the way we still say "don't go off half cocked."
Livy explains stuff like this, perhaps obvious to members of his own
generation, because he predicted that his work would be read for
several generations down the line when some idioms were lost in
meaning. I could make a short list of these idioms he explains, like
"why we say Thalassio at weddings."
gaius

[Paul Bardunias]

My understanding is that the meaning was more like the American idion
"the shit has hit the fan", which begs what fan was ever dunged. Not
so much a comment on the duration of battle, but the desperation of
having to rely on the last line of defense because something has gone
terribly wrong.

[Justin]

> If this suggestion has any merit, the implication would be that the
> manipular formation was primarily for ease of manoeuvre and that on
> the battlefield the hastati, principes and triarii would all form
> lines without gaps between subunits but with files at the usual
> interval of 3 feet per man. Retirement would pass the individual
> files of the retiring line though between the files of the line
> behind, making disengagement and handover easy.

Yes, very feasible and easier than the file-by-file plugging system.
One question though: the doubled centuries of the Principes - is that
an academic deduction or is it described as such in the sources?
Doubled centuries would imply either that there were twice as many
Principes as Hastati, or that there were century-wide gaps in the
Principes line.

Are there other documented cases of heavy infantry retiring directly
through lines of other heavy infantry? Does it historically pose any
problems?

On the subject, why would the Legion initially deploy in separate
blocks of maniples? (also please clarify the distinction in function
between maniples and centuries) I had the impression that armies
usually had plenty of time to deploy and would normally only fight on
an open piece of ground, hence pre-battle deployment and manoeuvring
would not have presented any real difficulties for a solid line - the
other side, without benefit of maniples or centuries, seems to have
managed it well enough.

On the "tum principium pugna erat; hastati sequebantur" one could
possibly interpret this line as: "the principes (files) moved up to
the fight, following which the Hastati moved off", fitting in with the
mechanism I hypothesise. But that might be straining the meaning a
bit, unless Livy is misunderstanding what he is describing and has mis-
paraphrased it.

Thanks for all the input.

Justin







On Jul 28, 12:06 am, Patrician <patrick.water...@mypostoffice.co.uk>
wrote:

[bard...@aol.com]

If this suggestion has any merit, the implication would be that the
> manipular formation was primarily for ease of manoeuvre and that on
> the battlefield the hastati, principes and triarii would all form
> lines without gaps between subunits but with files at the usual
> interval of 3 feet per man.  Retirement would pass the individual
> files of the retiring line though between the files of the line
> behind, making disengagement and handover easy.

Do you think 3' is enough room to pass two men abreast.  I have faced this problem with hoplites engaged in a countermarch.

There is certainly enough room, but I cannot say this can be done in the heat of battle.  I tend to think it could though.

 

I don't want to stick my neck out too far where Romans are concerned, but Hellenistic manuals detail a type of doubling,

where for every other file, the man steps behind the man on his right, that would clear a 6' lane, minus one body width, for the Hastatii to retire through.

 

Moving in and out of this type of doubling would be pretty quick and does not involve gross movement of the adjacent files or substantial gaps. 

[Justin]

> I don't want to stick my neck out too far where Romans are concerned, but Hellenistic manuals detail a type of doubling,
> where for every other file, the man steps behind the man on his right, that would clear a 6' lane, minus one body width, for the Hastatii to retire through.

I like this. It reinforces the use of the term "intervalla" to
describe the space between the "ordines" (I had been wondering if one
would speak of any real space or gap between men standing at three
feet intervals, even if other men could squeeze between them). I think
it a fair presumption that the Romans would have picked a lot of
military habits from the Greeks and that this kind of manoeuvre would
naturally have been part of their repertoire.

Oh, and forget my query about the doubled centuries of Principes - I
realise that of course they would have deployed into a single-century
deep line just before contact with the enemy. But I still find it a
bit strange why they would deploy doubled in the first place.

Cheers,

Justin

[Justin]

Unless of course I'm chasing a mirage. If the principes doubled their
'ordines' = files, not centuries, with one file inserting itself into
another, then the whole principes formation would still be a solid
line. Hence the whole Roman 'checkerboard' formation is a modern myth.

Input, anyone?

On Jul 28, 8:38 am, bardun...@aol.com wrote:

[Mark]

Thanks - this is good. I wasn't trying to render the literal sense;
just trying to give an example of a modern idiom, the origin of which
may be or become unintelligible because of changes in technology and
tactical systems.

Your expert insights are always welcome, to me at any rate!

Mark

[Patrician]

Paul, that seems an eminently sensible suggestion, and one that might
explain another, albeit later, minor Roman conundrum.

Vegetius (Book III) mentions, when dealing with how to draw up a lone
of battle, "we now come to the distances and dimensions. One thousand
paces contain a single rank of one thousand six hundred and fifty-six
foot soldiers, each man being allowed three feet."

This neatly fixes frontage; however, when we come to depth:

"We have previously observed the distance between each rank should be
six feet, one foot of which is taken up by the men."

Polybius (XVIII.30), conversely, gives each Roman soldier three feet
for both frontage and depth (treis podas kat' epistatēn kai kata
parastatēn).

There are two reasons why Vegetius might want a six-foot depth: the
first is that when hurling javelins, especially long, heavy ones, the
right arm goes back and so does the end of the javelin. This could be
bad news for someone standing about three feet behind: with a six-foot
interval there is no danger.

The second reason could be that, as per Hellenistic procedure (the
Romans were great borrowers), alternate files could slip into the
spacing available to allow the files of a retiring line to filter back
through without a tight squeeze, as you propose.

Polybius' three-foot depth would indicate that after letting the
retiring hastati pass through their lines, the principes would
reoccupy their old files, hurl pila if given the opportunity (possibly
the whole unit volleying simultaneously on command, the rear ranks
shooting indirectly) and then close up to a three-foot fighting depth
for melee.

The Romans might be able to do this with the principes in their
original files at an interval of three feet per man, though this
would involve the Principes standing still and holding their shields
with the side towards the enemy to increase the space available to the
retreating hastati. If each man of the principes could thus reduce
his frontage occupancy to one foot, that would leave two feet for the
hastati to back through (I assume they would try to file through
backwards, which may be what Livy means by 'retro cedentes'). It
would certainly be less easy than having 4-5 feet available, so the
alternate files approach has the advantage of greater feasibility.

Justin raises a few important points, starting with why there were two
centuries in each maniple. For this, we need a slightly more detailed
look at the Roman military system.

The essence of the matter is that in the legion described by Polybius
(c. 264-107 BC) each line had ten maniples and each maniple consisted
of two 60-man centuries for the hastati and principes and two 30-man
centuries for the triarii. How it got there from the earlier legion
may have gone like this.

Livy's legion (book 8, chapter 8), dated to the Latin War in 340 BC,
has a first line of 15 ordines (60-man formations) of hastati, each of
which has 20 skirmishing 'leves' attached, an early form of velite.
The second line, the principes, also has 15 ordines of 60 men each.
So does the third line, the triarii. There are two more lines, each
of which also has fifteen 60-man ordines: the fourth, the rorarii,
appear to be (in the one description we have of this legion in action,
Livy VIII.9 and 10) battlefield replacements, 'adding their strength
to' the ranks of hastati and principes after the battle has gone on
for some time. The last line, the accensi, consists of persons of
lesser ability armed and equipped as triarii, and their role would
logically be to add weight to the triarii in the final fight at the
climax of the battle.

The important point here is that each ordo consists of 60 men
(actually 62: 60 men, one of whom is a standard-bearer, and 2
officers). We may note that this particular legion, in which the
terms ordo and maniple both seem to refer to the 60-man subunit
(although the 20-man light infantry detachment is also a 'manipulus';
this looseness of application recalls that of the later 'numerus'),
does not seem to have paired up centuries (unless each ordo/maniple
was actually split into two weak 30-man centuries as some, including
Peter Connolly, have conjectured). The inference from this would be
that each line formed as a single entity without gaps, and hence the
method of relief would logically be by passing files between files.

When this legion was reformed to give us the legion described by
Polybius in Book VI of his Histories, the following changes seem to
have occurred. We begin with

15 x 60-man ordines of hastati (plus 15 20-man subunits of leves,
light troops)
15 x 60-man ordines of principes
15 x 60-man ordines of triarii
15 x 60-man ordines of rorarii
15 x 60-man ordines of accensi

Total: 4,800 troops. Also present were 150 officers (total 4.950)

From these:

15 x 60-man ordines of hastati plus 5 ordines of rorarii => 20
centuries of (new) hastati in 10 x 120-man maniples (1,200 hastati)
15 x 60-man ordines of principes plus 5 ordines of triarii => 20
centuries of (new) principes in 10 x 120-man maniples (1,200
principes)
10 x 60-man ordines of triarii => 10 x 60-man maniples of (new)
triarii each of 2 centuries of 30 men each (600 triarii)

10 ordines of rorarii are discarded (the legion size decreased from c.
5,000 to c.4,200)

(The Polybian legion had provision to raise strength to 5,000. In
such an event, all categories except the triarii received a
proportionate increase. Hence the troops that would have previously
formed 10 centuries of rorarii would be portioned out between the
hastati, principes and velites. By this time rorarii were no longer
being recruited as a distinct category, it presumably having been
decided that it made more sense to start them off where they would
anyway have to end up, i.e. among the hastati and principes.)

15 ordines of accensi plus 15 x 20-man groups of leves => 1,200 (new)
velites in 60 x 20-man groups. This is quite a career change, though
not a status change, for the accensi.

The new legion totals 4,200 men, officers apparently now being elected
from within the legion (Polybius VI.24).

The new (Polybian) legion thus has 10 maniples in each line instead of
15, making command and control easier, and each maniple has two
centuries. Eventually, legions would be reorganised (Marius' reforms
in 107 BC) so that one maniple each of principes, hastati and triarii
would be combined into a cohort, leaving the legion with ten manoeuvre
elements instead of thirty, and hence easier to control while actually
becoming much more flexible. The new legion is also 'combat front-
loaded', with most of its fighting power up in the first two lines
rather than being balanced before and behind the standards.

The reorganisation indicates why each maniple ended up with two
centuries, but not why they are referred to as 'prior' and
'posterior'. That particular set of designations indicates they would
have travelled in pairs, in tandem. Here we come to the point of a
manipular legion: instead of manoeuvring in one long line, the legion
manoeuvres in a number of small, easy-to-handle groups that can form a
line at very short notice.

I emphasise manoeuvring: if crossing any sort of undulating or uneven
terrain, any solid line has to do one of two things: either advance at
the speed of the most incommoded subunit or become disrupted. A force
in manipular configuration is moving in relatively small formations
which can alter their speed or their route to cope with obstructions
and inconveniences, and the force can assemble itself into an
effective battleline at short notice. It is a sort of early
illustration of Napoleon's concept of 'march divided, fight united'.

In Livy's Latin War legion, there were 45 subunits that needed to be
controlled when the legion was in motion. The legion of Polybius'
time had only 30 to control. Marius' legion (and its subsequent
variant used by Julius Caesar) had a mere ten manoeuvre elements to
coordinate.

To return to terminology, a maniple consists of two centuries (in the
Polybian legion, which we should really call the Punic Wars legion) or
one ordo (effectively a century equivalent) in Livy's Latin War
legion. Hence, a Polybian (Punic Wars) maniple has 120 men in two 60-
man centuries, and a Livian (Latin War) maniple (ordo) has (is, to all
intents and purposes) one 60-man century without trimmings. (Polybius
is of limited help with Roman terminology, as he gives Greek
equivalents: 'kai tagma kai speiran kai sēmaian'. Others will be
better than me at explaining what a 'tagma', 'speira' and 'semaia'
signify.

The two centuries in a Polybian (Punic Wars) maniple bear the
designations 'prior' and 'posterior'. This indicates the relative
positions they would have occupied while the maniple was on the move.
For action, they would line up one beside the other.

Assembling a solid line for battle requires units to form up from
their column of march to line of battle formation. In the case of a
hoplite or Macedonian phalanx, once you are assembled you are either
held up by your slowest-moving troops or at risk of outstripping them
and leaving them behind. The manipular formation used by the Romans
(after they borrowed it from the Samnites and similar hill and
mountain-dwelling opponents) allows a line of battle to travel as a
series of small columns, assembling at the very last moment before
fighting begins (the skirmisher screen delays the enemy a bit while
this is going on, and, just as importantly, stops enemy skirmishers
shooting your officers while you are deploying and assembling your
line). Most of Italy is pretty rugged (as the Allies found in
1943-45) and the ability to fight in less-than-amenable terrain was
important once Rome began expanding into the Apennines. Armies still
where possible fought in clear, level terrain (it made life, command
control and pursuit so much easier, besides being closer to
substantial food supplies), but the Romans simply had to develop the
ability to manoeuvre in rougher terrain in they wanted to hit the
Samnites where they lived.

Every army manoeuvred by subunits to an extent (the Macedonians at
Cynoscephalae seem to have ascended steep hills by files and then
assembled the files into units at the top of the hill) but the Roman
manipular organisation meant that they could travel in a condition
closer to battle readiness and react more snappily.

Back on "tum principium pugna erat; hastati sequebantur" it would make
at least theoretical sense for the principes to push through the
hastati and take over the battle, letting the hastati drop back.
Livy's terminology, (retro cedentes ... recipiebant) suggests,
however, that the hastati did all the moving. He follows this with:
"tum principum pugna erat", i.e. 'then was the battle of the
principes', implying they did not get stuck in until the hastati had
finished retiring. For this reason I am reluctant to assume that the
principes moved at all during proceedings, logical though such
movement might appear. Livy could still be less than completely
correct, of course.

It does increasingly look (at least to me) as if Justin's suggestion
that individual files rather than whole maniples were the basic unit
of retirement during Roman line relief has a great deal to recommend
it. I think a pat on the back may be due.

Patrick
.

[Mark]

This is indeed interesting, and though I also don't want to stick my
neck out with regard to Romans, I have a couple of observations:

1. Any explanation must take account of LIvy's express statement at
8.8.5 that the maniples were "a short distance" ("modicum spatium")
apart, and also take into account the other apparent references to
gaps between maniples in Classical sources. I don't have anything like
a complete list of those, but there are for example Liv. 8.9.12;
10.27.8; 10.41.9; 30.33.3 and Plutarch, Pyrrhos 21.10. Not that these
can't be explained - indeed many seem to be references to spaces being
created or opened up, which implies that there were no spaces or only
small spaces to begin with.

2. I don't want to harp unduly on an old subject, but it's at least
possible that Dionysius was not confused about the armament of the
principes, but rather was reproducing a description that he found in a
Greek source more or less contemporary with Pyrrhus - and that the
description would have made sense to the intended audience of that
Greek source. Incidentally, Lévèque (Pyrrhos, 1957) and J. Hornblower
(Hieronymus of Cardia, 1981) favour the identification of Hieronymus
as Dionysius' source because, inter alia, Dionysius appears to have
read Hieronymus; Plutarch cites Dionysius and Hieronymus together in
his account of Pyrrhus; and the fragments of Hieronymus seem to
indicate that he dealt with Pyrrhus' campaigns in some detail.

Mark

On Jul 27, 4:06 pm, Patrician <patrick.water...@mypostoffice.co.uk>
wrote:

[Mark]

In addition to the examples below, Marquardt cites "intervallia inter
ordines" (Liv. 10.5.6) and "ta diastêmata tôn sêmaiôn" (Plb. 15.9.7)
(Römische Staatsverwaltung, 2nd ed., Leipzig: 1884, Vol. 2 p. 351).


Mark

> > Patrick- Hide quoted text -

[Patrician]

An interesting set of references, three of which are notably useful.
Before turning to them, Mark's point about Livy VIII.8.5 deserves
consideration.

The relevant portion of text is "manipuli quindecim distantes inter se
modicum spatium manipulus levis vicenos milites".

The question here is whether the 'inter se modicum spatium' refers to
an interval between the 'manipuli quindecim' of hastati or between the
'manipulus levis' and its respective hastati maniple. I favour the
latter, on the basis that otherwise the leves are left hanging with no
orientation. If this understanding is correct, the text would
indicate fifteen maniples, with a moderate space between each one and
a maniple of twenty light infantry. The light infantry should,
tactically, be stationed ahead of the main infantry line: putting it
in gaps between hastati is pointless unless one is a Scipio with a lot
of elephants to deal with, which brings us to the next reference.

Polybius XV.9 is a completely clear indication that a 'quincunx'
formation was in use during the second Punic war, and by extension
probably for the entire period in which the legion described by
Polybius in Book VI was standard. So as far as this legion is
concerned we are back to the original question of how line relief by
maniples was carried out.

Livy X.5.6, contrariwise, supports the observation that intervals
between maniples in Livy's Book VIII legion were not customary. To
give the relevant passage:

"prima incedebant signa legionum, ne quid occultum aut repentinum
hostis timeret; sed reliquerat intervalla inter ordines peditum, qua
satis laxo spatio equi permitti possent."

The consul is leaving enough spaces between the 'ordines' to allow the
passage of cavalry, but advancing the 'prima signa legionum' so that
the enemy would not see or suspect anything, the particular 'anything'
being that he had left gaps between the 'ordines'. This would seem to
indicate that such gaps would be noticed as being unusual.

I must confess myself baffled by reference to Livy VIII.9.12, because
it describers the "cohortes Latinorum fugam ac uastitatem late
fecerunt" in the context of Decius Mus' devotio at Vesuvius in 340
BC. The Latin 'cohorts' (sic) fled leaving an empty space around him
(literally 'widely made flight and empty space).

Livy X.27.8 is another odd one in this context: "Whilst the two armies
were standing ready to engage, a hind driven by a wolf from the
mountains ran down into the open space between the two opposing lines
(inter duas acies decurrit)."

Livy X.41.9 is more useful: "His orders were carried out to the
letter; the legionaries opened their files (panduntur inter ordines
viae), the cavalry galloped through the open spaces, and with levelled
spears charged the enemy's centre." This at least indicates that
opening gaps or intervals was something the 340 BC legion needed to
make a special effort to achieve: such gaps were neither integral nor
natural to it. it is basically a confirmation of Livy X.5.6.

Livy XXX.33.3 parallels Polybius XV.9 as an indication that the
quincunx deployment was natural for the Punic Wars (Polybian) legion,
though Polybius' text is more useful and explicit in this regard.

Plutarch, Pyrrhus 21.10 is truly baffling in this context: "For he had
lost a great part of the forces with which he came, and all his
friends and generals except a few; moreover, he had no others whom he
could summon from home, and he saw that his allies in Italy were
becoming indifferent, while the army of the Romans, as if from a
fountain gushing forth indoors, was easily and speedily filled up
again, and they did not lose courage in defeat, nay, their wrath gave
them all the more vigour and determination for the war." Is this
actually the intended reference? Pyrrhus 21.6 has a reference to "he
[Pyrrhus] put great numbers of slingers and archers in the spaces
between the elephants."

What emerges from this is that Polybius' Book VI legion is explicitly
noted as using a quincunx deployment, so we are back to the drawing
board with that one, or, more accurately, back to looking at Justin's
original explanation.

Livy's Book VIII legion, however, as Mark notes, is depicted as having
to make gaps when passage through its ordines is required. This would
indicate that such gaps did not normally exist.

It looks as if we have two models of legion with two significantly
differing modi operandi. The Latin wars legion of Livy VIII.8 appears
to use relief by files and perhaps, like the Greek concept of epi
skelos anagein, retreated with their faces towards the enemy. The
Punic Wars legion of Polybius VI (and XV) lands us back with the
question of how they could have executed relief by maniples.

On a different subject, Dionysius actually cites Timaeus as having
written a history of Rome's wars with Pyrrhus, but whether Dionysius'
source was Hieronymus or Timaeus or both, if he had any grounding in
military matters he would not have been satisfied with the description
he gave whether it was in his source or his mind. It is like the
infamous English translation of Polybius which issues Hannibal's
skirmishers (lonchophoroi) with pikes: the weapon is incompatible with
the role.

While I respect any individual's right to understand a passage in a
way that makes sense to them, I would need to be convinced of the
practicality of using cavalry spears two-handed while at the same time
operating a scutum to block blows and cover the body (Polybius XVIII.
30.7).

Patrick

[Paul Bardunias]

OK, like the wiley tortoise I shall really stick my neck out now.
Just how far from the Hellenistic tradition was the roman system of
stacked units? Consider what Polybius suggests for the manner in
which Alexander moved his men over a long stretch of broken ground:

Plb. 12.20 Inaccuracies about Tactics
Again, he says that Alexander was marching in line when he was about
forty stades from the enemy. A greater blunder it is difficult to
conceive. For where could one find a ground, and especially in
Cilicia, twenty stades broad by forty deep, for a phalanx armed with
sarissae to march in line? It would not be easy to count all the
impossibilities in the way of such an arrangement and such a movement.
One that is mentioned by Callisthenes himself is sufficient to
establish the point. For he remarks that the winter torrents which
descend from the hills make so many gullies in the plain, that, in the
course of the flight, the chief part of the Persians are said to have
lost their lives in deep places of that kind. But, it may be urged,
Alexander wished to be ready for battle as soon as the enemy were in
sight. But what could be less ready than a phalanx in a disordered and
straggling line? Is it not much easier to form up a phalanx from a
proper column of route, than to bring a disordered and straggling line
back into the same alignment, and get it into order of battle on a
broken and woody ground? It was, therefore much better to march twice
or four times the ordinary depth of a phalanx1 in good order, for
which sufficient ground could possibly be found. And it was easy to
deploy his men quickly into the line of the phalanx, because he was
able by means of scouts to ascertain the presence of the enemy in
plenty of time. But in this case, beside other absurdities, while
bringing his men in line across the level, he did not even (we are
told) put the cavalry in front, but marched with them in the same
alignment.

1 That is, sixteen or thirty-two deep.

So Polybius was either projecting back the notion of advancing into
battle in stacked units with gaps between them from his knowledge of
Roman tactics, or, more likely I think, phalanxes advanced into battle
at times in a formation of doubled Speira. This makes me wonder about
the formations we read of at least since Pyrrhus where the phalanx has
units of other troops intercalated. I'd have to go through the
accounts to see if we could determine at what scale the alternation
occured, but lighter troops between stacked speira that could then
deploy seems very roman to me.


On Jul 29, 4:57 pm, Patrician <patrick.water...@mypostoffice.co.uk>
wrote:

[Mark]

Yes, this is getting to be more and more interesting. Can it be that
the notion of the chequerboard manipular legion really rests on such a
slender foundation? The sources cited by Marquardt do not seem
compelling, as Patrick has explained. And can it be that Roman
evolutions are not radically different from those described by the
Hellenistic tacticians?

The passage from Plutarch's Pyrrhos I was trying to refer to, by the
way, is 21.9-10 in the Teubner edition (and the TLG database) - I now
see that it's 21.6 in the translation on the Perseus website. Sorry
for the confusion. I'm posting a PDF file that includes the text. In
any event, it's not a compelling piece of evidence either. It's part
of the description of Asculum; in Dryden's translation it runs:

"The Romans, not having those advantages of retreating and falling on
as they pleased, which they had before, were obliged to fight man to
man upon plain ground, and, being anxious to drive back the infantry
before the elephants could get up, they fought fiercely with their
swords among the Macedonian spears, not sparing themselves, thinking
only to wound and kill, without regard to what they suffered."

But I need to go away and think a bit more about whether it's
reasonable to think that Livy truly was describing a different stage
in the evolution of the legion than Polybius!

Regards,

Mark

> ...
>
> read more »- Hide quoted text -

[Mark]

One fundamental question is whether Livy’s description really does
represent a different and earlier stage in development, or whether it
is just a confused and garbled account of the same organisation that
Polybius describes. Yes, Livy purports to be describing the legion of
340 BC, but also claims that the manipular organisation originated
when pay was introduced (about 406 BC as discussed here earlier),
which clearly is far too early. And he is inconsistent: elsewhere he
seems to assume that the Romans always used maniples, even from the
early days of the kingdom, e.g. 1.52.6; 2.20.7; 2.53.1; 4.37.11. So
it seems unwise to assume that his chronology is exact. So here are
few more observations on Livy’s description of the legion in 8.8.3-14,
based principally on Oakley’s commentary on Book 8.

Generations of Latinists have struggled with the text of 8.8. One of
the cruces is the final clause of 8.8.3: “postremo in plures ordines
instruebantur” (“finally they were drawn up in more ordines”). One
question has been whether “postremo” is the right reading. If it is,
it is part of a sequence “antea … dein … postremo” that implies that
the manipular organization was an intermediate stage, coming between
the phalanx and the division into “plures ordines”. But instead of
going on to explain about the ordines, Livy continues in 8.8.5-7 by
speaking of the manipulus. So is the mention of the ordo, in 8.8.4 to
be taken simply as reference to some future stage of development that
Livy doesn’t intend to describe? Editors have suggested amendments:
postrema (neuter plural), though this only seems to make the passage
more obscure, or postremi, which would make it say something like
“their final line was drawn up in several ordines” (Ortmann,
Scriptorum latinorum (1882), p. 12).

Another question is whether the definition of the ordo (“sexagenos
milites, duos centuriones, vexillarium unum habebat”) belongs in 8.8.4
or in 8.8.8. The manuscripts have these words in 8.8.4, after
“instruebantur”, i.e. they have Livy saying (with Ortmann’s
emendation) that the final line was divided into several ordines, and
an ordo is 60 men, 2 centurions and a standard-bearer. But this is at
odds with what Livy says in 8.8.7-8, where he tells us that an ordo is
186 men. And of course it is at odds with Polybius. Conceivably Livy
could just be talking about the organisation of the final line, but if
so he gives no indication that the rest of the army would have been
different. Or he could be using ordo to mean “century”, but then he
is giving his centuries 2 centurions – and ordo in 8.8.7-8 clearly
does not mean “century”.

Attempts to make sense of this have included deleting the first
“vexillum” in 8.8.8, so as to make Livy say that an ordo consists of
186 men. But in 8.8.4 Livy says explicitly that an ordo is 63 men.
This led Weissenborn to propose deleting the whole of 8.8.4 as an
explanatory gloss that had mistakenly been inserted into the text
(NJbPh 39 (1843): 243-283 at pp. 269-270). Mommsen argued for
deleting everything from “earum” in 8.8.7 to “errant” in 8.8.8 (Die
römischen Tribus (1844), pp. 128-130), but did not explain why the
passage proposed for deletion would ever have been inserted into the
text in the first place. Thompson and Stroth each proposed other
deletions.

But in 1824, Reisig had suggested that the phrase “sexagenos milites …
habebat” really belongs between “constabat” and “vexillum” in 8.8.8
(Jenaer Literaturzeitung, Suppl. R, pp. 38-41). The same proposal was
made by Conway in 1918, apparently independently (Walters & Conway,
“Restorations and Emendations in Livy vi-x”, CQ 12 (1918): 1-14 at pp.
9-14), and this has formed the basis of the present editions. This
amendment involves much less violence to the text than the extensive
deletions proposed by others, involving only an assumption that the
words had at some point been inadvertently skipped over by a copyist
and then inserted in the wrong place. And the amendment removes the
difficulties created by including the words in 8.8.4; it has Livy
saying that an ordo is divided into 3 parts (8.8.7), explaining that
these parts are vexilla, and then giving a total number for the 3
combined vexilla (i.e. an ordo). Two centurions for a vexillum of 60
squares with Polybius’ description of 30-man centuries for the
triarii.

But even with this amendment, Livy’s account leads to confusion about
the size of the legion: multiplying his figure for the ordo (189)
gives 8,505 – far more than the 5,000 he states at 8.8.14. At 6.22.8
Livy says the legion was 4,000, and at 7.25.8 he gives 4,200.
Certainly 8,500 is far too large for any legion we know of.

Of course, Livy may not have expected that all the ordines were the
same size. There have been many attempts to calculate how Livy’s
structure (15 maniples of hastati and principes and 15 ordines in the
third line) could be made to add up to 5,000, e.g.:

Niemeyer argued that the complement was 5,100 (“Zu Livius”, NJPhP 23
(1877): 179-183). He posited 1,200 hastati (15 x 60 scutati + 20
leves) + 1,200 principes (15 x 80 scutati) + 2700 others (15 x 180,
with each 180 comprising 60 triarii + 60 rorarii + 60 accensi). He
accounted for Polybius’ figure of 4,200 by assuming that at some point
the 900 accensi were deleted and the other maniples reorganised:
instead of 15 maniples of 60 heavy and 20 light troops in the hastati,
there would be 10 maniples of 120 + 40; instead of 15 maniples of 80
in the principes, there would be 10 maniples of 120 + 40; and instead
of 15 ordines of 60 + 60 + 60, there would be 10 maniples of 60 + 40.
But there is nothing in the text to support a figure of 80 for the
Livian hastati and principes.

Marquardt proposed different numbers (Römische Staatsverwaltung
(1881-1885), vol. 2, pp. 360-363). He had 945 (63 x 15) hastati, 945
(63 x 15) principes, 2,790 (186 x 15) triarii, rorarii and accensi,
and 45 vexillarii – a total of 4,725. But this seems to skate over all
the difficulties with the text of 8.8.4 and 8.8.8.

More recently, in the introduction to their Budé edition of Book 8
(1987), Bloch and Guittard (pp. cxi-cxii) suggest that the 5,000
comprises exactly 945 (63 x 15) hastati, 300 (20 x 15) leves, 945 (63
x 15) principes, 930 (62 x 15) triarii, 930 (62 x 15) rorarii, 930 (62
x 15) accensi, and 20 scutati. This is an elegant equation, but there
is nothing in the text to support 300 leves in addition to 945
hastati, or to support 20 separate scutati. And is there any reason
to accept the existence of a variation in the size of the ordo and
vexillum between the hastati/principes and the triarii/rorarii/
accensi?


There is, at any rate, plenty of cause for perplexity here.


Best,

Mark


On Jul 28, 10:30 am, Patrician <patrick.water...@mypostoffice.co.uk>
wrote:

> ...
>
> read more »

[Patrician]

Before we resume, a couple of quotes dredged from Greek historians.
The first is from Dionysius of Halicarnassus V.15.4.

"Titus and Sextus, the sons of King Tarquinius, put the left wing of
the Romans to flight, and advancing close to their camp, did not fail
to attempt to take it by storm; but after receiving many wounds, since
those inside stood their ground, they desisted. These guards were the
triarii, as they are called; they are veteran troops, experienced in
many wars, and are always the last employed in the most critical
fighting, when every other hope is lost."

This battle is datable to 509 BC (Tarquin's attempt to reconquer Rome)
and looks like being the first mention of triarii in any source - in
the role of camp guards. No wonder 'ad triarios redisse' originally
meant that things were going really badly: for students of military
scatological imagery, one can compare Ducrot's observation at the
Battle of Sedan (1870): "Nous sommes dans un pot de chambre et nous
serons emmerdes." I trust translation is not required ...

The second quote is from Diodorus Siculus, XXIII.2.1.

"The Romans, for their part, advised the Carthaginians not to teach
them to meddle with maritime affairs, since the Romans, so they
asserted, were pupils who always outstripped their masters. For
example, in ancient times, when they were using rectangular shields,
the Etruscans, who fought with round shields of bronze and in phalanx
formation, impelled them to adopt similar arms and were in consequence
defeated. Then again, when other peoples were using shields such as
the Romans now use, and were fighting by maniples, they had imitated
both and had overcome those who introduced the excellent models. From
the Greeks they had learned siegecraft and the use of engines of war
for demolishing walls, and had then forced the cities of their
teachers to do their bidding. So now, should the Carthaginians compel
them to learn naval warfare, they would soon see that the pupils had
become superior to their teachers."

The reference to 'rectangular shields' evidently predates the Tullian
'legion' and should strike a chord with anyone who has seen reliefs of
ancient Near Eastern armies. The shields in question probably went
with an earlier Phrygian-style military system (not the late version
with indented shields), which we can discuss in more detail if anyone
is interested.

One notes that when fighting Etruscans, the Romans adopted an Etruscan
system. The trigger for conversion to a manipular system was
experiencing such a system used against themselves, which suggests
either the Volscians or the Samnites as the triggering power. By the
time of Livy's legion (340 BC, in the middle of the Latin War), the
Romans and Latins used an identical military system (Livy 8.8.15). We
now turn to that system and compare it with the one described in
Polybius Book VI.

The salient features of the 'Livian' and 'Polybian' legions are as
follows:

Infantry strength: Livian 5,000 Polybian 4,200
Number of subunits: Livian 75, + 15 skirmisher groups (turbae),
Polybian 30, + 30 skirmisher groups
Number of troop types: Livian 6, Polybian 4
Number of lines: Livian 5, Polybian 3.

One point that should be made about Livy's descriptions is his habit
of using anachronistic terminology. His forces are liberally
sprinkled with 'cohortes' and 'manipuli' right from the time of
Servius Tullius, who was using a recognisably hoplite-style army. He
uses 'verutum' for the 'vericulum' in use among the lighter infantry.
The one time he attempts to use what may be period terminology, where
he is describing the composition of the Latin War legion in VIII.8,
his use of 'ordo' and 'vexillum' confuses rather than enlightens the
modern reader.

Time to strip away the confusion.

For a start, Livy's figure of 5,000 looks like a rounded rather than
an exact number. If we add up the troops we know about, we can make a
reasonable surmise about those not specified. For this exercise, we
take Livy's text at exactly face value.

Known troops: 15 x 20 leves (300)
15 x 186 triarii, rorarii and acensi combined (2,790)
TOTAL KNOWN = 3,090
UNKNOWN = 1,910

Unknown troops (antepilani maniples, i.e. hastati and principes): 30
formations amounting to approximately 1,900 troops.
1,900/30 = 63.33
Therefore approximate strength of each antepilani maniple = 63

Coincidentally, the ordines of vexilla (or vexilla of ordines, the
designations really do not matter but their sizes and roles do) in the
rear three rows each consist of 62 men exactly (1/3 of 186; 60 troops
and 2 officers). This may also be the exact size of the hastati and
principes formations, giving a legionary total of 4,950 which is only
1% less than Livy's 5,000 and well within the acceptable limits of
rounding. Giving each such maniple a 63rd man brings the legion total
to 4,980, even closer. The question is thus not whether the hastati
and principes maniples in Livy's legion were 60 or 120, but whether
they were 62 or 63.

This simple arithmetical exercise has been perfomed before, and is
quite sufficient to justify the organisation and troop strengths given
by Livy. We may note that Polybius' legion, when raised to a strength
of 5,000 by increasing all contingents except the triarii, also cannot
achieve 5,000 exactly (how do you divide 800 by 3 to increase the
velites, hastati and principes by an equal amount?). It seems safe to
take 5,000 as a rounded rather than exact number and spare oneself any
agonising over a putative and unmentioned 20 supernumerary personnel.
(If one feels compelled to attempt exactitude, count in tribunes and
their attendants.)

Polybius' legion is quite differently organised. It consists of 1,200
velites, 1,200 hastati and 1,200 principes with 600 triarii, each in
ten (not fifteen) maniples, with no rorarii and no accensi. Even Livy
would not mistake 'X' in his sources for 'XV' (and he used Polybius).
Even Livy would not invent two additional troops types out of his head
and then devise a spurious account of how they were used in a battle
(at least, I trust he would not: do I assume correctly that he does
not do so elsewhere?). Polybius' antepilani maniples are explicitly
120 strong, Livy's, as shown above, are 60 (plus officers). We are
dealing with different models of the Roman legion.

I have detailed in a previous post what seems to be the easiest
explanation of how the 'Livian' legion converted to the 'Polybian'
one. If one feels inclined to dispense with the accensi and
amalgamate the rorarii and leves to create velites rather than losing
the rorarii and merging leves and accensi, that is in all likelihood
as valid if not more, depending upon whether Rome had a surplus of the
young or the faint-hearted.

Just by taking Livy's existing text at face value we have a viable and
self-consistent account without problems (or at least none that I can
see, which may not be the same thing). Livy's loose tracing of
previous history and linking the manipular trend (not necessarily the
340 BC form of manipular legion) to the issue of pay suggests a change
around the time of Camillus (early 4th century BC) which would accord
with a change of principal opponents in the late 5th-early 4th
centuries. Continuing changes of opponent between the Latin War and
the First Punic War would provide a credible background for further
modification of legionary organisation and techniques over the
intervening 80 years. Even the 'Polybian' legion seems to have
evolved during its c.270-107 BC slot, adopting (gradually) cohort-
style formations and with specialist foreign missile contingents
interspersed while still retaining the maniple as the basic subunit
(Sallust, Jugurthine War, 46.7 and 49.6).

I cannot see any way in which Livy and Polybius could be describing
the same formation.

Thoughts?


Patrick

[Justin]

Looking again the the idea of the hastati retiring directly between
the files of the principes, which are assumed to be doubled, leaving
about 4' lanes between them, I see a possible problem. If I'm not
mistaken the Greek use of the doubled file was for manoeuvring
purposes, as in the case of the left-flank phalanx that marched up the
hill at Cynocephalae with the intention, once at the top, of deploying
into a line with files 3', or possible 1,5' wide. In other words the
doubled file is not a fighting formation, at least not for the Greeks/
Macedonians.

Presuming the Romans used it when they originally had phalanxes, it
would have occurred to them to adapt it to line relief, but eventually
they would have looked for a better alternative as it has a weakness:
Files of hastati that retire through doubled files of principes must
get clear before the doubled file can move up to fill the 4' gap. This
take a little time, enabling the enemy troops to catch the front ranks
at two men against one, and even penetrate a little into the principes
line, disrupting its formation. It would mean a number of front-rank
principes - the best troops - getting killed.
The problem was not such a major one that it immediately demanded a
reform of the retirement tactic, but it would have prompted the Roman
military minds to look for a better alternative.

Enter the maniple system. For manoeuvring in hilly and wooded
countryside it represents a good way of getting to the enemy without
having either to advance in column or stumble over uneven ground in a
disordered line. Independent maniples are small enough to wheel around
impassible obstacles like clumps of trees whilst keeping their file
structure intact (large lines cannot wheel). They do not need to worry
too much about maintaining exact intervals between each other as the
infantry line on average will keep its frontage constant. When near
the enemy the rear maniples supply the ruler by which the front
maniples adjust themselves and then all are ready to fight.

On the line relief mechanism itself I would like to add a couple of
modifications to my original sugggestion: I assume that the posterior
maniples of principes are directly behind the gaps between the prior
maniples. Throughout the relief process, the front troops are under
constant pressure from the enemy and must maintain an unbroken line of
men spaced at 3' intervals, men moreover, who must be able to devote
most of their attention to fighting the enemy troops in front of them.
Hence the manoeuvre must be done slowly, which seems indicated by
Livy: "If the hastati were unable to defeat the enemy, they retreated

slowly and were received into the intervals between the companies of
the principes."

Furthermore, after a prolonged bout of melee combat, the file
structure of the hastati line will have broken up. The hastati can
still move as mass, but they are not longer able to execute manoeuvres
file by file, or even century by century. The file, like the century,
is a pre-battle manoeuvre formation.

With this in mind:

The hastati line falls back to the prior principes, or perhaps the
principes advance to the hastati. The prior principes at this point
throw their pila to disrupt the enemy and lessen pressure on the
hastati. The volley of pila cannot stop the front ranks of the enemy
in their tracks as these are adjacent to the hastati and so not
targetable.

The rearmost hastati then start falling back through the gaps,
followed by the hastati in front of the prior principes, whilst the
front ranks directly in front of the gap stand firm.

The hastati line gradually collapses into a series of half circles
covering the gaps. The optios - possibly of the principes - control
the rearward flow, ensuring that the sides of the gaps remain covered.
Meantime the righmost (or leftmost) files of the posterior centuries
of principes move up one by one to narrow the gap. Each principes file
throws its pilae at this point. More hastati fall back and the half
circles shrink.

The gaps narrow and the last file of the posterior principes centuries
moves up to close it, leaving the last of the hastati to filter back
directly through the principes ranks. Minimal disruption, a constant
front maintained at all times, and a simple manoeuvre for the hastati.

I've included a pdf in the Files to illustrate this idea.

The triarii, being a residue of the old phalanx, would have used the
old doubled file line relief system when letting the principes
through. The Romans were of course traditionalist sometimes to the
point of irrationality.

Any comments?

On Jul 28, 8:38 am, bardun...@aol.com wrote:

[Paul Bardunias]

On Aug 3, 4:05 pm, Justin <hfro...@gmail.com> wrote:
In other words the doubled file is not a fighting formation, at least
not for the Greeks/ Macedonians.

That's something of the point. The Princeps are not in battle
formation either and are expected to deploy into line. Thus, the
formation with princeps formed in stacked units (don't say 'doubled',
for this has a different meaning and leads to confusion) may be
analogous to the hellenistic practice of coming to battle in stacked
units, then forming line when they engage after letting light troops
go to the rear.

We read that Pyrrhus alternated units of italian allies with his own
troops like "teeth". This is problematic because the benefit of
having the flexibility and presumably missile capabilities of the
Italians is surely outweighed by the coordination problems inherent
with troops of such differing mobility. Some have claimed this as the
reason for his high losses. I think it would be interesting to know
if perhaps he stacked his units like princeps and allowed the Italians
in the intervals with the idea that if they were defeated, his troops
could deploy into a solid line.

Cynocephalae shows an interesting example of (un-)doubling. Polybios
18.24:"while he ordered the peltasts and heavy armed to double their
depth and close up to the right. By the time this was effected the
enemy were close at hand; and, accordingly, the word was given to the
phalanx to lower spears and charge;"

This is usually taken to mean that his order meant that every other
file backed out to stand behind a file beside them, leaving twice the
interval between men (probably from 3'-6'). In order to ensure his
men did not fight at 6' frontages, he had the whole army march to
their right. For the second man from the right this was only 3', but
each man man in line to the left had to walk an additional 3' and you
can see how this took so much time that the Romans were upon them.

Could these units have stacked instead of undoubling, leading to big
gaps between rather that a 6' per man spacing? Engaging in undeployed
enomotia of ~24 men, without the usual interval between for the
intended deployment, would explain how the Thebans formed in 24 (25)
ranks as at Delium. Two of those stacked gives to 50 ranks for
Leuktra.

Note as well that: "At the same time Flamininus also, having received
his advanced party into the intervals between his maniples, charged
the enemy."

If you have never read any of the hellenistic tacticians, you can
download Aelian for free:
http://books.google.com/books?id=6tITAAAAYAAJ&printsec=frontcover&dq=Aelian+tactics&source=bl&ots=YMGdn-asYK&sig=9HH51vsfiC-VaycRo9RlXb1SXAg&hl=en&ei=5dZYTPihPIP2tgPB3r2hCg&sa=X&oi=book_result&ct=result&resnum=4&ved=0CCAQ6AEwAw#v=onepage&q&f=false

[Paul Bardunias]

> Files of hastati that retire through doubled files of principes must
> get clear before the doubled file can move up to fill the 4' gap.

Not for the type of doubling you will see in Aelian where every other
file simple steps sideways behin the man on their right. There would
be only moments as the princeps steps laterally to replace a
retreating hastati.