20, \S 5. tanti .~.~. co1ldonet. Caesar could not yet afford to
punish Dumnori2l: (he found an opportunity of doing so four
years later [v, 6-7]) for fear of offending the patriotic larty

· la8mo\S n3g) thes Aedui, with whom Dumnori~ was popular (3, \S 5;

21, \S l. sub mo11te. This hill must be identified with Sanvigne,
~bout 6 miles east of the river Arroux: for, as we shall see in
the note on 24, ~ 1, Caesar's ne2~t camp was hard by Toulon-sur-
Arrou~; the march by which he reached it was very short, a~
we may infer from the fact that the ~elvetii, whom he followed
took a fortnight or more to advance with their unwieldy wagon-
train from the point where they crossed the Saone to the
neighbourhood of Toulon (15, \S 5); and Sanvigne is the only hill
east of the Arrou~ and within a short march of it which answel~
to the descliption in 21, \S 1 and 22, \S 3. See Stoffel's Hist. de
Jules C~sa1;--Gue17e Civile, ii, 1887, p. 445.

\S 2. Iegatum pro praeto1e. Labienus was not only the ablest
of Caesar's generals, but the highest in rank. As legatus p1'0
praetore, he would be Acting Governor of Gaul and Commander-
in.Chief in the winter, when Caesar v~ as in Italy. See 54, \S 2.
\S 4. L. Sulla was the famous dictator who overthrew Marius.
M. Crassus ~as the millionaire of Rome, who, with Caesar and
Pompe~, formed the first Triumvirate, and who had defeated
Spartacus, the leader of the rebel s]aves, in 71 B, C.

22, \S 2. a (Gallicis armis) in this sense is estraordinary and
perhaps was not in the original manuscript. Indeed it is omitted
in L. If Caesar followed the usage of classical prose, he wrote
either (~allicis ar~is or ex ~allicis armis. Insig~ibus here means
' crests '.

23, \S 1. cum .~.~. oporteret. Even in this passage cum does
not tell us only ' how one action is related to another with
regard to the time of its occurrence ' (see the note on 4, \S 3).
It does not tell us onl~ that the rations would be due in 48 hours:
it tells us that their distribution would be a consequence_of the
expiration of 48 hours. The difference is subtle, but rea].

r~etiri. See the note on 16, \S 5.

e~istimavit, If the MS. reading is right, itao~Me, which Meusel

inserted in 1894, is evidently required, but he now adopts an
old emendation,--existimails.

\S 2. e~uitl~m Ga~lorum. The adjective of G'allus is Gallicus,
but Caesar never uses it as an epithet of living beings. The
phrase equites Gall~ is like mare Oceanus.

\S 3. e~lstimarent .~.~. co~derent. See the note on 6, \S 3.
Meusel (J.B., 1894, p. 332), remarking that both Caesar and
Cicero often use such verbs in the subjunctive where logically
the indicative might seem preferable, says that, strictly speak-
ing, the clause should run either seu quod .~.~. existi~labant or
seu quod, ?~t existimabant, Romani discederent.

24, \S l. proxim~ collem. This hill can be identified. When
Caesar struck his camp on the morning of the battle he was not
more than 18 Roman miles--and we may safely assume that he
was not much le~s--from Bibracte (23, \S 1), which was situated
on ~ont Beuvray (see p. 411): when he was attacked he was
marching towards that town, and the Helvetii, who, on hearing
of his movement, had reversed their march (23, ~\S 2-3), were
trying to intercept him. Evidently therefore, the battle-field
was close to a point where a road ieading to ~Iont Beuvray
diverged from the route by which the EIelvetii had been marching
to Saintonge (10, \S 1). It was also somewhere south-east south
or south-west of Mont Beuvray, bccause the Eelvetii couid have
had no motive for passing by the east and north of Bibracte in
order to reach the Loire, which they would be obliged to cross
on their way to Saintonge. It is now generally agreed that
the only site which fulfils all the conditions is about 3 miles
north-west of Toulon-sur-Arroux, and that the hill on the lower
slopes of which (~aesar formed his army was the hill of Armecy.
The site was first pointed out in 1867; and some years later
Stoffel discovered by e~cavation an entrenchment on the plateau.
This entrenchment could only have been intended to serve a
temporary purpose, for its shape was that of a crescent, the
width from horn to horn being only about 300 yards; and there
are irregularities in it which seem to show that it was con-
structed in a hurry. I am inclined to think that the men who
erected it may have stopped work when they saw that the battle
was going in favour of their comrades: this supposition would

account for its not }~aving been completed and made into an
enclosure.

But althou~h the battle-field is certain, opinions differ about
dctails Stoffel belie~ed that the hill to which the Helvetii fell
back after the failure of their first attack (25, \S 5) was just north
of the village of Montmort; but, as the plan (facing y. 25) will
show, in order to maintain this view, he was obli~ed to argue
that the Boi and Tulingi attacked the Romans on their left
flank, whereas (laesar says that they attacked them on the right
(see the note on 25, ~ 6). Besides, according to Stoffel, in the
first stage of the battle, while the fighting was going on on the
hill of Armecy, the c~tremity of the Roman right wing was
posted on a steep declivity, whereas the Romans preferred a
gentle slope. Colonel Bircher therefore modified Stoffel's theory.
He concluded that the four veteran ~oman legions (24, \S 2) were
posted on the lower slopes of Armecy, facing west-south-west,
and that the hill to which the E~elvetii retreated was on the
further side of the valley through which runs the road from
Toulon-sur_Arroux to Luzy. This was the road by which the
Boi and Tulingi would have marched to reinforce the E~elvetii.
Both Stoffel's theory and Bircher's are illustrated in the plan;
but Captain G. Veith and M. Jullian agree with me in following
Bircher (C. G., pp. 624-7).

\S 2. in colle medio,--' half-way up the hill '.

trlplice~ aciem inst~uisit. This was the normal formation
though Crassu~ in Aquitania formed his army for battle in two
lines (iii, 24, \S 1), probably because his troops were compara-
tively few; while Caesar in Africa once deployed only a single
line (~ell. Afi:, 13, \S 2), and in the battle of Pharsalia, for a
special reason, improvised a fourth (B. C., iii, 89, \S 3). In that
battle, according to Frontinus (ii, 3, \S 22), each line in Pompey's
army was tcn men deep. Frontinus seems to mention this as
e~ceptional; and accordingly it has been conjectured that the
normal depth of a cohort was eight men (C. G., pp. 587-8).

The thou~htful reader will probably ask himself certain
questions, which are suggested by (~aesar's account of the battle
with the ~elvetii, but which he does not answer, I suppose
because he ~vrote for his contemporaries, most of whom perhaps
knew enough about warfare to understand his book. Thc
soldiers in the front rank of the fighting line must have become
tired after, say, twenty minutes of hand-to-hand fighting. How
were they relieved? When and how did the second line come
into action ? What was the use of the third line ? The first
question and the last can be answered easily. When the fore-
most ranks became tired, or thinned by the loss of individual
soldiers, the rear ranks advanced between the file~ of those in
front, and relieved them. Superintendent Froest tells me that
this method would be adopted by the police in street fighting;
and indeed no other method appears practicable. The third
line served as a reserve: how it was used in this battle will be
seen in 25, \S 7; generally, when its services were required, as,
for instance, in the battle with Ariovistus (S2, \S 7), tbe cohorts
that composed it may have been directed against one or both
of the enemy s flanks or on his rear. It is the second question
that has most perplexed the commentators. Rudolf Schneider
has tried to prove that as soon as the light-armed au~iliaries
had done their work and hand-to-hand fighting had begun, the
second line was regularly incorporated with the first. But in
fhe battle with the ~elvetii the auxiliaries were far removed
from the fighting line: if, then, the second line wa~ from the
commencement of the battle incorporated with the first, why
was it formed at all, and what was the sense of the e2~pression
t~ iplicem acie~n ? It is clear from Caesar s narrative (25, ~ 7) that
even after the first stage of close fighting was over tbe firsttwo
lines still remained distinct. My own belief is that the second
line as well as the third acted as a reserve, that if and when
the fir~t line needed support, the second was incorporated with
it, and occasionally perhaps the third also. In the first stage
of the battle each cohort of the first line probably formed a
separate group: it would have been dangerous for the enemy
to attempt to penetrate the spaces between the groups, for they
would have been liable to be attacked and cut off by the fresh
cohorts in reserve; and, on the other hand, these could ad~lance
when they were wanted into the spaces and reinforce the first
line. It is not my business to give reasons for this opinion
here, for in doing so I should have to discuss many passages
which are not in Caesar's text: bnt the whole problem, which
is interesting, is thoroughly worked out in C. G., pp. 588-99.

\S \S 2-3. ~ s14mmo i~go .~.~. sarcinasque. The MS. reading,
which is untranslatable, is (ipse interim .~.~. veteranorum) ita ~ti
s~pra se [v. 1. sed] in summo i?~go .~.~. auxilia conlocar1, ac totu~
mo~item hominib~s comple~i et ~nterea sarcinas (in unum locum .~.~.
iussit). An attempt has been made to amend this by changing
conlocari and compleri into co~locaret and compleret. Meusel
deletes ita ~ti supra se and brackets ac totum .~.~. interea, which,
as ll~lotz remarks (C.S., p. 239), is a desperate remedy. I have
adopted Klotz's conjecture, which is at all events ingenious.
Fortunately the general sen~e of the passage is in any case
clear.

~ 3. sa~cinas means the bundles (analogous to knapsacks)
which the soldiers carried (see p. Ixv). Accordingly Stoffel
supposes that the heavy baggage (impedime~ta), which, as the
reader will have gathered from the preceding note, is not
mentioned in the MSS., had been sent on under a small escort
to Bibracte. But we should have e~pected Caesar to tell us
this. Moreover, as his army remained on or near the battle-
field for three days after the victory (26, \S ~ 5-6), it seems
reasonable to suppose that they must have wanted some of their
heavy bagga~e. On the other hand, as only two days' rations were
left (23, ~1), the baggage-cattle may have been sent to Bibracte
to fe~ch corn. The entrenchment on the hill of Armecy
was not large enough to protect the entire baggage train
(C. G., p. 628).

\S 5. phalange. The men in the front rank held their shields,
which overlapped, before their bodies, while those behind bore
theirs horizontally over their heads. ~f. Livy, x, 29, \S 6.

25, ~1. om)~iur~l evidently means only the mounted officers of
the legions, not the cavalry.

2. pllis. See p. Ixiv and C. G., p. 599.

~ 3. cl~m .~.~. in~e~i~set. See p. Ixiv. When cum is used in describ-
ing repeated action, it is generally coupled with an indicative.
The subjunctive, as Meusel observes (J. B., 1894, p. 371), is here
not only iterative but causal. Caesar does, however, occasionally
use the subjunctire in a purely iterative sense, e.g. in B. C., iii,

\S 5. spatio, which is evidently required, was supplied by
B. Dinter. Schneider's attempt to defend the MS. text fails.

\S 6. novissimis praesidio erant. These words have generally
been taken to mean that, after the emigrants retraced their-
steps (23 \S 3), the Boi and Tulingi served as the rearguard of
the whoie column, including the wagon-train, which they
marched past in order to come into action. But it is very
doubtful whether the first stage of the battle las~d~ong
enough to enable them to do this. M. Jullian understands the
Latin in the sense that the Helvetii had left the Boi and
Tulingi on the road to goard the wagons in front of which
they had been marching. I believe that Le is right- but if
so, the wagons must surely have been protected in their rear b~
another force,which took no part in the battle (C. G., pp. 629-30).

ex itinere shows that the attack began immediately after the
march ended. I translate thus:--'marched up, immediately
attacked,' &c.

<ab> latere apeito. The insertion of ab, as Meusel shows
(J. B., 1894, p. 299, with which cf. L. C., i, 36-9), is necessary.
The words mean ' on the right flank ', which was e~posed
(aperto) because the shiel(3 was worn on the left arm. This was
denied by Stoffel, who nlaintained that the words simply meant
'on the exposed flank',--left or right, as the case rnivht be;
but there are at least three passages in Caesar--iv, 26, \S 3-
v, 35, \S 2; and vii, 82, \S 2--which prove that ab latere aperto i's
a technical military phrase, and means what I have said. For
the troops which are mentioned in each of these passages were
exposed, as far as their position was concerned, on their left as
well as on their right- and therefore either ab latere aperto
signified 'on the right and unshielded flank' or it signified
nothing. The passage on which this note is written is fully
discussed in C. G., pp. 621-3, the arguments in which have been
accepted on the Continent as conclusive.

\S 7. Mommsen (J. B., 1894, p. 201) gives a sufficient reason
for regarding conversa ai an interpolation:--' the first two lines
did ilot change front.'

2~, ~ 2. ab hora septima. The Romans divided the period
between sunrise and sunset into 12 hours, ~vhich of course were
only equal to our hours at the equino~.

pugnat~ sit. If Caesar had written pilgnaretu~; he would, so
to speak, have been placing the reader at the standpoint of a
spectator of the battle; whereas the perfect nnerel~r states that
the batt]e lasted seven hours without calling upon the reader to
form a mental picture of it. See J. B., 1894, p. 3S7.

~ 3. rotas seems at first sight superfluous; and some editors
adopt Meiser's emendation, ra~das (cf. 51, \S 2). Perhaps, how-
ever, (inter carros) rotas(que) may be defended on the analogy
of i~ter carros impedimentaque (iv, 14, \S 4), if we may suppose that
in the latter case the baggage was i~ the carts; and, moreover,
the wheels may have played a part in the defence. See p. 436.

\S 5. E~ eo proelio .~.~. perve~ unt. Meusel (J. B., 1910, p. 39)
rightly brackets nullam parte~n .~.~. ii~termisso. The words are
absolutely super~uous if noctis denotes the same night as ea tota
nocte; and if Caesar had meant to describe a succession of night
marches and to imply that the Helvetii rested by day, which is
more than improbable, he would have written not noctis but
noctiur~ (C. G., pp. 632-3). ~Ieu~el (J. B., 1910, p. 55) also
regards the words triduum morati as interpolated, because,
being followed in the next sentence by triduo intermisso, they
are unnecessary. Suspicious they certainly are- but I can
conceive that, after writing them, Caesar wrote tr~duo inter-
tnisso, and forgot that he was repeating himself.

In translating die ~uarto we must remember that the Romans
as a rule reckoned inclusively. Thus if the battle was fought
on a Sunday, the Il:elvetii reached the country of the ~ingones
on Wednesday; but it is impossible to tell what point in that
country they had reached when Cae~ar overtook them. They
may have retreated to Dijon, the place which they would
naturally have made for if as is probable, they already
intended to return to Switzeriand. See the note on 30, \S 5, and
~. G., pp. 631-4.
27, \S 4. ea. See the note on 29, \S 2.

conquirl~ntur refers to obsides and servos, conferuntur to a~ma.

Schneider tries to reconcile nocte intermissa with prima nocte
by e~plaining the former as meaning 'after night began to
intervene'; but, as Meusel points out (J.B., 1910, pp. 56-7),
it can only mean ' after a night had passed '.

in ta~lta ml~ltitudine is virtu ~ equiva]ent to cum tanta multi-
tudo esset.

e~istimarent. See the note on 23, ~ 3.

28, \S 1. in hostiurl ll umero habuit- Probably the 6,000 fugitives
were put to death. Cf. Cicero, Veiv:, ii, 5, 25, \S 64; 28, \S 73;
Cat., iii, 10, \S 25.

\S 5. petentibus Haeduis is not dative, but ablative absolute.
C£ ii, 12, \S 5.
2~, ~ 1. litteris G1aecis. Greek characters were also used by
Druids (vi,