Alexander's Seventh Phalanx Battalion Milns, R D Greek, Roman and Byzantine Studies; Summer 1966; 7, 2; ProQuest pg. 159

Alexander's Seventh Phalanx Battalion 
R. D. Milns 

A SOME TIME between the battle of Gaugamela and the battle of 
the Hydaspes the number of battalions in the Macedonian 
phalanx was raised from six to seven.1 This much is clear; 

what is not certain is when the new formation came into being. 
Berve2 believes that the introduction took place at Susa in 331 B.C. He 
bases his belief on two facts: (a) the arrival of 6,000 Macedonian 
infantry and 500 Macedonian cavalry under Amyntas, son of 
Andromenes, when the King was either near or at Susa;3 (b) the 
appearance of Philotas (not the son of Parmenion) as a battalion 
leader shortly afterwards at the Persian Gates.4 

Tarn, in his discussion of the phalanx,5 believes that the seventh 
battalion was not created until 328/7, when Alexander was at Bactra, 
the new battalion being that of Cleitus "the White".6 Berve is re­
jected on the grounds: (a) that Arrian (3.16.11) says that Amyntas' 
reinforcements were "inserted into the existing (six) battalions 
KC1:TCt. e8vr(; (b) that Philotas has in fact taken over the command of 
Perdiccas' battalion, Perdiccas having been "promoted to the Staff ... 
doubtless after the battle" (i.e. Gaugamela).7 The seventh battalion 
was formed, he believes, from reinforcements from Macedonia who 
reached Alexander at Nautaca.8 

Now all of Tarn's arguments are open to objection; and I shall 
treat them in the order they are presented above. 

1 Arrian 3.2.9 for Gaugamela, where all 6 battalions are named; 7 battalions are named 
by Arrian at the Hydaspes battle: those ofCleitus and Coenus, 5.12.2; of Poly perch on and 
Alcetas, 5.11.3; of Meleager, Attalus and Gorgias, 5.12.1. I am not here concerned with the 
question of whether Antigenes now commanded Coenus' battalion (Tarn II.190 and 146); 
the important thing is that 7 battalions are here named. 

2 Das Alexanderreich auf prosopographischer Grundlage I (Munchen 1926) 115-116: hence-
forth cited as BERVE. 

8 Reinforcements: Diod. 17.65.1; Curt. 5.1.41-42; Arrian 3.16.10. 
, Philotas: Arr. 3.18.6. 
5 Tarn, Alexander the Great II (Cambridge 1948) 142-148; hereafter referred to as TARN II. 
tI Cf Arr. 4.22.7. 
7 Tarn II.143. 
8 Tarn II.147; cf especially his n.3. 

159 



160 ALEXANDER'S SEVENTH PHALANX BATTALION 

(a) Arrian does not say anything about "the existing (six) bat­
talions." Rather, he says TOVS 7T€~OVS 8~ 7TPOUE87JK€V 'Tex.L!) 'Td,~€UL 'Tex.tS 

&AAex.LS, Ka'Td. €8V7J €Kd,U'TOVS ~VV'Td,~ex.s-literally, "he added the infantry 
to the other battalions, assigning each man according to nationality." 
The emphatic position of the words 'Tex.Ls &AAex.LS strongly implies that 
there was a battalion or battalions over and above the ones to which 
additions were made. 

(b) The earliest mention in Arrian of Perdiccas as a uWfLex.'TOtPVAex.~ 
(which is what Tarn means by "the Staff") is not until three-and-one­
half years after the battle of Gaugamela, when the army was attacking 
the Rock of Chorienes9 in the Spring of 327. The appointment may 
have been made at the time of the Philotas affair, since Curtius 
(6.8.17) speaks of Perdiccas here as an armiger. But one cannot have a 
great deal of faith in Curtius' terminology; thus two sections later 
armigeri is used of a detachment of soldiers. However, such specula­
tion need not detain us; for at the battle at the Persian Gates Arrian 
specifically mentions Perdiccas' battalion as participating in the attack 
on the Persian Camp with Alexander; while Philotas is equally 
specifically mentioned in the same passage as being sent at this time 
to the Araxes with Amyntas and Coenus to bridge the river (Arrian 
3.18.5-6). When this is recognized, Tarn's argument for Philotas as 
being the successor to Perdiccas in the command of Perdiccas' brigade 
falls straight to the ground.10 

(c) Tarn's argument for the seventh battalion being formed at 
Bactra is based on the belief that reinforcements of Macedonians were 
received soon before this time at Nautaca. For this the crucial passage 
is Arrian 4.18.3: EclJ7TOALV 8~ Kat • E7T6KLAAov Kex.t M€vloex.v is Mex.K€OOVlaV 
., \ \ \ '116 C:-' 'A" l: T b l' €K7T€fL7TEL, 'T7JV U'Tpex.'TLcx.V T7JV €K lflaK€OOVLex.S ex.V'TlP ex.vaSOV'Tex.s. arn e leves 
that it was this u'Tpex.nd. iK Mex.K€8ovlex.s from which the new battalion 
was formed at Bactra (together with supplements for the other six). 
To get rid of the overwhelming difficulty raised by the problem of 
the time factor-it would have been absolutely impossible for the 
three men to return to Macedonia, then come back to Bactra with the 
army in the time that it took for Alexander to move from Nautaca to 

o Arr. 4.21.4; Chorienes is to be found in Berve under Sisimithres, no.708. 
10 It may be noticed en passant that appointment as a CTW/4ar0<pvAag need not, as Tarn 

seems to believe, disqualify a man from the command of a military unit. There is the 
example of Ptolemaeus, son of Seleucus, (Berve 00.670) described at Arr. 1.24.1 as a 
CTW/4ar0<pvAag and found as a battalion leader at Issus, where he met his death. (Cf. Arr. 
2.8.4; 2.10.7.) 



R. D. MILNS 161 

Bactra-Tarn resorts to the expedient of claiming that the three com­
manders met the army en route, "perhaps at no great distance" from 
Nautaca: that, therefore, "they were not of course sent ES MaKE8ovtav, 

as the text says-that has merely been put in from EK MaKE8ovlas." 11 

This highly suspicious line of argument has recently been followed 
and taken even further by P. A. Brunt in a recent article, in which he 
conjectures that there reached Alexander at Nautaca in the a-rpana 

EK MaKESov{as ca. 6,000 infantry and 500 cavalry.12 Now there is no 
justification for Tarn's excision of ES MaKESov{av. Repetitions of words 
and proper names in this fashion, though stylistically clumsy, are a 
regular part of Arrian's method of writing, intended to give greater 
clarity.13 Further, apart from the fact that there is no record in our 
sources of any such army reaching Alexander at this time, it is in­
teresting to note that of the three commanders only one is ever 
mentioned again in Alexander's lifetime. This is Menidas,14 who 
reached Alexander at Babylon in 323, in the company of the finance 
officer, Philoxenus, and the Carian satrap, Menander, with a force of 
cavalry (Arrian 7.23.1). 

In the light of these considerations I feel that Tarn's arguments for 
328/7 as the date of formation of the seventh battalion must now be 
treated with a great deal of caution. 

It seems only reasonable that, unless Alexander was going to form 
the seventh battalion by drawing detachments from the original six 
(and thus decrease considerably the numerical strength of each 
battalion), the only time that this formation could be created is after 
the receipt of reinforcements of Macedonians.15 Now between 331 

(Gaugamela) and 326 (Hydaspes), we know of six occasions when re­
inforcements reached Alexander.16 On only one of these occasions do 
we hear of Macedonians among the reinforcements; and this was in 
331 at Susa. It must therefore follow logically from this that Berve 

11 Tarn II.147 n.3. 
12 P. A. Brunt,JHS 83 (1963) 27ff, "Alexander's Macedonian Cavalry"; see especially the 

table of reinforcements on p.39. 
13 See, for example, 2.14.1ff; 5.2.1. 
14 For Epocillus, see Berve nO.301 ; for Menidas, Berve no.508; for Sopolis, Berve no.736. 
16 Berve's hypothesis, 1.116, of an "Angliederung von Nicht-Pezhetairen an die make-

donische Phalanx" is untenable. The phalanx retained its purely national character right 
till the end of Alexander's reign, when the King, after dismissing the 10,000 veterans at 
Opis, began to experiment with a new mixed phalanx (Arr. 7.23.3-4). The only Macedonian 
infantry who can be called "Nicht-Pezhetairen" (in the sense that Berve here uses the word, 
i.e. members of the phalanx-battalions) were, of course, the hypaspists. 

16 See, most conveniently, Berve !.lS2, for a table of reinforcements and references. 



162 ALEXANDER'S SEVENTH PHALANX BATTALION 

(1.116) was correct in his conjecture of 331 for the formation of the 
seventh battalion. 

There is, moreover, a piece of indirect evidence to be found in 
Arrian that lends strong support to this conclusion, and which, so far 
as I know, has not previously been noticed. This occurs in Arrian's 
account (3.19ff) of Alexander's campaign in pursuit of Darius, which 
began in Spring 330 from Ecbatana. First, however, we must 
make some attempt to assess the number of Macedonian infantry 
(exclusive of hypaspists) that Alexander had when he set out for 
Ecbatana. 

It is almost certain that the battalions at Gaugamela were at full 
strength (i.e. 9,000+ altogether), since the numbers that Alexander 
had for the battle, as given by Arrian,17 are virtually the same as the 
numbers of the army at the beginning of the expedition.Is The only 
known reinforcement for the Macedonian infantry between 334 and 
331 occurs at Gordium, where 3,000 infantry and 300 cavalry arrived 
at the army.I9 Added to the original paper-strength of the phalanx, 
these would now give a total of 12,000 phalangites at the battle. From 
this, total losses must now be deducted (amongst whom can be 
included Macedonians detached from the main army on garrison 
duty). In this respect the sources are not always helpful because of 
the tendency to ·write down' the Macedonian losses and exaggerate 
the enemy's with a view to the glorification of the King; it is reason­
able in such cases to assume that the highest figure given in the 
sources for the Macedonian losses will probably be nearest the truth. 
Also where garrison troops are left behind, it is not always made 
clear who comprised the force. The following tables will give the 
figures most conveniently for Macedonian infantry troops lost to the 
army from combat service either through death or garrison duty 

17 Arr. 3.1Z.5; ca. 40,000 infantry and 7,000 cavalry. 
18 That is, the 3Z,000 infantry and 5,100 cavalry as enumerated in the army list by 

Diodorus (17.17.3-4) + the ca. 10,000 advance guard (Diod. 16.91.Z and Polyaen. 5.44.4). 
These latter included a contingent of ca. 1,000 cavalry; see P. A. Brunt, op.cit. 33f and my 
own forthcoming article in ]HS 1966. 

19 Arr. 1.Z9.4, specifically named as Macedonians. The reference in Callisthenes fr.33 
(Jacoby) to reinforcements reaching Alexander fdMov-n (8') ~ls KrJ..tKtav €f'fJ&Mnv must 
mean Gordium, which was the last important stopping place, apart from Ancyra (which 
merits merely a passing reference in Arrian), before the army invaded Cicilia. I feel that 
Callisthenes' totals of 5,000 infantry and 800 cavalry (of unspecified nationality) can be 
reconciled with Arrian's totals, if it be accepted that Callisthenes' totals include non­
Macedonian units, such as the zoo Thessalians and 150 Elians mentioned by Arrian, and 
other units (e.g. mercenaries) not mentioned by Artian. 



R. D. MILNS 163 

from the crossing into Asia up to, and including, the battle of 
Gaugamela.20 

The Granicus 
Halicarnassus 
Issus 
Tyre 
Gaugamela 

Total 

Halicarnassus 
Celaenae 
Egypt 

Total 

LOSSES IN BATTLE 

ca. 30 (Arr. 1.16.4) 
10021 

ca. 300 (Diod. 17.36.6)22 
ca. 400 (Arr. 2.24.4)23 
ca. 500 (Diod. 17.61.3)24 
ca. 1,330 

GARRISON TROOPS 

ca. 300?25 
ca. 150?26 
ca. 1,00027 

ca. 1,450 

Together these two totals add up to 2,780-say 2,800 for a round figure 
-Macedonians lost to active service between 334 and 331. The bat-

80 Cf Berve 1.179-181. I assume that in battle losses, where our sources give merely a 
figure, this figure refers to Macedonian dead only. This assumption is not an arbitrary one; 
in the major battles and military operations the brunt of the fighting was borne-with the 
exception of one or two small specialised units, such as the Agrianians-by purely 
Macedonian troops. It is, of course, impossible to say what proportion of the infantry losses 
were from the phalanx and what from the hypaspists. 

11 This is a very conservative estimate, since 40 men were killed in one sally alone by the 
defenders (Arr. 1.22.7). 

IS Cf Arr. 2.10.7, who says that 120 Macedonians-the context shows that they were 
phalangites-died in one stage of the battle alone. 

23 Again, the context shows that the ca. 400 were Macedonians: cf 'TWV /Lv, Tvptwv • •• 'Tw" 
MaK€86vwv 8E: ••• is'T€'TpaKocrlovs. 

U These would be mainly on the left wing of the phalanx, which received a severe 
mauling from Mazaeus. Note that even on the victorious right wing ca. 100 cavalry­
Companions seem to be meant by Arr. 3.15.6-were killed. 

26 Arr. 1.23.6 says 3,000 mercenary foot and ca. 200 cavalry were left at Halicarnassus as 
the garrison for the city and the rest of Caria. It seems highly probable that they contained 
at least a stiffening of Macedonians; I have assumed ten per cent. 

u Arr. 1.29.3 merely says ca. 1,500 soldiers. Again I have assumed a mercenary force 
with a stiffening of Macedonians. 

27 Curt. 4.8.4 gives this figure. In view of the great strategic importance of Egypt-and 
perhaps also warned by the recent experiences of Amyntas, son of Antiochus, of the 
attractions of Egyptian wealth to mercenary soldiers under an unscrupulous leader-the 
proportion of Macedonians left in the country is likely to have been much higher. This 
seems to be borne out by Alexander's administrative arrangements for the country as 
given by 3.4.3ff, where the Aetolian, Lycidas, is put in command of the mercenary forces, 
but two distinguished Macedonians, Peucestas and Baiacrus, were appointed as CT'Tpa'T7}"ol 
of the whole army of Egypt, perhaps of Upper and Lower Egypt respectively, and inde­
pendent of the civilian administration. 

6-G.R.B.S. 



164 ALEXANDER'S SEVENTH PHALANX BATTALION 

talions of the phalanx were thus each somewhat more than 100 men 
overstrength before Gaugamela, and still each about 30 men over­
strength after the battle.28 

Between the end of the battle and the army's departure from Persis 
early in 330, the following losses of manpower to the army are known: 

Persian Gates 

Babylon 
Susa 
Persepolis 

Total 

LOSSES IN BA TILE 

ca. 100129 

GARRISON TROOPS 

700 (Curt. 5.1.43; Diod. 17.64.5) 
1,000 (Curt. 5.2.16) 
3,000 (Curt. 5.6.11)30 

ca. 4,800 

Thus the phalanx, between Gaugamela and Ecbatana, received 
6,000 reinforcements at Susa and lost from active service ca. 4,800. 
As has already been shown (see n.28), the probable strength of the 
phalanx after Gaugamela was 9,200 men. Between Gaugamela and 
Susa, where the 6,000 reached the army, a total of about 1,800 men 
were taken off active service, and a further 3,000 soon after at 
Persepolis. This gives a total of ca. 10,400 phalangites left on active 
service when the army left Persepolis, and this figure, I would 
suggest, must represent seven phalanx battalions of ca. 1,500 each. 
Whether the introduction of the new battalion was actually at Susa 
on the arrival of the 6,000 (as with Berve), cannot be stated with 
certainty. The descriptions of military reforms at this time, as given 
by Arrian, Curtius and Diodorus,31 seem to make this highly likely. 
The battalions would certainly be overstrength for a time,32 but 
Alexander must have had a very good idea of the opposition he was 

18 That is, if we take the reinforcements and losses (both deaths and garrison troops) as 
far as Gaugamela, we get a battle strength of 9,700 for the phalanx; after the battle this 
drops to 9,200. 

111 No figures are given in the sources, but both Diod. 17.68.2-3 and Curt. 5.3.18ff' em­
phasise (what Arrian glosses over) that the Macedonians suffered heavy casualties in the 
first attack on the wall. 

80 Berve, ll.278 s.v. NICARCHIDBS, no.563, believes that this number is exaggerated. I 
cannot see why; the difficult entry into Persis and the unsettled condition ofthe very heart 
of the Persian empire must surely have demanded a much stronger garrison-and a much 
more reliable one-than elsewhere. 

al Art. 3.16.W-U; Curt. 5.2; Diod. 17.65.2-4. 
81 Nearly 1,930 men per battalion. 



R. D. MILNS 165 

going to meet in Persis, with the need for a large detachment of 
Macedonians to garrison the country and a consequent reduction of 
battalion strength to normal. No doubt many of the 3,000 left in 
Persepolis were, as at Sus a, "aetate gravibus" (Curtius 5.2.16). 

We may now return to the piece of evidence which, in my opinion, 
affords conclusive proof that the seventh battalion had been intro­
duced at least before Ecbatana, in the Spring of 330. When the army 
advanced to Ecbatana, the monies that had been taken since 
Gaugamela were transported there to be handed over to Harpalus. 
Parmenion, with a strong force of mercenaries, was appointed to 
guard the treasury. However, it was first necessary to clear up the 
tribes in the mountainous area along the southern shore of the 
Caspian. Accordingly, Parmenion was ordered to take his mercenaries 
and make an expedition which would reduce the tribes along the 
western part of the southern shore, while Alexander himself set out 
in pursuit of Darius (Arrian 3.19.7-8). When Darius had been either 
captured or killed, Alexander himself would turn back into the 
mountains and reduce the eastern part of the southern shore.33 

Whilst Parmenion and his troops were away, the task of guarding 
the treasury was entrusted to ca. 6,000 Macedonians, with some 
cavalry and a few light infantry. On Parmenion's return these 6,000 
were to be relieved of their task and make their way eastwards (to 
Parthyaea) to join up again with Alexander. They were to be led by 
Cleitus "the Black," at that time ill in Susa, but expected to have 
recovered by the time of Parmenion's return.34 Unless evidence can 
be produced to show that an increase in the manpower of a phalanx 
battalion was made about this time, the 6,000 Macedonians can only 
represent four battalions that were left at Ecbatana. 

Now the troops that Alexander took with him on his pursuit of 
Darius and subsequent operations in Hyrcania are named by Arrian 
(3.20.1) as follows: the Companion Cavalry; the Prodromi; the 
mercenary cavalry of Erigyius; the Macedonian phalanx, "apart from 
those assigned to the treasury"; the archers; and the Agrianians. The 
"Macedonian phalanx" must include the hypaspists who, in fact, are 
mentioned along with their commander several times during the 

33 This follows from the course of military events after Darius' death. 
3' Arr. 3.19.7-8. The 6,000 did rejoin Alexander, as their instructions stated, at Susia, on 

the Parthyaean-Arian border; cf Arr. 3.25.4: op.oii ij01J EXWV -ri]v 7Taaav ovvaJ.Uv. They have no 
connection with the 7,000 foot-soldiers who reached Alexander in 326 at the Hydaspes 
(Curt. 9.3.21). 



166 ALEXANDER'S SEVENTH PHALANX BATTALION 

pursuit.Sli The rest of the phalanx, after deducting the hypaspists, 
amounted to three battalions, as can be seen from the battalions 
which are named with their commanders during the pursuit and 
operations in Hyrcania. These are those of Craterus, Amyntas, and 
Coenus; no other battalions or battalion commanders are mentioned 
by Arrian either during the pursuit of Darius or in the subsequent 
operations.S6 There were thus certainly seven battalions in the 
Macedonian phalanx with Alexander by Spring 330.37 

That the introduction of the seventh battalion is not mentioned in 
our sources (especially Arrian) need not cause surprise when it is 
remembered that other, much more important military reforms 
receive no mention (cf. the reorganization of the cavalry into five 
hipparchies at some time before the Indian campaign). Again, to the 
best of my knowledge, it is nowhere specifically stated in our sources 
that six battalions of the phalanx crossed into Asia with Alexander. 
This is a deduction based on the known numbers of the Foot Com­
panions and the known names of battalion commanders as given in 
battle lists. Likewise we are left to deduce, from our available evi­
dence, the introduction of the seventh battalion. 

UNIVERSITY OF NEW ENGLAND, 

NEW SOUTH WALES 

March,1966 

36 cf. Arr. 3.23.3, 3.2I.S. 
3S For Craterus and Amyntas, see Arr. 3.23.2; for Coenus and Amyntas, see Arr. 3.24.1. 

That there were only the battalions of Craterus, Amyntas and Coenus with Alexander is 
based not only on Arrian's silence; with the number of Macedonian troops available to 
Alexander, as I have demonstrated, there could not have been more than these. 

37 The above arguments might well be used to show that in fact Perdiccas was appointed 
"to the Staff" as early as Ecbatana (but after the Persian Gates-see pp.159f), since so senior 
a battalion leader as Perdiccas would probably not be subordinate to a squadron leader of 
the cavalry. However, it is perhaps possible to underestimate the importance of both 
Cleitus himself and of his position as leader of the Royal Squadron. It was, after all, Cleitus 
and not one of the senior battalion leaders who was appointed, along with Hephaestion, 
as hipparch after the Philotas affair (Arr. 3.27.4). Perdiccas, Craterus and Coenus had to 
wait till conSiderably later before receiving a hipparchy.