Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

48
MEN-AT-ARMS SERIES 89 BYZANTINE ARMIES 886-1118 rmm:I MILITARY I:\l" HE:\TH ANGUS :\lcBRIDE

Transcript of Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Page 1: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

MEN-AT-ARMS SERIES 89

BYZANTINE ARMIES886-1118

rmm:IMILITARY

I:\l" HE:\TH ANGUS :\lcBRIDE

Page 2: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

MEN-AT-ARMS SERIESEDITOR, MARTIN WINDROW

[i!lm] 89MILITARY

BYZANTINE ARMIES886-1118

Text byIAN HEATH

Colour plates byANGUS McBRIDE

Page 3: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

IlltroductionFor !.he era in which the)' li\'ed the Byzantines had aremarkably sophisticated approach to politics andmilitary strategy. Unlike most of their can·temporaries, they learnt very early in their historythat winning a batllcdid not necessarily win a war,and they frequently bought off their enemies withtreaties and bribes rather than squander men andmateriel in potentially fruitless campaigns. Al­though, evell as carly as the 6th century, thehistorian Procopius bad shrewdly qbservcd that thepayment of tribute to one type of enemy en­couraged the aggression ofanother, still the overallsuccess of this policy is well-testified by theEmpire's survival, despite ilS limited manpowerand frequent internal dissension, right up 10 1453.Besides, since another aspect of Byzantine dip­lomacy \\as the playing off of one enemy againstanother, the attraction of additional foes was onlyrarely a problem which gold and honoul1i, falsifiedletters or sponsored revolts could not solve, and theEmperor's first-class intelligence service, the Officeof Barbarians, kept him well abreast of currentmoods and trends at all times.

Alas, the Empire's contemporaries did notalways understand thc complex motives ofplot andcounter-plot, nattery and threat, which werecsselltia I ingrcd icnts ofByzant inc politics, and mosttcnded to regard the diplomatic manoeuvres andskullduggery of the Emperor and his ambassadorsas underhand and two-faced (which it was)withoUl appreciating its true politico-militaryvalue. The 'bad press' that Byzantium has receivedfrom historians and chroniclers over the lastthousand years has done littlc to enhance itsreputation. to the point where evcn today tortuousand underhand beha\'iour is sometimes describedas 'Byzantine'.

But against this backdrop ofdeceit and intrigue

Byztl1ltineArmies 886-11/8

there is one essential fact that muSt not beforgotten; that such a policy of threat and briberyinevitably presupposed a strong military establish­ment. The Byzantine amlY of the loth and earlyI nh centuries, at the height of its power andefficiency, was the best-organized, best-trained,best-equipped and highest-paid in the knownworld.

ByrA!lwu: warrio... orthe late loth CIt>ltury. UllfortWUltety thelI.rt.;"t;c: slyle ul.itillecl Hlnty....;ps or II O;lroOll dallo;ic:alin8ueoc:e and "tne iaac:c:untcilt'S kaye Lherebr *. in­troduced. The otd.f:....hio.ec1 ptilyltS, ror i.a.tan~, lire hi&bJyi.a:>probabJe lit this dale. The lanle1larc:onelet, orldibaNOII, orthe risht-h.aad 6pre is 1Ic:au-ale ftIO. howevft-.~ thetnusdecllnthu c:onelet orhis c:ompaaion is ora type thai kadprot-bJ)" bent obsolete ror m.a.ay hwtdreds of ,.ears. 80thc:orwdets ....ye thiclr.lnthu strips c:aIled P'tUUl" ("(nth""')~ rrom waisl aad ahouJder. The c:&-ks iadic:allt thaithftllt are probabty hOrHIDnL

3

Page 4: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

OrganizationAlthough the Byzantines clung tenaciously to theirRoman heritage in a great many respects (lheyeven continued to call themselves Rhomaioi orRomans) army organization was not one of them,and as early as the late 6th or carly 7th century,when the Emperor Maurice's military manual, theStralegicoll, appeared, hardly a vestige remained oftbe old Roman military system. The organizationwhich Maurice's work outlined remained praCli­cally unchanged until at least the late loth centuryand probably up until about a century later thanthat, and il is rcpeated almost verbatim in anothermilitary manual, the famous Tactica, written at thebeginning of the I ath century (c. 903) by EmperorLeo V I the Wise.

The basic unit for both cavalry and infantry inLeo's day was the bandon, alternatively called in theearlier Strategicoll a tagma or arithmos (the laner astraight translation into Greek of the Latin lIume­rus). The term bandon itself was derived from theGerman word for a banner, and bears witness to theforeign influence prevalem in the army at the timethat this panicular type of unit evolved in the 6thcentury. Infantry banda consisted of sixteenlochaghiai, each of sixleen men commanded by anofficer called a locllag/ws or 'file leader'; he wasassisted by a dekarcllos, 'leader of ten', a pentarc/lOs,'leader of five', a tetrarchos, 'leader of four', and anouraghos, 'file closer'. Each four lochaghiai con­stituted an allaghion or 'winglet'; these were usuallypaired ow. In heavy infantry units three-quarters ofthe men were spearmen called skutaloi and one­quaner were archers, the archers presumablyorganized as a separate lochaghia within eachallaghion or as a separate allaghion. Light infantryand guardsmen would not have had the splitbetween spearmen and archers, consisting insteadofonly one troop-type; it has even been suggcstedthat light infantry lochaghiai might have com­prised only eight men rather than sixteen.

At the time when the Strategicoll was writtencavalry banda had been subdivided into threehekatolltorchia, each commanded by a lIekatolltorc/IOSof whom the senior acted as second-in-commandand was called an illarches, the bandoncommander-in both infantry and cavalry units-

4

being called a komes or count. By Leo VI's day,however, the hekatontarchion bad disappearedand the bandon was divided instead into sixallaghia (probably commanded by officers calledpenlekon/archai). These were generally paired off asin the infantry bandon and each pair was stillcommanded by a hekatontarchos (or kenlarchos).Each of the six allaghia had fifty men, organized infive dekarchiai of ten men each, comprisingdekarchos, pentarchos, tetrarchos, ouraghos andsix men. On the battlefield the cavalry dekarchiausually formed up in two files five-deep with thedekarchos and pentarchos in the front rank,followed by a rank of lancers, then twO ranks ofarchers, and finally the tetrarchos and ouraghosclosing the files; all four officers were lancers.

Basically, thcn, by the beginning of the lothcentury the standard infantry unit consisted of 256men (sixteen times sixteen) and the standardcayalry uni t 01'300 (six timcs fift y), but the manualsadvise us that unit strength could in fact varybetween 200 and 400. Thosc in excess of officialstrength were apparently not usually taken intoaction and pl'Obably accounted for wounded andsick men and horses and raw recruits. It seems moreproba ble anyway that uni tS generally took the fieldunder-, rather than ovcr-, strength. Standard­bearers, musicians, and officers above the rank oflochaghos and dekarchos do not appear to beincluded in these figures.

One of Empcl'Or Nikephoros II's works (ruled963-g6g) indicates that by the second half of theloth century the cavalry bandon could in fact beonly fifty strong, but this is quile probably a slip ofthe pcn and it seems more likely that the allaghiollis meant. However, it is not impossible that theterm bandon might have changed its meaning inthe sixty-odd years since Leo had written. One ortwo sources also imply that by the late loth centurythe smallest infantry unit may have been ten ratherthan sixteen men (with an archer: spearmen ratioof 3:7), though Michael Psellus' Chrollographio,written in the last quarter of the 11th cenlury, stillrefers to sixteen-man locbaghiai.

At a higher level cavalry (and presumablyinfantry) organization was in moirai (commandedby moirarclwi) or dhollllgoi (commanded by dhoullgariior d/wullgarokometes) and turl1lai or merai (com­manded by tllmlOrclwi and merarchai respectively).

Page 5: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

\\

.'• II'• J. • ,'. I'. ,

The moira or dhoullgos appears to have consisted ofan apparently variable numberofbancla, probablyon average between twO and five, while the turmaor mcros (the lattcr tenn somewhat archaic by theloth centul)') seems to have consisted of threemoirai. The earlier Stra/~gi£o" records the moira ascomprising 2,000-3,000 mcn and the mcros as6,000-7,000, bUI by Leo VI's time we must assum~that the strength of these larger units had declinedconsiderably since even the biggcst theme (aprovincial ann)' corps-sec later section) couldraise only 15,000 horsemen, and the smallest only4,000! Certainly in 838 lurmai of 'not more than2,000 men' arc reported.

PayByzantine soldiers appear to have been generallywell paid. The sources indicate that thematic(pl'Ovincial) troops received one or one and a half1lOmisma{a (gold coins weighing 117'2 ofa pound) permonth, therefore twelve to eighteen nomismata, or1/6-1/4 ofa pollnd ofgold, per annum. In additionthematic soldiers also had grants of land, whicht.947 Constantine VII Porphyrogenitus decreedmust be worth at leasl twO pounds of gold (144nomismata) for Imperial seamen (and, possibly,infantrymen) and four pounds ('288 nomismata) for

Spirited cavalry enp!erneatt fron> the Josl".. Roll, <latin! tothe 6nn half of the loth CftItury. Th.. meat w.,;or hip-Iensth_raete.. with pte....!" at ahol1lder and waist, ph,. iron hel­mel and I....t.hu IoanlKS ofb.--.t-baDd and PouId..r-pieces.Note also t.heir Iart'" shields and kODtaria. Th.. offittr6prt atleft sporta a.. impreui~hehnel CrtSt. (BibliotKa Apo$tolicaVaticaa.a., Rome)

thematic seamen and cavalrymen, this rising totwelve pounds for cavalrymen by the end of thelOth century. According to one source an ad·ditional nomisma was paid for each year ofthematicscrvice up to twelve years. However, a 9thcentury Arab wriler recorded Ihat thematic troo~

were paid only once every three years, or in somecases every four, five or even six years; ConslantinePorpbyrogenitus says that thematic troops weredivided into four groups each paid once every fouryears, but be explains that this was 'the oldpractice', withoUl enlightening us as to what thenormal praclice was in his own day. This seems tomean that thematic troops served for a full year ona rota basis once every three to si~ years, so that asmall core of regular troops was available at alltimes. Alternatively it may indicate a sup·plememary payment made at periodic training or• •mspecuon musters.

In addition, soldiers received rations duringactive service, and occasional special bounties and

5

Page 6: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

•• •

a share of the spoils t'akcn on campaign, whiledisabled men received a pension, and widows ofmen killed in action sometimes received com­pensation in the form ora lump sum (five pounds ofgold in the 9th century).

Rales ofpay for officers in Leo V I's reign wcre asfollows: dckarches received onc pound ofgold perannum (72 nomismala); pentckontarchai (com­manders of fifty) two pounds; komeles (bandoncommanders) three pounds; fifth-class sim/cgoi or'generals', five pounds; fourth-class stralcgoi (thenaval commanders orthe Kibyrrhaiols, Samos andAegean Pclaghos Themes), ten pounds; third-classslralcgoi, [wellly pounds; second-class slratcgoi,thirty pounds; and first-class slratcgoi, [ortypounds (2,880 nomismala). Salaries of hekaton­tarches, moirarchai and Illrmarchai arc not re·corded, those of the laller probably varyingdepending on what grade of general they servedunder. These salaries only apply to the EasternThemes anyway, and officers of the WesternThemes probably received lower pay, drawn frompl'Ovincial taxes rather than coming directly fromthe Imperial Treasury.

EQUIPMENT

Whether in reality such complete equipmem as isoutlined below and in the colour plate captionsactually appeared very often (if at all) seems

6

Avar horseman from a 6th century rock carving.Much Byzantine military equipment was copiedfrom the Avars in the 6th.,.m centuries, includingclothing, stirruplll, borlile-annour. and the wide­spread Ulile oflameUar.

dubious; certainly the pictorial sources do nOtencourage us to believe so. But there is little doubtthat the quality of such equipmem as was issuedwas ofa very high standard, and it is worth quotingthe shrewd observations ofCatacolon Cecaumenus,holder of various military POSts in the EasternThemes, from his own Stralegicon of c. 1070: 'Aboveall,' he writes, 'insist that your horsemen have goodmounts, and complete and well-kept equiprnem,and saddle-girths and boots that fit. For you can besure that a horsem~n with a good horse, a smartuniform and good quality weapons will, if he'sbrave, become doubly so, or-if timid-will takecourage and do his bit. But ifhe is badly equipped,with too big a saddle, boots that don't fit, and agood·for.nothing horse, then you can bejust as surethat however brave he may be the only thing he willbe thinking of is how to save his own skin, by takingflight at the first opportunity.'

AmlOur

We are amply provided with detailed informationon arms and armour of this period by the varioussUl\living military manuals and documents andlarge numbers ofcontemporary illustrations. Fromthese it is clear that the three main forms ofarmourin usc were mail, scale and lamellar, with lamellarpredominating.

Lamellar armour comprised small, basicallyrectangular plates (either long and narrow or very

Page 7: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

nearly square) laced togethcr in rows by threadingleather thongs through punchcd holes, thc rowsthcn being laced to each other overlappingupwards (unlike scale armour, which overlapsdownwards). The lamellac were most commonlyiron, but leather and horn also feature prominentlyin thc sources. The resulting corselel, characterislicof Byzantinc military equipment, was called akLibanion (a name derived from the La tin clipanarills,a heavily-equipped cavalryman), usually sleevelessor shon·sleevcd and reaching only to lhe waist.Some knee·lcngth lamellar corselcls with longsleeves are to be found even in II th centurymanuscripls, but these were rare.

Being stiff and somewhal inflexible, scale cor­selets, where lhey occur in the illustrative sources(and they arc not always casy to distinguish frommail owing to the artistic techniques utilized), tcndto cover only the torso and arc invariably sleeveless.?'lail corselets (called <.abai or Lorikia, cf. LatinLorica) are rarest of all; they are usually depictedknee-length and frequently have sleeves reachingto elbow or wrist. Mail hoods were also worn.Klibania are oflen shown being wom over mailcorselets.

Padded and quilted cotton, leather, wool and feltbody-armour, a minimum of tin thick, was afso inuse under various names, such as epiLorikioll,

kabadion and bambakio1/. All were slccved, and theepilorikion and bambakion at Icast also had hoods.The former was worn over, and thc lattcr under,the klibanion or lorikion. Epilorikia were usuallyworn by cavalrymen, kabadia by infantrymen. Anarrow proof felt eloak is also Il1clllioned in theTactica, as is a thick felt cap.

In place of sleeves, and owing to the faci thatthey arc usually only hip-length, mOSt scale andlamellar corselets (and many mail and quiltedcorselets 100) had hanging strips called plerllges atwaist and shoulder, either oflealher, quilled cottonor even splint-armour. (Helmet aventails weresimilarly often of lamellar, lealher or cottonlappets, though others had mail aventails leavingonly the eyes uncovered.) Forearms and lower legswere protected by vambraces (cheiropseLla or mal/­ikelia) and greaves (podopselLa or chaLkotouba) re·spectively, generally of splint construclion. Thesewere usually iron, but leather, wood and felt wcrealso somclimes used. In addition the tall, square­toed boots, which were a standard part of Byzan­tinc equipmcnt, also had a defcnsive value, being

Two fully-equipped hors~mt:n on a lion hunt, probably inAnatolia or north~1'D Syria, t.ak~n from an ivory casket inTroyefi Cathedral dating from th~ 11th c~ntury.Th~ c:on;d~ts

are pre"wnably 1an>~lIar, though th~ large upward_pointingscal~s are somewhat unusual. Absence of stirrups here muStbe .rti"tic licence since they OCCUr in another panel on theliame casket. (Tresor de I. Cathed....le. Troyes)

7

Page 8: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

infantry was about 5lJ.in deep, that of th~ cavalrysomewhat smaller, apparently 3 40in (which,intercstingly.tallies with a 40tin shield recorded atthe beginning of thc century by Leo). Byzantinemanuscripts firsl begin to depict such kite-shields inthe mid-loth century, where they seem to be abouttwo feet broad at their widest point. Later vcrsionstaper considerably more, in keeping with those inuse in the W t which themselvcs probablyevolved frolll the earlier Byzantine type). Judgingfrom illustrative sources the k..ite became thepredominant shield-type amongst both infant!)'and cavalry during lhe course of the II th ccntu!).

Il'taponsTIl(" RntfilHill/' "lkJif'r', ma.n anm roO,klnl t,rIUllr.- aud '''''JIlt t lw lunuu VoU lht, , .... UI.( ,~~... ~t" l"MUIItlr"o 1"04tll", hu"t ItK". 1­1~1~l'h ;llnll.\L.,P1o, flirt ""f" Ird H Ie" ......, HI" n,",

.anu lI..: ••LUII: IJI lUIucvlud lwl/.l1,'J" foe U1f.llurr

U, n ~ 1'C"1 <t .It -'lid I.t\"b~ ",...door,t.tfNt~

tit ..t.t.M' II" dutlt- Wt ,,In,.-I Ihlh. IhltlwhtllQJ L. we~ d L.u.w::wnltQI Iud du:

frIf..".j,." (....'1.. 1ltf' "'" 1\'1" "'"'-Jluh In ,h,. \'.,IJ t 'r.'KNasa ~nd ,hr ~\;.(( nJ'>"",~ 11"".... uf '.upl...", II I'lIOl;n, ...l.a\ J.n~LIl UM'C1 lit .. IJn~.wtiu.. of Q1 l:.alkd.,...1 in u(h bon mbntn' Imit Tilt:H'·Pln ..,DduWa. u~ m=un.Jcd • ..wl bt!JUIin . hi'" Uhtl".,. t.1M" ...me "."~ ...t..,,a.ppul"nth in th,. .mh rrntUf'1" 001\ ht'2Y\ mnlnWIItd~ lb,.t., k:uJ.llI~ uti tl:uts,. ",I"cb "U1"

urtW:'<1 U1 ., "" II Ifw ~f1.-.11,t" n.nrd .... ill. m.lINlIII I,..m- 'cr. l..illdD IfttJrItJa

lUM.1'UUM11 f"C.ll(·Jtn uK.lbdit'lot' euJlItlutC t.lw ILllL.lit l~l,1j-d ~ ... ,...-.c f,n'-l1 \U'lftlIlnl f'~M" "'bolldrill: nnd !llIn III Ih" Int hlp 'Ib.. othr.r rruln

1!\tIV" ~f ll;\:Uft.t .... ~J' JldJrhUff I ~~w-~ '0"1l:Illpcun ~llhC' II..,. Qtll.'(11Ulr\ It ".h ,11'....r,.ltll) "'Mn-h'IQI'I1 ~u f~1I1 III ltw 11K" t,.nlt,tl 0' th...IUUII~'Il1 4l~1 .... '1\Jllf' I '-luM) .It '!\tiP" ~J" '-It....

l.i1.tfLr Ihr '1Jo1Uu.tn II "'14t ~f1k<1 i,u dtf' 'A14bt.

'"jl~'

:JJ=o,eH- ~'b-.o-_.*1

dnnhll!1t' (if poLld~. 1I..1.IlIII1 kt'1'T' iWlm•.umt::lpAtt IW bot ~Lhn. uudt:u, tU.lUun'f"tJ h.h IUJi• tli.. DiW' 11Il"d.ra-boa1.1 (;I\.itl')

Ildml'1) ;app:ar 10 ~'U. btnllll;~~"" (.'ll; tin~.U:tU'\ft"a.I pt:HNft hJd\ I'~' lnt'r""",nUUtUulbMUrt~IM""') t"'f1lHCS h ....n U. ",_'0 11(..,n..t.n...hh. lr'pU~I'"on .I.JIrd.

\-"',lltudili lAJUC UJ .. 'l'wlt=l\' ur tblVO' .ud SlLa.. 'I Ix

1Ir3" ,dll"" . I,hl, d. u..a.t " •., tl.lftt a."fl­mllnl~ • ')1",-' loo,..tnnt O\.d Ihouqh HlilM::

(:unll~ I n.H.'ubl IlUdd 1)(" llJUl dwUClU wluclal,('il ,inn,1 .u .II Jiw'-t, . liII,. r.t It'} r.tnh·.clnrnll.a.t Ihldrh uf a timlbr tDL. L.htht inlimlr\.LIN C'Cf, CW"llt'll ooh' a \'CI) IIwll. 1 lid IWcld urf ..riM ''''''I1I1("r. -..a tlJlI4t t.. Ibt 'IU"""''' d.1m.un 11I'~1\ Gllllln fpat1inlbTh In:bfn. 'WoolJt.D.U!rIt' ul tllal br ""ae uwLlto InC' .&J'lduuK&an=t'1. M:i4D) tJ~ "'.ouk! 'K" h.s\, atrirtl "'1bldrt ,u .all in Lro'. timr). 11\ th~ time: Ih1 _,Nap­T«nr. . WO'"» ..", lIIftl t. !la. wuoQlt'Jtncd'1birl. CJ il,..katl~ "~11 l.il..4.d,.Lk, "",,.,,. ..1 .inUK" 'UDom:" bnch mfantn .and lCn":a.trT. '1Uit ..,rtil"

8

Page 9: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Othcr sccondary weapons includcd axe andmace. Exccpt for the t<:,ikourioll (thc old Romanstcuris), a common infantry sidc-ann, axes wercuncommon amongst native Byzallline soldiers,though they were the principal weapon of Va ran­gian guardsmen. TIle mace (mtJt<:,oukion or bar­doukion, the laner meaning 'sledgehammer') wasused mainly by cavalrymen, though the Pratctplalists it amongst infantry equipment. Horsemen keptit in a leather case anached to the saddle.

Missile weapons consisted chiefly of bow andsling. The Byzantine bow was a composite weapon45 to 48in long with short, powerful limbs,probably originally adopted from the HUllS. It wasused by both cavalry and infantry, though its useamong the former was on the decline during thisera, the m:yority of Byzantine horse-archers beingprovided by Asiatic mercenaries. In fact archery ingeneral bad been on a steady decline in Byzantinearmies since the 8th century, so much so that in hisTaclica Leo VI had cause to complain that 'archeryhas wholly been neglccted and has fallen into disuseamongst the Romans'. Ht profK>Sc<l the rein­troduction of enforced archery practice (even formen exempt from military service) and ordered hisstrategoi to ensure that every household possesseda bow and forty arrows, but apparelllly to linleeffect.

The reorganization ofheavy ca\'alry units SO thattwo men out of every fj\"e were archers and nolonger carried lances at all was probably one ofLoo's refonns too, aimed at ensuring that bowswere kept out of the hands of those incapable ofputting them to good use (though he was basicallyfollowing the carlier Slralegium, where Mauricerecommended of kalaphracloi that though double­armed the best archers should fight mainly with thebow and the best lancers with the kontos). The factthat many archers were also issued with a sling(sptndone) may be further indication of Byzantineinefficiency with the bow. Spendobolon was anothertenn somctimes used to describe the sling, thoughthis was tcchnically the name used for the four tofour-and-a-half-foot staff-sling. Bow, slingand staff­sling, together with javelins and a bolt-throwingweapon called a solmarion, were all used mainly bylight infantry.

The solenarion occurs in sevcral sources andappears to have been a type of crossbow; It IS

10

usually mentioned in conjunclion with a quiverand short arrows called mel/ai. TIle S;;'lIoge Tacti.corum describes the solenarion as very eflective inbattlc because its arrows were fired at such velocityLhal no armour was adequate against them and thenaked eye was unable 10 sce Ihem, in addition towhich Lhey could fire to a very great range. LikeVegetius' arcuballisla SC\'en centuries earlier, IheS;;'llogt adds that the solenarion \\'as a "'eapon usedby light troops. For reasons unknown it seems to

have dropped out of usc around the middle of theloth century orsc:xm after, to be reintroduced eitherat the end of the II th century or the beginning ofthe 12Lh through contact with the ~ormans.

One fin~l weapon which needs to be mentionedis the rhomp!wia, with which many Byzantineguardsmen were apparently armed. It is notaltogclher clear exactly what this was and therchas been prolonged but inconclusive debate as to ilSidcmity. It was clearly ofa dist..inctive shape, butthe military manuals, thorough though they are, donOI cven mention it, let alone offer any kind ofdescription. The most convincing thoory, however,and the only one which seems to fit lhc little writtenand archaeological evidence that is available, isthat it was afalx-like weapon with a slightly curvedblade ofabout the same length as its handle.

Tile IagmataThe nucleus of the Byzantine anny during this erawas provided by the regular guard units based inConstantinople. Of these the Scholae, E."cubit..i,Arithmos and Ikanatoi cavalry regiments col­leClively comprised the Tagmata, though in alooser sense this term was also sometimes used toinclude thc Numeri, Hetaereia and other unitsstaLioned in thc capital, and could evcn include thcImperial Fleet (or at least Lhose elements of it thatwere secondcd to the Emperor's personal service).

Of the four principal regiments the Scholae wasthe most senior and probably the oldest; scven$Cholae are recorded in the early 5th century NOliriaDignilatum, and Procopius tells us that the scholae ofhis day totalled 3,500 men. By thegth century theirnumbers may have risen to 4,OOO-such at least isthe infonnaLion which Kodama, a loth centuryArab author, gleaned from an earlier work by al-

Page 10: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

......" -:r.~ftnftll ~ft:II kOllltarioo-armm cavalry, thislime from e heavily illustrated lIUlDuKripc ofScyLian, DOWU. Madrid. n .. ill"min_ticms probably post.-dallt thi. "'n1 butappear to baye bHD bredy based 00 1.tb-eeDturyo~•.PoiDU 10 DOte h...... are the qui~ of Lhe two bodi" in thero~ and !.he ....dooHype .l2Ddanb with th~r 10DfI:,~u-eame ...like tails. Annour is maimy scale (or mail?).(Bibliol_ Naciolllal, Madrid)

L._'_I I

Page 11: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Gill bron~e plaque or the 11th century depicting a m;!itarysaini (Theodore) in the gWlle or a dillmounted Byzantinecavalryman. His equipment comprilleli leathe... rringed lamel­lar k1ibanion, sword suspended rrom a baldric, lance (Ilhor_lened 10 iii the plaque) and decorated circular shield. The sashlied round hi" chest OCCurli in a number or sources and iiiprobably an indication or rank. MOIlI dismounted hOrliemenwear cloak" in the pictorial sources. (By courte"y or theTrullleeli or the Brilish MUllewn}

Carmi which datcd to 838-845. On the other handanother Arab, Ibn Khordadbah, apparently alsodrawing on al-Canni, seems to imply either anolJerall Tagmata strength of6,000 soldiers and 6,000servants (i.e. one servant per man), which wouldmean approximately 1,500 men per regiment if weassume all four regiments to have been of similarsize; or that each of the Tagmata regimcnts was6,000 strong. Certainly the great J. B. Bury seemedto favour the latter interpretation, for he acceptcdthe 6,000 as applying only 10 lhe Scholae. \Vhat IbnKhordadbah actually says is that 'the Emperor'seam p, in his rcsidcnce or ill the field, consists of fourdivisions of cavalry commanded by a patrician,

12

under whom arc 6,000 soldiers and 6,000 servants'.That there may have been 1,5°0 Scholarii, how­ever, is indirectly supported by another souree, thelate loth eentury Anonymolls VlIri, which says thatthe regiment consisted of thiny banda eaeh underits own komes. Admittedly, at Tactica-strength of300 men pCI' bandon this would give 9,000 mcn, butif we were 10 accept Nikephoros II's fifly-manbandon we would arrive at a much more accept­able total of 1,500.

Each of the Scholae banda may in fact becomparable to the original scbolae since these toowere once commanded by kometes; certainly an8th-gth century scat is preserved which refers to the'komes of the Fifth Schola'. It should be noted,incidentally, that the term sc!w!arioi sometimesoccurs in the sources as a general description forsoldiers ofall four Tagmata regiments, and perhapsthe Al/onymolls Vari's thirty Scholae banda should besimilady treated.

The Scholae were commanded by a domestikos, aswere the other Tagmata regiments (with theexception of the Arithmos or Vigla, commandcd bya dhoungarius) with a topo/ere/es as his seconcl-in­command. By :\Ilichacl Ill's reign (842-867) thcDomestic of the Scholae had become senior to allbut the strategos or 'general' orthe Clite AnatolikonThcme, becoming the most senior army officer, andcommander-in-chief in the Empcror's absence,during tbe course or the 10th century. The futureEmperor Alexius I bimselfheld this office in 1078,but the Scholae he commanded was no longer thesame regiment, as is witncssed by a reference in hisdaughter Anna Comnena's Alexiad 10 a Frankishmercenal)' in its ranks.

Nexl in seniority to the Scholae carne theExcubiti, probably established by Leo I (457-474)and at first commanded by a komcs, replaced by adomestikos in the 8th century_ The latter was seniorto the strategoi of the Western Themes at least by89g. rl~he regiment's strength is not recorded, but areference dating to 773 supplies us with lheinformation thai there werc alleast eighteen bandaand possibly morc, bUl lhe number of men in eachor these is nOI specified and it seems unlikely thatthey totnllcd the 300 men specified in the Strategicolland Tactica. In addition to the Excubitores, theDomestic of the Excubiti seems to have bcenresponsible ror the men callcd skribolles who were

Page 12: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

attacbcd to themc units as medical orderlies.Thc tbird Tagmata regiment was known by two

names, Arithmos and Vigla, of which the lattcr isthe more common in the sources. They seem tohave existed at leaSI as early as 559 and may be evenolder, Bury putting forward the hYPolhesis thattheir origin may be connected with the ComitesArcadiaci established by Arcadius (395-408).Unlike the other regiments, this onc was com­manded by a dhoungarius, though its banda werestill commanded by kometes_ On campaign theVigla performed special duties, guarding theEmperor's tent at night as well as conveying hisorders. They were apparently also responsible forprisoners-of-war.

The Ikanatoi was youngest by far of theTagmata regiments, having been established in the9th cent ury by Nikephoros I (802 8 [ I). Imerest­ingly Kodama, still drawing on ai-Carmi, docs notlisl this tagma at all, inslead subslituting Fid­aratiyin, which is clearly the same as Foederati,which had been an C1ite unit at Ihe time Ihal theStra/tgicon was written and certainly still exisledunder thai name in Ihe early 9th celllury andpossibly as late as Sgg, though its name seems lalerto ha\-e been changed to Hetaereia.

As already nOled, Kodama claims that eachregimelll of the Tagmata comprised 4.000 men andIbn Khordadbah implies either approximately1,500 or 6,000 men per regiment. The only otherapp..1.relll reference 10 their strength occurs in theAnonymolLS Var; of c. gSa, which says thai oncampaign the Emperor should be accompanied bya minimum of8,200 horsemen, which implies Ihatnormally there wOlild be more-the figures of10,000 12,000 have been suggested. Of the 8,200,1,000 were I-Ietaercia guardsmen, leaving a min­imum of 7,200 cavalry; Bury assumes these Lainclude thematic soldicl'S, but it seems moreprobable that the Anon)'11/ous is referring only to

Ihe Emperor's own regiments (undoubtedly in­cluding mercenaries) accompanying him on cam­paign from Constantinople. Even then, probablynot all of the Tagmata are intended; undoubtedlysome units remained behind to guard the C<1.pital,and we know that detachments of each regimentwere also posted in ~Iacedonia, Thrace andOpsikion. Such provincial detachments wereusually commanded by the Domestics' lieutenants,

Saines Geo",e Uld Oemecri", dernOllJitratUl! cavalry equip-<neae of c. 1100. Th., _ .....1 appear 10 be of mail aad quilerupecti,·.,ly. 80ch han plrnlfl:K. Th., 10ft!: twc.ia may beevid....ce orF...........d1 i.rt.Bu....ce.

the topoteretai, though c. 975 we hear ofprovincialTagmatic units, by now posted all over the Empire,commanded by dukes. Pro\·incial detachmentsparticipating in the 9.J.9 expedition againsl Cretecomprised 493 Scholarii, 869 Viglae, 700 Excu­bitores and 456 Ikanatoi.

Another regiment which Kodama gives astrength of 4,000 men was the Numeri, an infant!)'unit permanently stationed in Constantinople.Probably the figure is relatively accurate since it isimpossible to believe that the city could begarrisoned by any less, and even 4.,000 docs notseem sufficient for the purpose. The unit is only firstrecorded with cenainty by Kodama, but theevidencc indicates that it was much older. As in thecavalry regiments, its commander was a domes­tikos, assisted by a topoteretes, but Ihe individualbanda appear to havc been commanded nOI bykometes but by lribouni or tribunes, a title whichleslifies to the age of the unit.

13

Page 13: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Probably the responsibilities or the Domestic orthe Numeri did not go beyond the TheodosianWall, since a separatcofficer called the Count orthe\'Valls seems 10 have been militarily responsible rorthe Long Wall or Anastasius, presumably with hisown unit.

The Hetaereia Basilike, another unit associatedwith the Tagmata, has already been mentioned. Itfirst appears soon aner the mid·9th cemury andwas probably no more than the old Foederali undera new name, in which case-ir Kodama is to bebelieved-it numbered an improbable 4,000; theAl/ol/)'IIlQus' 1,000 seems a more realistic figure. Theunit's name derives rrom the Greek word rorcomrade (hetaeros) and should perhaps be trans·lated in rull as 'The Emperor's comrades·in-arms'or 'The Emperor's retinue'. It was a largelymercenary guard regiment probably comprisingboth cavalry and inrantry, its members, thoughmainly roreigners, also including native Greeks. Adetachment or the Hetaereia seems to haveaccompanied the Emperor at all times whenever helert the city, eilher on campaign or on huming trips.

This regiment at first comprised three individualbodies, the Great Hetaereia, Middle Hetaereia andLittle Hetaereia, commanded by their respectiveHetaereiarchs, but the Lillie Hetaereia was abol­ished during or immediately ancr the reign or BasilI (867-886). Commanders orthe sub-units oreachHetaereia seem likewise to have been titledHetaereiarchs, distinguished rrom the seniorofficers by having no prefix (i.e. Creal, Middle orLillIe). As with the Tagmatic regiments some of theHetaereia were apparemly posted in the provinces.

In addition there were other mercenary unitscovered by the gencral description or HClaereia, inthe same way that the Hetaereia itselr was, ingeneral terms, considered part or the Tagmata.These consisted orthe Khazars and Pharganoi, therormer recruited rrom a people or Turkish orHunnic origin sellied in the Caucasus, the latterrrom amongst Central Asian Turks living in thevicinity or Ferghana (hence their name). IbnRusta, who wrote c.903, records that 10,000 (!)Khazars and 'Turks' (undoubtedly the Pharganoi)accompanied Ihe Emperor on parade, and like theHctaereia some seem to have accompanied theEmperor on campaign and in the hunt. Anotherunit, the Maghlavitae, also appear to have been

14

associated with the Hetaereia; these may have becnWestern Moslem (i.e. Maghribi) mercenaries. Allor these were cavalry.

Enrolment as a guardsman in anyone or theseHetaereia units was clearly a much-sought-artel'privilege, and membership was in ract purchased.It is on record that entry into the Greal Hetaereiacost a minimum orsixteen pounds orgold; into theMiddle Hetaereia, ten pounds; and into theKhazars and Pharganoi, seven pounds.

The lftralzgian GuardOne other elite unit has yet to be discussed-theramous Varangian Guard, the only Byzantineregiment that most people know by name. Theirrascination derives chiefly rrom the incongruity orfinding such men-a warband or lusty, hard­fighting, hard-drinking barbarians ('wine-bags',some sources call them!) from the far North orEurope-serving the great Empcror or Byzantium,the Christ Incarnate, amidst the pomp andsplendour or the dazzling courtS and sparklingpalaces of the Holy City or Constantinople.Anecdotcs and stOries of their exploits abound inthe Icelandic sagas or the I 2tb-13th centuries, butthe glamorous reputation that tbey have somehowattracted over tbe last thousand years does notappear to be entirely deserved; one modernauthority goes so rar as to dcscribe them as 'prizedror their ability to act as thugs and desperadoes'!

Their nal)1e 'Varangians', used by the Russiansand, through them, the Byzantincs to describeScandinavians, probably derived rrom the OldNorse word var, meaning 'pledge', used to describea band or men swearing loyalty to one another,observing a common code or conduct, and sharingout profits rairly amongst themselves; all in all agood description of Viking activities in their rolesboth as merchant-traders and mercenaries.

Scandinavians (Swedes rrom Russia, in ract) firstvisited Constantinople in about 838. Some time

Military nints (ThNdore now joins George and Deonetrius)equipped as horseonen of the early 'llth century. ThNdore, onme left, wears a ktibanion over his l11ail corselet; George, inthe <:entre, wears a kJibanion with pteruges at shoulder andwaist; and Detnetrius has now changed into a knee-length mailloriltion with elbow_length lileeves. Alt thl"fl: have kite-shields.

Page 14: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118
Page 15: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

This is taken from a BY'lantine bowl of me I'lmcentury, probably depicting the folk hero DiogenesAkrit'ls. It is the only contemporary, or near_oontempo..... ry, picture which appears 10,> exist of thesboulder tufts mentioned in Leo Vl'll Tactica. ItseemS fairly certain that they would not nonnallystand upright like mis, probably a ",sult of artiSliclicence.

latcr, in 860, following an abortive altack on thecity by othcr Scandinavians scttled in Russia, atreaty stipulated for the first time that the Emperorshould receive a military levy from these 'Rus', asthey were callcd. Though this treaty may ncverhave been honoured, similar treaties of 91 I, 945and 971 probably were; 700 Rus wok part in aByzantine attack on Crele in 91 I, 629 in anotherattack of 949, Rus auxiliaries fought against thcArabs at the Battle ofHadath in 955, and two Rusships anended a Byzantine fleet sCnt against Ilalyin 96B.

None of these bands, however, constitutedregular units in permanent Byzantine employ, thefirst instance ofthe laner apparently occurring onlyin 988, when Emperor Basil II received as many as6,000 men from Prince Vladimir of Kicv. These'cxcellent fighters' were immediately established asthe Emperor's bodyguard, largely because, asPscllus puts it, Basil 'knew the treacherous dispo­sition of tile Romans', whom he could not trust; thein ference tha t far greater trust could be placed in theloyalty of the Varangians is bornc out by Anna

16

Comnena, who wrote that 'they regard loyalty tothe Emperors and the protection of their persons asa family tradition, a kind of sacred trust andinheritance handed down from generation togeneration; this allegiance they preservc inviolatcand will ncver brook the slightcst hint of betrayal'.

Not that they were always all sweetness andlight, and nor was their loyalty always abovesuspicion .... While an officer inthc Guard in 1042,the future Norwegian king Harald Hardraada wasaccused ofhaving misapproprialed Imperiallaxcs,and he is claimed in the sagas 10 have himselfblinded Emperor Michael V and kidnapped theEmpress Zoe's niece Maria. Michael VII, too, wasattacked by Varangian guardsmen, and in 1079 aband ofdrunken Varangians on dUly in the palaceattacked yet another Emperor, Nikcphoros IIIBotaniales, and tried to kill him. Much later, in1204 when the army of the Fourth Crusade wasbatlering at the watls of Constantinople, we evenfind the Varangians only agrceing to fight for a newEmperor on condition that he paid them at anexorbitant rate, 'making the vcry acutcncss of thedangcr an opportunity for driving a hard bargain'as a contemporary succinctly put it.

NOI that thcir usual salary was exactly poor. The12°4 cpisode also supplies us with the informationthat the Varangians received high pay, welt abovethat ofother mercenary troops. They seem to havereceived as much as ten to fifteen nomismata permonth (one and two-thirds 10 two and a halfpounds of gold per annum), as well as specialgratuities, a large share of Ihl:'; booty taken oncampaign, and possibly the right to panicipatc in asort of ritual plundering of the Emperor's privatcchambcrs on his accession if the sagas arc to bcbelieved. Harald Hardraada amassed a vasttreasurc in this way, so great a hoard 'that no one innorthern Europe had cver seen the like of it in oncman's possession before'.

Towards the end of the 11th century thecomposition of the Guard began to change. Thcpoint is still much debatcd, but it would appcarthat in the first few decadcs following the NormanConquest of England in 1066 an unknown numberof Anglo-Saxon and Anglo-Danish emigres bcganto lake service under the Byzanline Emperor.Cecaumenus' Sirategicoll of c. 1075-1°78 appears to

contain the first contemporary reference to them,

Page 16: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

and they are occasionally referred to in documentsand events of the 1080s, notably in a description ofthe Ballie ofDurazzo in 1081, where lhe inference isthat these Englishmen formed part ofthe VarangianGuard. However, though there certainly wereEnglishmen in the Guard from c. 1081, they at firstremained a minority element and many sources ofthe 12th century continue to speak of the Guard asbeing comprised of 'Danes' and 'Northmen'. Axe­armed Danes are recorded accompanying Alexius Iin Anatolia in IOg8, for instance, and only a fewyears later Saxo Grammaticus describes how 'menof the Danish (i.e. Old Norse) tongue occupy thefirst place'.

The Byzantine sources refer to the Varangians as'axe-bearing barbarians' or 'the axe-bearingGuard', sometimes describing them as 'those whohang their swords from lheir right shoulders' (acurious description referring to the use of rhom­phaiai) j the term Tauro-Scythians is also some·times used. The commander of the Varangianswas similarly sometimes called 'the leader of theaxe-bearing Guard', but his proper title wasAkolouthos or the Acolyte, 'Follower', undoubtedlya reference to his constant proximity to theEmperor. Both he and other Varangian officerswere usually of Scandinavian (or English) originlike their men, accompanied by Byzantine in­terpreters so that there was no language problem(though it is apparent from various anecdotes thatmany Guardsmen themselves learnt Greek). Westill have the names of several of the Varangians'commanders, such as Ragnvald, a Swede of theII th century who, on his memorial stone, isdescribed as 'leader of the war-troop in the land ofthe Greeks'; Harald Hardraada, who, though notthe Acolyte, held a senior rank in the Guardc. 1035-1044 and command~dabout 500 men; andNampites, Acolyte in the 1080s, probably aScandinavian nickname meaning 'Biter ofCorpses'or 'Bird of Prey'.

Tile Theme SystemThemes, or themala, provincial army-corps districts,first began to appear at some time during the 7thcentury, initially only in Asia Minor. ThoughEmpcror Heradius is usually credited with the

introduction of this system it seems far moreprobable that Constans II and Constantine IVwere responsible, establishing the first themes inAnatolia to defend the Empire's eastern fromieragainst Arab incursions.

By the loth century most themes were com­manded by military governors called strategoi, orgenerals (though the Opsikion Theme had a komesand the Optimaton a domestikos), each of whomhad his own full-time, personal mililary retinue ofspatharioi organized in units of 100 men com­manded by officers called kt1/larchai spolhariorum j

the size of this retinue varied from theme to theme,but we know from one source that the retinue of tilestrategos of the Thrakesion Theme comprised sixkmloTchiai (i.e. two banda). In addition thestralegos was assisted by three civic officials-theprolonolaTius, responsible for financial adminis­tration (induding the soldiers' pay), the praetor forlaw and administration, and the charlldarius fortaxation and revenue.

The forces at the disposal of the strategoi weresoldier-farmers, freemen each holding a plot of

H)th century Arab horsem.an. The Arabs had been theEm.pire'. m.ain enem.y since the 7th century, and Emperor LeoVI said thai il wa$ the ever-present Moslem thr_1 thatconvinced him. of the need to write his m.ililary maaual.

17

Page 17: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

agricultural land within lhe theme in part­payment for mounted mililary service performedby one of the estale-holder's family (or by proxy)as, when and where required. These smallholdingswere hereditable by the holder's eldest son so longas be also took on his father's military obligations.Many thematic soldiers bad servants, and thericher ones could even afford to own slaves, and itwas these richer or greater landowners whoprovided the 'firsl-class' thematic cavalry men­tioned by Leo VI in his Tactica, lheir higher incomeguaranteeing the superior equipment lhal thisstatus would require. Others were comparativelypoor, and in the early 91h century legislation hadeven been introduced obliging poorer thematic

Key to map: I Kalabria, ~ Langobardia, 3 Dahnatia, 4 Si....mium, .5 Dyrrachi...... 6 Bulgaria, 7 Nlkopoli.., 8 Kephalonia,9 Peloponnesos, .0 HeUa5, .1 Thessalonika, .2 Strymon, '3Macedonia, 14 Paristrion, 15 Thrace, .6 Ahydo.., '7 Chi05, .8Aegean Peiagb05, '9 Krete, 20 San>Os, 21 KihyrThaiots, 22Thrakesion, 23 Opsikion, 24 Optimaton, 25 BuIleUarioD, 26Paphlagonia, ~7 AnatOlikOD, 28 Seleulleia, ag Kypros, 30Kappadocia, 3' Kilikia, :JlI OIarsianon, 33 Armeniakon, 34Sebasteia, 3~ Lykandos, 36 Antiod>eia, 37 Teluch, 38 PoleisParephratidiai (Euphrates Ciriefl), 39 Melitene, 40 Koloneia, 'I'Me50potamia, 42 Taron, 43 Iberia, 44 Chaldia, 45 Theo­do&iopoli& (Talk'), 46 Vallpurakan, 47 Cherson (Gothia). Inaddition the Serbs and Croats, a550rted. ArmeniaD aDd Iberianprincipalitie5, the Lombard principaJitiu of Salerno, Capuaand Benevento, and the Atnirate of Aleppo aU paid tribute,while Venice, Napl", Atnalfi and Gaeta were nill nominallyByzantine towns.

soldiers to club togelher so as to adequalely equipone of their number to serve (a thematic soldier'sarms, armour and horse belonging not 10 theindividual himself but to the estate for which heserved). Nikephoros II seems lO have introducedadditional legislation so that wealthier soldiersmighl now be expected to help equip their poorerneighbours. Even then equipment was sometimessllb-standard, and uniformity was probably gen­erally lacking; at besl only shields, lance-pennonsand crests arc likely to have been uniform wilhin athemalic unit, though it is possible that some mayhave worn llnironn-coloured clothing like thecentral army corps in Constantinople. Menfmancially unable to rulfil their obligalion tosupply an armed horseman, even with the assis­tance of others, were evicted from their land andtransferred to the irregulars, called literally 'cattle­lifters' (which neatly describes their principal formof sustenance), responsible for supplying garrisonsfor lhe Empire's many fortresses. Deserters sufferedthe same fate.

New themes were created either by lhe sub­division of old ones or by the expansion of thefrontier, in which case units would be detachedfrom other themes to form the military nucleus for

THEMES AT BASIL 11'0 DEATH, 1025•,

18

,~, .- ---

ABBASIDS

Page 18: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Magyar horlleman ""th captive. At the time that Leo VIactually wrote hill Tactica the Magyars were the En>pire'.orner principal foe, though they could also be found allmercenariell in Ihe Byzanline anny. The annour worn berecould almOllI be 1I1....;gbl from an Imperial arsenal, ""th mailcoif and corselet, iron helmet and IIplint.annour greaves andvaJnb ces. Hill captive wears lamellar. (KunsthistorischcnMuse , Vienna)

the new province (for instance when Leo V Icreated the Charsianon Theme he transferred to itthree banda from Bukellarion, two banda fromArmeniakon, and a whole turma plus olle bandonfi'om Kappadocia). Some themes grew from otherfonified fromier districts called kleisourai (the namemeans 'moumain passes') where noble familiescalled akritai or 'borderers' maintained their ownforts and retinues in almost feudal fashion. Youngersons of thematic landholders often lived on frontiercstates too, supplememing the manpower of themeor kleisoura.

The troops of each theme were organized intotwo or three turmai or mcrai which, as explainedearlier, were in turn subdivided into moirai ordhoungoi, each comprising a number ofbanda. IbnKhordadbah, who wrote c.845, describes theorganization of one of the larger themes ([0,000men) as two turmai each comprising five 'banda'(presumably moirai arc meant) of I ,000 men, eachof these 'banda' being subdivided into five units of200 men called pt1llarchiai and commanded by

kOllletes (therefore probably banda). Each pelllar­chia, Ibn Khordadbah tells us, consisted of fiveunits of forty men commanded by 'kontarhin'(probably meaning kentarchai or pelltekonwrchai,the units therefore presumably understrengthallaghia or hekatontarchia).

It seems apparent, in fact, that the strength ofthematic units must have varied considerablyaccording to the size and manpower of eachindividual theme, and the variation in totalnumbers of troops available from theme to theme isapparent in the lists of the Eastern Themesrecorded by Ibn al·Fakih al-Hamadhani (whowrotc c.902) and Kodama (c. 930), which aregiven below:

Ibn al-Fakih Kodama

Anatolikon '5,000 15,000Thrakesion 10,000 6,000Chaldia 10,000 4,000Al'meniakon 9,000 9,000Bukellarion 8,000 8,000Opsikion 6,000 6,000Paphlagonia 5,000 5,000Seleukeia 5,000Macedonia 5,000Thracc 5,000Kappadocia 4,000 4,000Charsianon 4,000 4,000Optima ton 4,000 4,000

85,000 7°,000

These lists together provide a relativcly complcteinventory of the Eastern Themes as thcy stoodc. goo cxcept that they omit Koloneia (probablyabout 6,000 men), Mesopotamia (probably about4,000), the kleisourai of Sebasteia, Lykandos andLeomokomis (no more lhan a couple of thousandcach), and mislakenly include lhe non-militaryOptimaton Theme, which had perhaps beendemoted to its reduced status (as an army servicecorps) as a result ofsome earlier rebellion (it is lastrccorded as a combatant corps in 773). Whenlisting the Eastern provinces the Arab sources alsomention an additional theme called Talaya,Talaka or Tafla in lbe vicinity of Constantinoplcitself. This is Olherwise unknown and its name ispossibly no more than an Arabic corruption of

19

Page 19: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Tagmata; as a region it may represcnt thecommand of the Count of the Walls (scc carlier),the Long Wall of Anastasius marking itS westernfrontier. Unfortunately the lists do nOt give us itsstrcngth.

Xor do the sourccs givc us the strengths of the\\'cstern or three i\aval Themes, which Leo VI'ssalary schedule lists as comprising KibyrrhaiotS,Aegean Pelaghos and Samas (naval) and Pelopon­ncsos, Nikopolis, Hellas, Sicily, L'lIlgobardia,St!)'mon, Kcphalonia, 111cssalonika, DyITachium,Dalmatia and Cherson. However, the Westernnlelllcs were always regarded as inferior 10 thestralegically more important Eastern Themes andIheir strengths would have been correspondinglylower.

Faidy certainly the figures given by Ibn al-Fakihand Kodama only represent the cavalry, whichwere the backbone of every Byzantine army. Leo'sA nwnber of loth-Illth-eentury ivory callkets dep;CI figurell&uch liS tneH. They fairly certainly rqorewent A&iat;c merCe­naries, probably Pauio"k&. The apparent 'trouHred' tunic illprobably II long coat split at front and bade for riding. Note thepoiated felt or fur cap, c:haraCleriStiC of Pat~nak dr-.......(ViCloria and Albert MusetUn, London)

Tae!iea tells us thai each theme could provide at themOSI 4,000 first-e1ass cavalry, who he describes asktl!llphrae!Q; and \Vere therefore presumably moun­ted on armoured or half-annoured horses, so thebalance were presumably less wdl-equipped and oflower quality. Certainly it seems to have beenpreferable 10 summon first-class soldiers from aneighbouring theme in an emergency rather thanto rely on one's own second-class troops. Leo alsoSlates that withoul drawing more than 4,000 menfrom each the Eastern Themes could pUI 30,000horsemen in the field, which \\'ould still leave ahealthy reserve 10 fall back on.

In addition there were also lhematic infantry,about whom litlle information is available. It ispossible that lhey did nOt receive land grantS likethe caval!)' but were instead recruited by con­scription, lhough some must have been employedon a more permanent basis. The Ttlelie(/ seems toimply tbat a single theme eOllld muster as many as24,000 infantry, who would have presumably beendivided into first and second classes like the cavalry.Many of these would have bccn lighl troops.

20

Page 20: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

I

,; iit,'1'J" ,,wI tU i.,;"bl,.~~,.j;; I~'~ k';"~I~~ X"ll)Il'p";"7'll'4JU<llj , 7IJi1~'~.k"';,, ""("I"';,MI.~.;;;,:, I 'I: ;'f"I' ' ,"-

~ ,...~ ... t 7'Yf,

As time wenl on the thcmesySlcm began to breakdown, since the stratcgoi and their senior officers,who had become big landowners within theirthemes, began to utilize their combined civil andmilitary authority \0 transform their lands intohereditary semi-independent possessions, thesoldier-fanners of their estates becoming a son ofprivate anny. Though lhey were aware ofwhat washappening the central administration could doliule to prevent this transformation. In fact theImperial government had no option but to granteven more rights and concessions to this newprovincial aristocracy in order to retain its militarysupport, even to the point of issuing chrysobullswhich exempted the estates of the greatest land­owners from paying any taxes at all. Inevitably,and despite assorted legislative measures, more andmore of the small thematic landowners wereslowly squeezed out and their smallholdingsabsorbed into the great estates, which effectivelyreduced the fighting strength of the theme; theHudud al-Alam, a Persian source written C.982,states that each of the Eastern Themes could raiseonly 3,000 to 6,000 men, while a Byzantine sourceof c. 970 rates the full cavalry strength of eachtheme at no more than 3,000.

It was as a result of tbis decline that EmperorNikephoros I I Phokas (963--96g), himself risen

Turks attack a By-u.ntine fortress, from the Scylililes manus­cript. On the lefl are unarmoul'ed bors_reber.., on the rigbtheavy cavalry in corselets and belmets. Tbe defenders areburling down javelins from the balliement~. (BibJiolecaNaeionaJ, Madrid)

from the ranks of the thematic landowning aristoc­racy, ran down and disbanded the native troops insome of the themes, the strategoi now frequentlyemploying mercenary units in their place. Thisincreasing reliance on mercenaries, compared tothe low quality to which some of the nativethematic soldiers had sunk, is apparent in thedescription ofa thematic muster of I 067, where thearmy is described as comprised mainly ofMacedonian (Slavs), Bulgars, Kappadocians (pro­bably Armenians), Uzes, Franks, Varangians and'othcr mcrcenarics who happened to be about'.The fcw nativc Byzantinc troops arc described as'bent over by poverty and distrcss and devoid ofarrnour. Instead of swords and othcr military arms... they carried hunting spears and scythes ... andthey were without horses and other equipment.Because no Emperor had taken the field for manyyears they were (considered) for this reasonunprofitable and unnecessary, and their salary andcustomary maintenance had bccn taken away.'Most alarming of all, this is allegedly a descriptionof the Anatolikon troops, elite of the EasternThemes.

21

Page 21: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

,

ANGUS MoBRIDE

-

,

• SkuCato"" "Ch-'OJlh " ..nlud..",

OJ SkuCatos, ".95°3 Pelt""'lo"'_ "'975.. Skutatoli, 'GCh «ncury

3

A

Page 22: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

,

o s"rv...c, ,och "och ~ncuri_

" Pack_mules, .och cencury3 Vn....moured ;nr....ryman., .,ch•• 2th cenCUr;e1i<4 P1i;IoIl, .och «njury

,

4

B ANGUS Ml;BAIDE

Page 23: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Kilu.phr;&CIOS, 'olh cenlur)'

"Cavalry siandard·hearer, ",h-'<I,h ..en,uri"s

3 Kilillphrllclo", ".105°

,

A'lGVS McBRIDE

3

c

Page 24: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Klibllnophoro.., ".910

D ANGUS McBIlIDf

Page 25: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

,

I Guard officer, <:.880~ ElTlpo"r .or In parade armour c3 Member of the 8asilikoi ,.'0'7

Anlhropo;, <:.860

ANGUS McBRIDE

3

E

Page 26: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

F

Ru~ n,ercenilry, e.9.5"2 Vilr ..ngiiln G ....rd~m..n, c"ooo

J V .. r ..ngh.n G ..ardsm..n in dress ..niform, C.IOJO

)

3

. ..

ANGUS McBRIDE

Page 27: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

• Trapez;los, loth "",nlury

\I Pala;nak m<lrc"nary, 11th century

....NGUS MoBRloe G

Page 28: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

• Stljuk ",trCtnllry••• Ih-.,..h ctnlurits2 &. 3 hlllo·Nor",lIn ",trctn..rits, h'le I1lh Ctnlury

H ANGUS McBRIDE

Page 29: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Mljuk c:bif,ftaiD in o::banlal'ri$Lio:: CI>Illwn~, c:oonpl~l~ wi!hcued bow at IdllUp. Betw_!.he Mlj.......' fi"t app~ara.u:eon!.h~ Eonpire's ~I~"" f.....Lier .. an obsa>re T..rkish tribe and!heir dKisi"e ddNt of Ihe Byu.otinM al Maazikll:rt in '07', aperiod of ollly fifty yean bad elapsed. (Me1.ropolila.a M ..saunof An. Ham. Bmbaae Dick fll.Od '9S.)

Ri.flot:Victorious Byzulline Emperor reo:eivins!.he .lIubrni.llsion ofaDenern.y city, from !he Troy" caaket. u«pI for !he crown !.heeqwprnenl wo.... ia id~nLica1 10 Lhal of !he ho~en nn !h~

fronl of!h~ ca5k~t,5hown On pa«~ 7. (Trkor de Ia Ca!hidra.l~,

Troyea)

NornuLD Imi«bls from a porch al !.he ch..rch of Sa.. Nicola alBari, c. 1087. The cily ofBari,!he Empire'slalll stronghold onIbe lIalian mainland, f",11 to Ihe Nonn..... under RobertGwscard in '07'. The fipres are yery aUnilar to Ibose in !heBaye"" TapetUry. NOIe both underann ....d oyerann u.... of!he11lD~. (TUn Benlon)

22

Thc principal rcason for this drastic dccline intbe quality of tbe thcme armies, other than thcterritorial ambitions of thc stratcgoi, was Ihestruggle for political power which had broken outin the course of thc loth ccntury bctween thegenerals in thc provinccs and thc burcaucrats inConstantinoplc. The gcnerals' ambitions werc heldin check relatively successfully up until Basil II'sdeath in 1025, but thereafter the struggle becamemore violcnt. Thc ncxt thirty years saw an averageof one major provincial rcvolt by the generals perannum, and in 1057 onc of thcir number, IsaacComncnus, actually succeeded in seizing thcImperial throne for two years.

Inevitably, whcn they gOt the chancc thcbureaucrats' reaction to these constant rcvolts wasto disband units, to convcrt the obligations ofothcrsintoadatratio a fonn ofscutageor taxation in lieu of

military scrvice), to dismiss and cxccute generals,and to Cut ofr Ihe pay and maintcnancc grants onwhich thematic soldiers livcd. Constantine IX, forinstancc, who rulcd 1042-1055, cntirely disbandcdthc army oflhe important fronticr themc of Ibcria,perhaps 5,000 men, and convcrted ilS obligationsfrom military scrvice to the payment of tax, and wcfreqllcntly rcad in the SOllrces of other thematicarmies 'in want of their pay and deprived of theprovisions which were uSlially sllpplied to them'.

It was such measures as thesc, together with thedrcadful squandcring ofmanpower which the civilwars involvcd, and the stcady elimination of

Page 30: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

military smallholdings by the provincial magnates,that led to the deterioration and demoralization ofthe Byzantine army in the years which precededManzikert, and nOI even the disastrous defeat thaithey suffered there could halt the process.

M ERCE:"ARI ES

Though Byzantine armies had nearly alwaysincluded considerable contingents of mercenariesin their ranks this particular perioo saw a rapidincrease in their numbers, as a result of Ihedeclining strength of the thematic annies. As thestruggle belween bureaucrats and generalsdraggedon, more and more mercenary troops came to beemployed in place ofunderstrength,low-quality ordisbanded thcmatic units. By Nikephoros II's reignmOSI if nOI all light cavalry consislcd of Asiatichorse*Mchcrs, and by the mid-l llh century morethan half Ihe men in most Byzantinc armics weremercenaries of diverse ethnic origins.

Most of these mercenary soldiers were suppliedby various Turkish peoples, amongst whom thePalzinaks took pride of place, but many olhernations were also represented. Frankish chroniclersof the First Crusade frequently refer to the manymercenary types in Alexius I's annies, particularlythe Patzinaks, Cumans, Uzes and Turks. Oneanonymous chronicler recorded that in 1096 therewere in Constantinople, in addition to Creeks,'Bulgan, Alans, Cumans, Patzinaks, Italians,Venetians, Romans, Dacians, Englishmen, Amal­fitans, and even Turks and many gentiles,Jews andproselytes, Cretans, Arabs and peoples of allnations'; and although some of these-such as theVenetians, Amalfitans and Jews-were merchantsor artisans most would undoubtedly have beenserving in thc army. Certainly contingents of mostof these pcoplcs, and more besides, were present inthe Byzanline army destroyed at J\llanziken (seebelow). Anna Comnena refers to mercenaries in herfather's lime as 'horsemen and footmen coming outof all lands' .

Many were supplied as self·contained unitsunder their own leaders and officers by theEmpire's satellite states and vassals. uch con­tingents were referred to as Jjmmochoi or allies, likethe old Roman Jjmmochiorii. The tenn was used ofPatzinaks, Serbs, Uzes and Russians during theInh century.

••.­.-,.

A ponrail ofc. 10'7 ofBni] II (g,6-10II5), called BI4«IOroc'Ofl08or 'Killer of Bulg.....'. A ....p<l'rb soldi....., eonlnnpo ri....~ned Ibal allb.. yery sighl ofms ba.ruler Ibe eoemy ed 10fI_. scrrami.ng. 'R_l R_! II ~ Ibe Emp<l'ror!' Ie wa. be wbolotally ....bjllpled B g.ria, in addiuon scoring Yittorie. OYft'

Ann~.. Geo.pa A...bs ...... No.......... Had b.. liyedaaolber t.... yran be would hay.. .-:onquued probably IbeEmpire'.IODg.t_lllalia........~ provmCf!S 100. tfu; own~ foUowed 0 .. from IbOH or two olber brillia.D1 gftlera.l..tbe Empero.... Nlkepboros D Phoka. (WbOM .I...penclousYietories oyer Ib.. A ...bs WOOl b.in>. from Ibeir own li~ theutle of 'The Whit.. Oealb'), aAd Joha.l TzirniS':~.. wIto bul forbU prem.at......... dealb in 976 would haOle 1'f!C01loCI....-ed «>0>­

pleld,. Ill.. Holy Laad, 1_, '0 Ibe Arabs •• 10011 ...... ':w- Notwithout rea.- .... Ib....... ""'-10115 beta cbn.lftled by_e•• Byzanuwn'. 'Age of to...q......,·. (Biblicueca Nazinnal..Marcia..... Venice)

23

Page 31: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Bearing in mind that even at their zenith in the9th and 10th centuries only the largest themescould raise as many as 4,000 first-class cavalry, suchmercenaries were employed in considerable num­bers. 15,000 Patzinaks were hired in 1049, therewere 3,000 Normans in Byzantine employ in theimmediate post-~'!anzikenera, and in 1078 there,,'ere in various armies 2,000 Seljuk Turks, 6,000Alans and 8,000 Normans, in addition to unknownnumbers of Patzinaks and Italians. In logl theremay have been as many as 40,000 Cumans fightingalongside the Byzantines at Levunium.

Inevitably there were drawbacks, however. Theloyalty of mercenaries was often questionable­particularly lhal of Frankish contingents, who hadan inflated opinion of their own value-and bribesand other expensive inducements sometimesproved necessary before mercenary troops wouldfighl at all. And if their pay was not forthcoming(which proved so often to be the case) mercenarieshad distressing habits like changing sides halfwaythrough a battle or campaign or looting friendly(Byzantine) territory!

Tile llJTible Day':Mal1zikert 1071

Armenia, once the Empire's principal recrUitingground but now left practically defenceless by thedisbandment of its thematic armies, was in thehands of lhe Seljuk Turks by 1067. This loss wasfollowed by almost continuous Turkish incursionsinto the Anatolian heartlands of the Empire, and itsoon became painfully obvious that its Easternfrontier could only be successfully defended by thereconquest of the lost Armenian territories. In 1071there came an opportunity to achieve this.

Early that year the Seljuk Sultan, Alp Arslan(the name means 'Mountain Lion'), had set out forSyria with the intemion of seizing Fatimid-heldDamascus prior to launching an attack on Egypt.En route he attacked several Byzantine·held towns,as did one of his lieutenants, Afsin, whose ownactivities included the capture of the fortresses ofManzikert and Argis during January. Afsin's

24

forces, however, were now snowed up in theTzamandus Ilass and stood no chance of movinguntil the spring lhaw set in.

Meanwhile, taking advantage of Alp Arslan'sextensive advance, Emperor Romanus IV Dio·genes had prepared an offensive in the Seljuk rear,assembling a huge army at Erzerum, some eightymiles from Manzikert, in spring 1071. The actualsize of the army is debatable; two sources claim300,000 men, while others claim 200,000, 400,000and 600,000, Matthew of Edessa even going so faras to say it was one million strong! Exaggerationsaside, however, all the sources agree that it was anabnormally large army. The 12th century Moslemchronicler Imad ad-Din has left us with delailsofitscosmopolitan composition, recording in addition toByzantines contingents of Russians, Khazars,Alans, Uzes, Cumans, Georgians, Annenians andFrank!; (who appear to have been chiefly Normansand Germans); Matthew of Edessa adds CrimeanGoths, Patzinaks and Bulgars. The nalive Byzan­tine elements had been gathered from most of theWestern and all of the Eastern Themes, though theonce·cJite Anatolikon troops were notable only bythe smallness of their numbers. In addition theEmperor \\'as accompanied by a contingent ofVarangian Guards and the cavalry ofthe Tagmata.

The largest pan of the army consisted ofengineers, labourers and servants, the formernecessary to operate the siege·engines whichconstituted a major part of the baggage-train of'thousands' of wagons. At the most, then, it seemsprobable that only some forty per cent of the 'hugearmy' actually consisted of combatants, of whomonly a fraction were Byzantine and very few ofthose regulars; the remainder were seemingly poorin quality, ill·trained, ill-equipped and ill­disciplined thematic troops, their low standardbeing a direct consequence of the running down ofthe military establishment by Romanus' anti·military predecessors. In addition some of themercenary elements were so unruly that there waseven a pitched battle with the German contingentfollowing looting incidents; it proved necessary tocall the rest of the army to arms before the Germanmutiny could be put down!

It was only after the beginning of May, whileencamped before Aleppo, that Alp Arslan receivedthe news of Romanus' advance on Armenia.

Page 32: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Immcdiately he abandoned his planned move onDamascus and withdrew towards Mosul, so rapidlythat his army was scallered far and wide, many ofhis untrustwonhy Iraqi auxiliaries taking theopportunity to desert. In addition the Seljuks lostlarge numbers of horses crossing the Euphrates. Infact, to a Byzantine observer in Syria, Alp Arslan'sretreat gave the impression ofa veritable rout, andit was probably this information which finallyconvinced Romanus that he should launch a full­scale attack rather than merely contenting himselfwith partially restoring the old Armenian frontierdefences.

Therefore in June or July, in preparation for hisgeneral advance into Vaspurakan, he divided hisarmy in two, sending out a large force of Frankishand Turkish auxiliaries under the Norman com­mander Roussel de Bailleul to lay wastc the regionround Manziken and Akhlat. Advised of theByzantine movements by refugees from this area,Alp Arslan now set out northwards from Mosul tointercept Romanus, sending ahead one of his mostdistinguished officers ('Soundaq the Turk' tht:sources call him) with about 5,000 men to reinforceAkhlal. The Sultan himself was accompanied atfirst only by his 4,000 personal mamluks, bisscattered army having failed to reassemble, and theseriousness of the situation did not permit him thetime necessary to return to the heart of his ownterritory in the East to gather fresh troops; insteadhe summoned them to join him on the march, andrecruited in addition some 10,000 local Kurdishtribesmen.

Romanus in his turn had probably got wind bynow of Alp Arslan's approach; he despatched abody or(allegedly) 20,000 men, apparently Cumanor possibly Russian heavy cavalry under a Geor­gian officer, Joseph Tarchaniotes, to the aid of theFranks and Turks now approaching Akhlat,enabling the latter force toset itselfup safely beforethe town. In the meantime the remainder of thearmy successfully retook Manzikert aner thebriefest of sieges.

I t was only then that the Byzantines first becameaware that part of the Scljuk relief force hadactually arrived in the area, for 011 the morning of16 August Soundaq encountered and defeated alarge foraging party. Romanus immediately des­patched one of his generals, Nikephoros Bryennius

(undoubtedly the same man who later became DuxofOyrrachium and rebelled against Michael VII)to deal with Soundaq, but he was repulsed and hadto be reinforced by a second detachment underBasilakes, strategos of the Theodosiopolis Theme.Weight of numbers now forced Soundaq towithdraw. Whether this withdrawal was a feignedflight or not will never be known; but the pursuingByzantines, caution thrown to the wind, werecaught in a sudden counter-attack in whichBasilakes himself was captured, together with hisstandard, and Bryennius was wounded. At thesame time Tarchaniotes and Roussel de Bailleul,also having suffered heavy losses in engagementswith Soundaq and receiving news that Alp Arslanhimself had now arrived tOO, pulled out of Akhlatand withdrew as far as Melitene.

By the time Romanus had marshalled the bulk ofhis army Soundaq's force, in true Seljuk style, wasnowhere to be found. So the army returned to camp(a contingent of 'loyal' citizens recruited inr."janziken taking the opportunity to desert), andan anxious night was passed under the eyes andarrowsofthe Seljuks who, joined by Alp Arslan andhis main army, now set up their own camp onlythree miles away.

Yct the Sultan's army, as Romanus' scouts mustnow have informed him, was considerably smallerthan that of the Byzantines. The lowest recordedestimate is 12,000, while Ibn al-Athir says 15,000,but the higher figures of 30,000-4°,000 or possiblymore seem more probable. But, at the same time,the Byzantine army was itself now considerablysmaller than it had been at the outset of thecampaign; the detachments of Roussel and Tar­chaniotes had not returned, there had beencasualties in the skirmishes with Soundaq-reallybattles in their own right-and to feed his massivearmy it had becn necessary for Romanus to sendout large numbers offoraging parties, even as far offas Georgia, to gather provisions. In fact, theByzantine army had apparently been reduced toonly 100,000, and of these many must have beennon-combatants.

Next morning an embassy was sent by Alp Arslanwith an alTer of peace, which was scornfullyrejected-since it would be financially impossibleto raise such an army again for a long time 10 come,Romanus had little choice but to force a decisive

25

Page 33: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

solUlion there and then. Besides, he had the utmostconfidcncc in the size, if not the quality, ofllis armyand in his own ability to achieve a signal victory.Further, he suspected that the Sultan had only nowrealized the still considerable numerical superiori tythe Byzalllines enjoyed and illlended the peaceoffer merely as a delaying tactic while reinforce­ments were summoned.

Admitlcdly a dclay might also have been on thesidc of the Byzantines-it would have givenTarchaniotes, Roussel de Bailleul and at least someof thc foraging detachments time to return; but atthe same time delay would also give the army,discouraged by the inauspicious handling of thecampaign to date and distrustful of the Armenianand Turkish contingents in its midst, the oppor­tunity to sink to an all-time low in morale and tobecome even more undisciplined lhan il alreadywas. In addition, more mercenaries might mutinyor desert. Logically, therefore, a delay which wouldprobably see the Byzantines stronger in numbersbut dangel'O.usly low in morale, and the Seljuksgreatly increased in both numbers and confidence,could not be contemplated. It is hardly surprising,then, that despite advice to the contrary from hismany generals, Romanus decided that he wouldcommit lhe army to ballle. The die was cast.

On 18 August both sides prepared for theforthcoming engagement, which the Scljuksplanned to take placc on the following day,Friday-the Moslem Sabbath. None of Romanus'largcr detachments returned during the day, in factquite the opposile-a number of Uz mcrccnariesunder a certain Tamis decided to change !'Ii de!>during 17 or 18 August, possibly because they wereTurks like the Seljuks, but more probably becausetheir pay was months in arrears. They slippedquietly away and entered the Seljuk camp in abody, where further Moslem rcinforcements fromAkhlat and Manzikert had also arrived. Tar­chaniolcs, meanwhile, apparently had no intentionof rClurning, and as it happened neither hadRousscl. The only action oflhe day saw the archersof the Byzantine army marching out of theirfortified camp and successfully driving off AlpArslan's skirmishers with heavy losses.

Details of lhe Battle of i'\'lanzikert itself areunclear and arc sometimes contradictory since ofall the sources only one (the Historia of Attaliates)

26

Figun:s of this type seent to start appearing in theloth Cnltury and are orten thought 10 be VarangianGua..dsnten. They only eve.. appea.. in biblicalcl"\lcifixion scenes, howe"e.., and wea.. a type orheadd..ess which in Byzantine a"l ;s norntallyassociated wiLbJews. But al the Sam.e tinte Lbeir richpanoply certainly suggests Lbat artists ntay ha"eused the eqwpntnll of guardsnten as their m.odel,and inlerestingly this 6gun: has a ra"en_like birdentblazoned on his shield, which could certainlyassociate hint with a Scandinavian origin.

was written by an eye-witness, but it is cenain thaton the morning of 19 August Romanus drew up hisarmy in lhe customary two lines. The first consistedof three divisions-the right of lhe lhematic troopsof Kappadocia, Anneniakon and Charsianon plusUz mercenaries, under Alyattes (strategos of theKappadocia Theme); the centre of Ihe centralEastern Themes and the Tagmata under Romanushimself; and the left of thc Wcstcrn Thcmes,together with Patzinaks and other auxiliaries,under Nikephoros Brycnnius. The second, orrcserve line was under the command ofAndronikosDukas, a nephew of the previous Emperor andtherefore no friend of Romanus-a factor whichwas going to prove decisive; it was compriscd ofGerman and Norman mercenary heavy cavalry,the majority of the archontes (noblemen and theircontingents) from the Eastern borderlands, and a

Page 34: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

large part of the Hetaercia. No troops at all werelert to guard the camp.

Their advance across the plain of Manzikerttowards the Seljuk camp, which could nOt havestarted until well arter midday, was virtuallyunopposed except for skirmishes on the extremeflanks, where the Byzantine commanders musthave kept a wary eye on their own Turkishauxiliaries following the desertion ofTamis' Uzes,particularly since the flanks lacked any security inthe open. But tbe bulk of the Seljuk army retiredsteadily before them in feigned f1igbt, drawing tbeByzantines on until in the late afternoon or earlyevening they came to the abandoned Seljuk campsite. At this point Romanus seems to have doubtedthe wisdom of advancing further, and fearing aSeljuk atlack on his own camp, undefended in bisrear, he gave the order to retire, turning theImperial standard towards the rear.

But the order was misunderSlOod. In the failinglight only the centre tumed as ordered while theflanks hesitated, apparently confused. Simul­taneouslya rumour swept through the reserve line,

treacherously put about by Dukas himself, that theEmperor had been killed. Watching from theheights nearby, the incredulous Seljuks saw in thechaos 011 the plain below them the opportunity theyhad been awaiting. At that moment, with theByzantines' first line disorganized and facing inevery direction with gaps between its centre andflanks, Alp Arslan led 10,000 (j'esh Seljuk cavalry tothe charge.

Almost immediately a rout ensued as theByzantines panicked, believing themselves bet­rayed by either the Armenians or the army'sTurkish auxiliaries; in faelthe Armenians were thefirst to flee the field and practically all got away,while by contrast the majority of the Uzes andPatzinaks remained loyal to the end. The rightwing of the Byzantine army soon disintegrated,

A rare picture of bearded Varangians in full annour, hereanending the execution of an Anti.Em.peror. Eighteen of theirfamous axes are in evidence, together with lipears and unitstandards. Note the m.ixture of drcular and kite-shields.(Biblioteca Nadonal, Madrid)

,

... \ .. II.........._.. _._ _"

27

Page 35: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

•,

••

"'"-,,•\

••

. . '-_ ..". ..., \'/0"' .. r. f••• ';, tl""'-:-"-~'""l~....To o:. __~., .,',ra I

• -••w,

Early roth_lury .kulato;" (rom the Joshua Roll. Note' ..hield.wall' "I." of the ....rge shi"ld... The 6!,,~. with cloakswrapped round their bodin are officers, one ofwhom cam" abaudo", lilandal'd. (Biblioll!Ca Aposlolica Vaticana, Rome)

weakened by the gradual descnion of the Kap·padocians, though me left, despite being cut offfrom the centre and attacked in flank and rear, onlybroke after a hard fight. Dukas, blithely disregard­ing his duties as commander of the desperatelyneeded reserve line-which might still have savedthe day-had already withdrawn from the fieldwithout commitling his meo. Bryennius, anotherByzamine historian, Slates flatly 'the rearguardwithdrew immediately'.

Only the centre stood, where Romanus foughton, wounded and with his horse shot from underhim, until be was recognized by his COStume and bythe Varangian Guards crowded round him andcaptured by a Scljuk slave-soldier-the first timethat a Byzantine Emperor had ever been takenprisoner by Moslems. The last few Byzantine unitsstill holding together gave way at this, and a closeand bloody pursuit of the scattered army continuedthroughout the night.

One Emperor had once said that 'the army is tothe state what the head is to the body'. That bodyhad now been decapitated, and the executionmarked the end of Byzantine military greatness.

TilePost-MollzikeJ1Period

~Ianziken saw the end of the traditional Byzantineanny. ~Iost of the Tagmatic regiments weredestroyed on the battlefield and those that surviveddisappeared from the scene soon after. Likewise theforces ofthe Eastern Themes, understrength as theywere, had been annihilated j and with them gone,much of Anatolia was soon overrun by thevictorious Turks, many of whom were shon­sightedly introduced as mercenaries by Romanus'immediate successors, Michael V II (107 I-I 078)and Nikephoros III (1078-1081).

The destruction of the Eastern thematic armiesin faCl left the Empire practically defenceless, andthough a new central army soon began to appear inthe capital, responsibility for the Empire's exposedEastern frontiers devolved largely OntO an ever­increasing number of mercenary regiments com­prised mainly of Normans, Turks, Cumans andPatzinaks. TIle widespread disappearance of the­matic troops and the parallel increase in theEmpire's reliance on mercenaries were withoutdoubt the principal results of Manzikert militarilyspeaking, setting a trend which was to continueunabated for the rest of the Empire's history.

28

Page 36: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Cash was short, however, and the ImperialTreasury exhausted, and before very long manysoldiers, both natives and mercenaries, were livingon, and being paid with, new land·grants calledpronoiai, in existence as early as the loth century butonly becoming regularly associated with militaryservice in the post-Manzikert period. Unlike theholder of a thematic land·grant the holder of apronoia, called a stm/ioles or pronoiarios, received inaddition tax·relief as well as extra paymelll in thefonn of cash andlor kind, so he was financiallyconsiderably better off than an old thematicsoldier-though pronoiai were not hereditable (atleast until the 13th cemury). Many Byzantinenoblemen were also now receiving pronoiai inexchange for military service as cavalrymen,perhaps accompanied by personal retinues ofpredctermined size. TIle one rcal disadvantage ofthe pl'onoia system, however, was lhal it cost thcEmpire a huge amount of moncy in lost taxationrevenue which, in turn, meant crrectively that itcould aRord to maintain fewer and fewer soldiers ona regular basis.

TIlesoe detail. from a locJ>...c:-hu'y ivory aoal<et depictAnnnoiaa iAfaauy m ByzantiDe e>nploy. M0lI1 oftheEmpire'. tnilitary arUlooacy wftt of Armnaiaa_cnlry and m the gth .....et lodI. """",nu-in Ar­m_i_n. fonned abolll no.-_1y-6ve per CftII of theEmpire'. anneel forces or pos";'bly evea mo",.

S'f.""ICWA£\.•

The Batde of Durazzo, 108•. Alexiu. planned 10 .!tacl< theNonnan camp from th..- directions, bUI Robert ClUscardadvanced 10 meel the main Byzantine anny on the pial",dKlroyiag the bridge behind him to preY....t Ri~ht; though hewu 1101 aware of il this also fruslrated Alexlu.' _drdmgtDovnnau. The Nonnaa nyot ft.an.I< p" way wbea it came up.pinst theVa~Guard, bul th_ in IU.... we.... defealedby the No........ Lar...try wh.... the')' bad ad,_ced too far fromtheir main body to receive supporL M.a.ay V....as:i.a. tool<muge m doe cJ:".rcb of 51 lltticbul, to which the Nonna.nJI set6ft. TIle Byzaatine CftItre pve way after a hanI fight .....etbrokem ....1.

The actual reformed army nucleus of theimmediate post·Manzikert era at fir.it seems tohave included the remnants of the old Tagmata,the Scholae, Excubiti, lkanatoi and Hetacreia allbeing recorded on occasion in theclosing decades ofthe 11th century, but these seem to have fadedaway to nothing before the end of Alexius I's longreign (I I [8). Nikephoros III seems .to have madelhe first concerted allempt to reorganize thc centralarmy, establishing both the Phrygian Chomatenoiregiment and, \vhilst Logothete of Michael VII,the Immortals. The laller appear to have beenraised from amongst the remnants of the EasternThemes and according to Bryennius were 10,000 innumber; he adds that the titleof'lmmortals' was atfirst applied only to the unit's officer's but wassoon used of the whole regiment. AJexius I himselfraised another regiment, the Archontopouloi or'Sons of leaders', recruited from amongst officers'orphans and numbering about '2,000 men. All threewere cavalry units, and in addition there were 'the

29

Page 37: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

debacle could only be paid for by levying cash fromhis family and friends and by expropriating churchpossessions. This was really the nadir of Byzantinefortunes, and though Alexius achieved severalnotable victories in the latter part of his reign (suchas over the Patzinaks at Levunium in 10gl, andover the Seljuks at PhilomeliOIl in 1(16) themilitary revival tbat took place under his auspiceswas of brief duration only. Though the army heldits own under his successors John II and i'vlanuell,succumbing to considerable Frankish influenceduring the reign of the latter, it never reallyrecovered from the disaster of Manzikert, a defeatwhich destroyed forever the Empire's credibility asa world super-power and dramatically marked theend of Byzantine mililary suprem:lcy.

"

pt..•<"< •·••,l••_" ••;-,;, ••;,.....". ••

• '{';.f'.,...:. ..." ••~.. ,,-'"'-.-.~; ."

'-·4 ., ..... - " ~_.." ..._~'. ~, ,., '

-'_.~--"'.'.".'...,.,. ",•,

.. ~..--,

, . '.'w"" ..~h. \

I

Another fine lltudy of skutalO; equipnn,nt from the JoshuaRoll, "bowing corselet" with pterugn, breast-bands andshoulder-pieceto. The neck..guards of their helmet,. appear tobe flexible 5Oa~probably luther. Uniforrnsin this &Ource arechiefly red, 50meumu blue. (Biblioteca Aposlolica Vaticana,Rome)

,...~._~-,..,...-,-~....,.,~~.....,-~~

soldiers of the E.mperor's household', the Ves­liarilac, probably comprising both cavalry andinfantry. Collectively it was these units, togetherwith the Anglo-Scandinavian Varangian Guard(probably included in the Vcsliaritac) and twO newTurkish units, the clitc VardariolS (ChristianizedTurks from the Vardar valley and Ochrida,probably Uzes) and the Turkopouloi or 'Sons ofTurks' (probably comprising the offspring ofTurkish mercenaries in Byzantine employ, some ofthem Christianized or of Byzantine mothers),which actually p!'Ovided the central army for theremainder ofAlexius' reign.

Several of these units fought against RobertGuiscard's Normans at the Battle of Durazzo(Dyrraehium) in 1081, the very year of Alexius'accession to the throne. It is clear, however, that hisarmy on this occasion was a pot-pourri of Illerce·naries and allies, with no more than a sprinkling ofnative Byzantine troops; and although the Varan­gians at first gave Guiscard's Italian allies a hardtime it was almost inevitable that when it came tothe crunch the whole polyglot collection failed tohold together. The Varangians, exhausted by therapidity of their advance and losing touch with themain body, were annihilated by Robert's infantry,the Vardariots fled, and the Byzantines' Serb allieswithdrew without striking a single blow. It was adisaster from which Alexius himself only narrowlyescaped.

The army that Alexius resurrected f!'Olll this

ThePlates.4 I Skulalos, Illh-J2/h eenluriesThough the old skuta remained in limited usc llntilthe end of the 11th century, by the I 080s at the verylatest the majority of heavy infantrymen insteadcarried the kite-shield (though unannOllred foot­soldiers such as 83 arc frequently depicted withcircular shields of 24-30in diameter as late as the13th century). This was probably as much a resultof Norman influence as anything else, thoughchanging tactics were undoubtedly a contributoryfactor, the day of the well-drilled, close-orderinfantryman being at an end.

.42 Skula/os, c. 950This is the way that skutatoi appear in mostpictOrial sources of the loth century. The lamellarklibanion is characteristic, usually only hip-lengthand often sleeveless. It was most often put on like ajacket and buckled down lhe front 01' back, thoughsome may have been put on like a poncho andbuckled clown the sides. His helmet is of spangen·helm construction, with reinforcing bands runningfrom edge to crown. The iron ring on tOp wOlildtake a crest like those of A3 and A4; thoughspecified in the manuals, such creslS arc only rarelyto be seen in contemporary illustrations.

Byzantine unifonllS arc usually coloured red orblue in surviving sources, though shades of green,

30

Page 38: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

mauve and occasionally purple also appear. Unitswere identified by their shield patterns or, a t least inthe case or cavalry units, by the panieu[ardislinctive shade or their lance-pennons, mainshield colour, crests and standards.

Though he is dean-shaven, neatly-trimmedbeard and moustache were more common by thelate loth century.

113 Pcllastos, c. 975Peltastoi first appear III the Sylloge Tacticorum,written in the second halrorthe loth century, andwe later encounter them, under the years 108 I and

These figurell reprelienting pliil"i ..re lllken fro.n ......5SOrfnlent of loth-century ivory caskets. The firstthing we see ill that. d ....pite the .nilil"ry m ....ual5·51..teonenls 10 the contrary. heltnels appeu.r 10 h ..vebeen in widespread use a.nong"t light infantry.Bows and swords are the armll mOllt COmnlonlydepicled on the caskelli. but nOle that of the lOp twofigures One hall a slightly cu.rved w_pon and theother a sabre-hilted sword, both probahly one-edgedparameria.

A lIIim.ilar lIIelection of skutatoi. the lOp three fromloth-eentury casketlll and the other two from laterm ..nulilcripts. The loth-century figures all wearbasically the saUle equipUlent. i.e. II. hip-lengthklibanion with or withoul sleeves ....d .. helmel withOr withoul aventail. Shields "re all rather IImall forskuta> and appear 10 be circular. The la"t figurecomK from the Madrid Scylitz.... hilll ..rmo.... reallydiffering little froUl that or the .oth century. Theshield ofthe lalll figure. whodaleli froUl c. IIIO,i5 theindigenoulll Byzantine 'three-cornered', or kite.shield; he appears to wear a padded cor~letofsomekind.

1084, in Anna Comnena's Alexiad. They were less­heavily equipped than the skutatoi, which wouldseem to mark them ror a linking role benveen heavyand light inrantry, like the peltasls or classicalGreek armies. However, their evolution may in raCthave been the result or cconomy measures necessi­tated by the Empire's stcadily increasing financialdifficulties, the skutatos's heavier and more expen­sive armour ha\!ing become less widely available;cenainly the Sylloge recommends that the mail orlamellar armour or heavy inrantrymen should be

31

Page 39: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

worn when available. It seems likely, then, that thedescription of such troops as peltasts reflected ontheir equipment rather than on their militaryfunction. The thickly-quilled bambakion is worn,plus an open-fronted helmet with neither hood noraventail. The Sylloge states that they carried thecircular thureos rather than the oval skuta. Armscomprised kontarion, javelins and, instead of asword, the sabre-hilted, one-edged paramerion.The end of the scabbard illustrated was square-cut.

A4 Skulalos, according 10 loth century miLitary manualsThe arms and armour listed in the various militarymanuals of this era probably represent the idealrather than the norm that a soldier could expect toencounter during active selvice. Quite probably itwas based on the equipment of the guard regimentsthat were based in Constantinople.

Body armour of skutalOi consisted of mailcorselet or horn or iron klibanion, though Leo VI'sTactica states that such armour was onen only wornby the first two ranks (sklllatoi generally formingup eight or sixteen ranks deep), those withoutsubstituting a bambakion. This was a padded andquilted corselet with hood and eighteen-inchsleeves, its name deriving from the Arabic wordptlmbuck, meaning cotton, from which it was largelymanufactured. Leo also mentions that some skuta­toi might in addition wear epilorikia, a similar typeofcorselet, over their mail or lamellar armour. Theleather harness of breaststrap and shoulder-piecesshown here is not mentioned in the manuals at allbut appears in the vast majority ofpictorial sources,worn mainly by foot-soldiers but also frequently byhorsemen.

Additional armour comprised greaves, vam­braces and leathergauntlets. Leostated that onl y thefront and rear ranks were 10 have greaves, while themuch earlier Strategicon records greaves being wornby JUSt the front (wo ranks. However, manuals ofthe later loth century seem to imply that greaveshad become standard skutatoi equipment, as 100

had a mail coif.The skUlatOS's main weapon was the twelve- to

fourteen-foot kontarion, made of light wood with asocketed blade at least eighteen inches long. Inbattle it was tbrust at cavalrymen and hurled atinfalllrymen. A few carried heavy javelins calledmenaulia in place of the long spear; these were

32

made of non-splintering wood such as cornel oroak. Other arms comprised sword or paramerionand tzikourion.

The large three by four-foot oval skuta can beseen cleady here. This was slightly curved but notconvex and had a brightly-painted face.

B1 Servant, 10tll-1 1 til untun'esEach guardsman, Tagmatic cavalryman or first­class thematic cavalryman, three or four second­class thematic cavalrymen and sixteen infantrymenhad a slave or paid groom or selvant 10 look afterthe baggage and perform menial chores. Selvantsof the Tagmatic guardsmen were provided by theOptimaton Theme, and the fact that this Themecould mustcr 4,000 men lends support to theargument that, c. 850-900, the Tagmata regimentscomprised 4,000 mcn in loto.

The infamrymen's servant drove a light mule­drawn cart which contained, among other items, ahand-mill, saw, twO spades, mallet, wicker basket,scythe, bill-hook and twO pick-axes. Since theservants were responsible for emrenching thearmy's camp each night the purpose of these tools isfairly obvious. If there was a shortage of scrvantsthe worst soldiers (probably defaulters) had tocarry out their chores. Known collectively as theluldum, servants carried a sling for self-defence andare sometimes to be found on the extreme Ranks ofa Byzantine battle-array.

B2 Puck-muLes, folll centuryPack-mules of the Tagmata were provided fromImperial ranches and stud-farms administered byan official called the Logothete of the Herds, as weremany of those used by the Imperial household, thelatter also receiving contributions of mules fromstate and church officers. A1l1hose of the Imperialhousehold wore red housings. The Imperial bag­gage alone required hundreds ofmules and horsesfor its transport, including 100 for cooking utensilsand silver table-ware and thirty for chandcliers,tapestries, silver bowls and cooking cauldrons.Other items accompanying the Emperor on cam­paign included a library, a complete pharmacy andwardrobe, Turkish leather baths, and a privatechapel complete with portable altar and ikons!

Page 40: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

83 Unannouud i'!fanlT)'11wn, II th-12th trnluritsInfamr)'lllcn of this type occur with incrcasingregularity in sources of thc II th-13th centuries asthe role of the fooHoJdier in Byzantine tacticssteadily dcclined. This particular figure is based onilluminations in the Scylitzes manuscript in ~I,ad·

rid. Despite thc absence ofamlour he is not really a

Detail. worth;, ofnotice i.a this lotb-ceatvol')' David .nd GoliatlailI"minnOon roD> tlae Paris Psalt....~ ... the upper seene,Goliath'. erested b"'hnd with .vftlULil oflnth.....tnpl' and b.ir;jaV<ilit> with b.att..pille, aDd ... the low" seene David'. em...qed panun.....o Notl! also the .piked bdnteu .1 t",Ft &adri&bl. Tbe sbi",ld fairly certainly COII~~bf:ff. and .............• ppears to be I"-cloer. (Bibliolbiq_ N.IioaaI... Paris)

33

Page 41: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

'light' inf.'1ntl'yman in the lnle sense of the word,since the sources tend lO show such soldiers fightingin fairly tight formations; more probably he shouldbe classified as a 'light-medium' infantryman,successor to the pe1tastos described under A3.However, psiloi of the type described below alsosurvived into the 12th century, being both illus­trated in the ylitzes manuscript and recorded inthe Alrxiad. '11e spear he carries is a rhiplarion orakoulion, a light throwing weapon.

84 PsiloI, Quording 10 Jotll untmy mllilary manualsThe principal weapon of the psilos or lightinfantryman was the composite bow drawn to thechest, though some substituted javelins, staff-slingor crossbow, illld archers were often issued inaddition with a sling as a reserve missile weapoll.The bowcasc/quivcr, slung from a Strap across theleft shoulder, carried fonyarrows. I n Leo's day thepsi los's only defensive wealx>n was a light hand·axe, the tzikourion. Armour was not usually worn,though the m:llluais say that light mail or lamellarCOrselelS should be supplied to as many archers aspossible ifavailable, at the same time admitting thedifficulty ofgctting them for more than a few menequipped thus were probably to be found only inheavy infantry units. A small shield oftwelve inchesdiameter was probably also carried, though Leoforbids it and Ihe earlier Strattgit:on says it was oftendiscarded in aClion as 'too heavy'! Kikephoros II,however, records that his archers are to have asmall shield, as well as twO quivers (one of sixtyarrows and one of forty), twO bows, four bow·strings, sling, sword and tzikourion.

Thou~h ofi.....produced, thill ivory "",..kel from the ViClor;'"...d Albert MUMum 1I1i11 ............11 one of the be.t ....p~ient.tio... of loth-c,.ntury "kut.to;' AU we.r lanlellar4:Qr""leuI, dU'1 nfl.he Rated genu'al ..11,.fl ruching to 1m......S

well all etbow, and carry konl..ria iO loag th..1 (h..,. diliJlppe..roul of Ihe lOp of th,. panel. Though badly cut th,. lutherbarn"' ofbrealit-band. and lihouider-piec:etI i. alto "pparent.Two large oVll..l.hieldll of the type called ,,"uta; ppa....nl,whiJ,. the RlIUI at e.trftne le:ft carri,... .....t,.,w circv.harthu.-- Th,. two 6pres at th,. Of'I""ite nod of th,. pand .....• ppa...... tly w,..riftS ...........dallie .nnour, .,illa,.r te-th,.r Orponibty quiJlM f.hric, .nd ..... perh.plI pel..stoi. AU havebe1n>t:es willa seal,. avftltail!l, sunnount«t by .. rinllo whida ..Cl"'"$1 would be .ttached OD parade. (V;t:eoria IUId AI'-rtMulOftUD, Londota)

H

CI Kafapl'Yat:IOS, aaordillg fo loll, (millry militarymanuals1110ugh inevitably the descriptions contained inthe various military handbcX>ks differ in detail, theequipment they list is similar enough to arrive atthis composite figure. Body·annour comprised amail corselet or. ifunavailable, a klibanion of iron,horn or leather lamellae. &th the Tatlita and lheSylloge mention that a klibanion could in fact be,,"om over the mail corselet. and it is clear from thepictorial sources that this was indeed fairly com­mon practice. At the shoulders smaller versions ofthe helmet crest were worn. Additional armourconsisted of vambraces and greavcs (here, ofwooden strips), leather gauntlets and-whenavailable-a mail hOClCi attached 10 the brim of thehelmet. Leo mentions that a padded wool or linengorget could be worn if the mail hood W:lS

unavailable, but he is quoting the earlier Strategicollpractically verbatim at this point. Padded armourcould also be worn under or Ovtr the corselet.

Kalaphractoi of the loth century carried as theirmain weapon either lance or bow. The lancc wasthe slender, twelve·loot komes (meaning 'barge­pole' !). more commonly referred to as a kontarionby this time, with ilS ten·inch head and colouredpennon (Leo says that the pennon was remo\'ed inbattle). It would seem lhat two lances \\ere issuedper lancer in Leo's time, the spare probably beingcarried in the baggagc train. Those equipped witha bow instead carried the quiver suspended at theright hip and the watcrproof bowcase, completewith a pouch containing a spare bowslring,

Page 42: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

suspended at the left. The quiver contained thirtyto forty 27in arrows. Bad archers substituted twojavelins for lhe bow, and during the second halfofthe century two javelins and a spear could also besubstituted for the lance. Other arms comprisedsword and dagger. Officers at least had a mace too,carried in a leather case attached to the saddle.

The kite-shield carried here is the indigenousByzantine type. Small circular targets twelveinches in diameter are specified in Leo's Tactica,but larger circular shields of twenty-four to thirtyinches diameter and, later, ki te-shields are far morecommon in pictorial sources. Archers officiallycarried no shield at atl, but it seems likely thatregulations were often ignored and most bow­armed katapbractoi probably carried the smalltwelve-inch target, strapped to the left forearm.

C2 Cavalry standard-bearer, J f th- J 2th centuriesThis is the type of cavalry equipment mostcommonly depicted in contemporary sources fromthe mid-I I til century onwards, comprising kite­shield, helmet with leather or scale aventail and ahip-length corselet with pteruges at the shouldersand, less often, the waisl. The sources seem mostcommonly to depict the corselet as scale armour,though the artislic convention followed couldequally well portray mail, or possibly on occasioneven lamellar. Note lhat the kile-shicld is now ofthe more characteristic 'Norman' type.

The horse accoutrements arc also fairly stan­dard; though one or two of the manuals mentionplumes sllspended from throat-lash, breast and

rump Slra ps, these appear only rarely in illustrativesources. Harness was most commonly dyed red orblack. olherwise being left as undyed yellowishleather. Saddle-cloths were often some shade ofred. f\ote, incidentally, the absence of spurs on therider's boolS, these apparently not being adoptedby the Byzantines until quite late.

The standard carried here is from the Scylitzcsmanuscript and is probably the type called abandon, used by both infantry and cavalry units byLeo V I's time. The size and shape of the bandonappears to have varied according to the size of theunit, those of dhoungoi and tunnai apparentlybeing similar to but longer than those of banda.Those in Scylitzcs have between three and eighttails, the number of tails fXlSSibly indica ting the sizeof the unit. The cross appears to have been the mostcommon standard device, often embroidered ingold and silver.

C3 Katapllractos, c. 1°50Most cavalrymen seem to have had cloaks. Leo'sTactica describes a waterproof, sandy-bro\lln col­oured lype whieh appears to have been army-issue,bUllhosc to be found in contemporary enamels andmanuscripts arc fairly certainly non-regulation andquite probably of civilian origin, being brightlycoloured with richly embroidered hems and panels.The panel on the fi'onl of the cloak, the characteris­tic shape of which can be seen here, was called atablioll. The cloak was HOi normally \110m in aClion,inslead being rolled up and strapped behind thesaddle.

35

Page 43: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

Detail from another .o,b-e:entury ivory panel depictingskutatoi, their large oval shields again much in evidence. Theequipment portrayed differs from that of the last picturemainly in the addition of pleruge" 10 the cor"e1et". (Metropo­litan Museum of Art, gUt ofJ. Piertnont Morgan '9'7)

D KlihanopllOros, c. 970The klibanophoroi were a revival by Nikcphoros IIof the truc caiaphracl, which had not bccn seensince late Roman times. The NikepllOri Praecept(l/.,;/ ilifaria describes Ihe armour of Ihese super-heavycavalry as a lamellar klibanion with elbow-lcngthsleeves, and over it a Ihick, padded epilorikion. Thehcad was protcclcd by an iron helmct with a mailhood tWO or thrcc layers thick which left only theeyes uncovered, Ihe forearms and lower legs beingprotected by splint-armour vambraces andgreaves, with any gaps filled by pieces of mail. Inadditioll mail-slrenglhened gaulltieis appear tohave been worn, while fcet were probably pro·teclCd by a metal overshoe. Their stout horses werelikewise heavily armoured, wearing klibania ofox­hide, split at the front for case of movement andleaving only thc eyes, nostrils and lower legsunprotccted. Other forms of horse·armour men­tioned in the sources include two or three layers offelt glued togcther; horn or iron lamcllae; andmail. An iron chann'on might also be worn.

They drcw up in a wedge formation on thebattlefield, Wilh tWCnty men in the first rank,twenty-four in thc sccond, and four more in eachconsecutive rank; lhe last rank (the twelfth) could

36

comprise as many as sixty-follr men, which meanstherc were 504 in thc whole unit. Apparently a unitof384 was more common (i.c. only tcn ranks). Thefront four ranks in eithcr case carried marwbar·boula in addition to the usual sword and komarion,and some mcn, lighlcr-equipped than the lancers,were armed instead with bows; if there were 300lanccrs there could be eighty archcrs, and if therewere 500 lancers there could be as many as 150archers.

Because of Iheir cost klibanophoroi were pro­bably limited to the Tagmata regiments, and itseems likely that j'vlanziken saw the end or Ihem.

E f Guard rdficer, c. 880Almosl identical figures appear in Byzalliinesources as early as the 4th century. This man, indrcss uniform, is probably a member of theHetaereia or one of the Tagmata regiments, andthe red cloak and circular cmbroidered panels onthe skin of his tunic seem to indicate thaI he is anofficer.

E2 Emperor in parade amlOur, c. 10/7Though armour such as this, complete with crown,is frequently depicted in contemporary sourcesbeing worn by Empel'Ors in bailie this is un·doubtedly artistic licencc, and there is little doubtthat in reality il was reserved for state occasions.Probably equipment more like that orCI was worn

Page 44: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

on campaign. Howcver, more functional vcrsionsof the gold armbands, sometimes ofscale btll moreoften of iron, appear in many sources from the 11thcentury onwards, oftcn cngraved to look likeptcruges.

£3 J/onbn of Ihe Basi/ikoi Anlhropoi, c. 880The Basilikoi Amhropoi (the 'men of the Basileus'or Emperor) ,,'erc court attendants ofa military orsemi.military nature. comprising spatharokandidatoi.spalharioi (called spalhorioi basi/ikoi to distinguishthem from those comprising the retinues ofstrategoi). Ilratores, kandidaloi and mandalores. col­lecti\'ely under the command of an official calledthe protospathariol by this period. We know fromConstantine Porphyrogenitus lhat they bore theImperial arms on parade, and it i~ clear frOIll theirnames that both spatharioi and spalharo­kandidatoi were, originally at least, sword·bearers.This particular figure. however, is p.-obably akandidatus. who wore white uniforms and were

David ....d Goliath ..,tne from the Mnool08hun of Basa II,c. '0'7' 'Goliath' is a Sood eotample of .. lItb-etntllry heavyiJIl....t ..ynuu'. WfllrinS the usuallameUar c:onelet with frinSeuad pler1lSes plus the~hof.. offiCft". His helm..t, ..d thoseor th.. l... th..r-annoured 'Is lites· c:rowded ...... ind th.. h.ill, i.of a ...e...- brimmed vari..ty, jnisce.u of a ke1t1e-b"lrnet,whidl ......... 10 6 ...1 litan apptarias at aboul thi. dat... n ..6~ of David, .... lower-da..ss C05tum.. -.ad armed -.ly with asliDfI:, .,.... probahly be lak<ra U rq>rtHDtati"e of th" ap­prs.......c:e of soldiers' HrvaalS OD tIH battldi,,1d (_ Plat.. s.).(Biblioteea NaUonaI.. Marriana, V~)

distinguishcd by their gold 100xIlIes or neck·chains(characteristic of most Byzantine guardsmen fromlate-Roman times until at least the llih century);perhaps it was such sword·bcaring kandidatoi whowere called spatharokandidatoi. On state occasionsthey could be mounted, and wore gilded armourand helmcts and had \\'hite cloaks and standards.The kandidatoi were probably the oldest clementof the Basilikoi Anthropoi, its members originallybeing selected, on the basis of their size andstrength, from the ranks of the Scholae. Thekandidatoi and spatharioi at least had their ownhalls in lhe Imperial Palace.

FI Nus mercenary. c. 950Many ofthosc Scandinavians who settled in Russiawere soon influenced by the dress of their Slav andAsiatic ncighbours. This man, for instance, wearsthe bleached while linen tunic charactcristic of tileSlavs, and his stripcd, baggy trousers arc probablyof Asiatic origin; another Asiatic trait adopted bysome Rus was the tattooing of the hands and armsup to the shoulder. Boots and a cloak clasped at theshoulder completed their costumc. Most wore mailcoats, and anns comprised spear, axe, sword and

Thous;b datias to th.. '3th «utury ou. is the ,,"pot ofhelmet U i .....0.... by Goliath .... th.. M.....1op1Ut> pie Thispanic:ular one i. v...-y ornate, "",«,,nd ..ilb and....besq.... pan........ and ..... clearly thai of ... oflie.-r.

37

Page 45: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

dagger. This man's helmet is ofSla" design, as is hisshield, illustrated in the accompanying mono·chrome sketch. Rectangular shields of this typepersisted in Russia for many centuries.

In appearance the Rus were 'tall as date palms'with red or blond hair and ruddy complexions.~Iost \\'ere bearded, though some affected onlydrooping Turkish-style mousl3chcs. PrinceSvyatOslav of Kiev even shaved his head Turkish­fashion, leavingjust two long locks ofhair to signifyhis rank.

F2 Varangian Guardsman.. c. 1000The most distinctive feature of the Varangian'sequipment was undeniably his axe, which appearsto have been retained in preference to the rhom­phaia more usually carried by Byzantine guards­men. Psellus. however, claims that every Varan­gian 'wi Ihout exception' was armed wi th shield andrhomphaia, 'a one-edged sword of heavy ironwhich they carry suspended from the right shoul­der' (perhaps meaning it was sloped across the rightshoulder when not in use).

TIlough the two-handed axe was their mainweapon spears and swords are also memioned inthe sources. It is clear from the sagas that many menkept their own swords when they entered theGuard, and since their axes too were fairly certainlybrought from home we ha\'e leave to doubt justho\\ much of their equipment (as opposed tounifonm) was actually official Byzantine issue.~Iost probably a mixture of Scandinavian andByzamine ge.tr was in usc, the latter probablybecoming predominant the longer a man stayed inthe Guard as his own equipment wore OUI.

We know from Anna Comnena Ihal Varangianswere generally heavily armoured, and this man hastaken full advaillage of access 10 the Imperialarsenals to supplement his own equipment withvambraces and greaves. Their shields probablyremained circular throughout the 11th century,but in 1122 we hear of Varangians with kite­shields.

F3 Varanlian Guardsman in dress uniform, c. 1030Laxdaela Saga records several ex-VarangianGuards wearing searlel clothes when Ihey rcturned10 Iceland in about 103°01' 1040. The equipment ofBolli Bollasson, their leader. is described in some

38

• • • • •

f:,J II, ~,: •

,

o 1/), .

• , •,

\• • I,I~,• ,,/ ~ , •

j .• . I I •

\ I,• • • • •

detail as comprising silk clothes (presented by theEmperor himself), a scarlet cloak, a gilded helmet,and a scarlet shield decorated with a warrioroutlined in gold, probably all Byzantine issuc. Inaddilion, his sword had a gold-decorated hilt andgrip, which is of intercst sincc wc know that theright to \\'ear a gold-hilted sword was one of theprivileges that accompanied the COllrt rank ofmanglahiles, which was later to be held by HaraldHardraada as an officer of the Guard; Bolli,therefore, may likcwise have held this rank.

GI Tra/Ndlos, according /0 loth unlury militarymanuals~ative light cavalry, called lra/N{itat, are describedby both Leo VI and the author of the SJ·UogtTaclicorum. They were unannoured (though somemight wear a hcwxl ofhorn scales) and armed withsword, kootarion and IWO or Ihree javelins, thelatter apparently not to cxceed ninc fect in length.It is also possible that some were equipped as horse­archers. The shield they carried appears to havebeen the large infantry skuta or lhureos, thecircular thureos apparently remaining popularamongsl light cavalry untillhe 13th century. TheS]lIogl' also ll1f'ntions a light cavalry shield oflwellly·seven inches diameter, and the smalltwelve-inch target seems to have been used too.

G2 Pat{inak mncmaries, I Ilh cmturyThe Patzinaks, or Pechenegs, were a Turkishpeople often to be found in Byzantine employ fromthe lategth century onwards, comprising one of thelargest mercenary elements of lhe anny by themiddle to late II th century. ~Iany dctachmentswere employed as a son ofprovincial police by that

Page 46: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

•,- ....-.-.-

"\'.. •• I:;:-_.~~ H TWN (HLH"}1 'rA'~ .

lime, in which role we find them dogging theunruly march of the Fil'st Crusade through theEmpire's EUl'opean provinces.

Like all Asiatic peoples their main weapon wasthe composite bow, but javelins, spear, sabre andhand-axe were also carried, as well as a lasso used toentangle enemy horses and riders in close comb-"ll.For defence a small circular shield ofosiers, wood orhide was carried. Body armour of lamellar con­struction could also be worn depending on thesocial status of the wearer; chieftains and theirretinues, for instance, were normally armoured.

HI Stljuk mtrUlIary. la/t II th ernlurySurprisingly, Seljuk Turks did not appear inByzantine service until oJttr their vicwry overRomanus IV at ~Ianzikert, when between 1071and loBt successive Emperors and generals,desperate for troops, falher shortsightcdly intro­duce<llarge nllmbers of them into Anatolia, mostof which they soon overl'an. They were first-classsoldiers ofa rathel' fiercc and savagc disposition­'ferociOliS beasts covered in blood' is the way oneArmenian described them. At this time the Byzan·tines employed them exclusively as light cavall)'horse-archers.

This man, a chieftain, wears the characteristicwide·skirted topcoat which had a flap called amuqallab passing diagonally across thc chcst fromright to left; this was tied by tapes under the armand down the left side. From the belt, which ismade of silver plaques, hang his bowcase, quiver

'Cneek Fi ', or Sn Fi.... (pyr tl,,0I1I66'0,,) 1111. the ByzanlinHthemsel" c:all«l it. had been invenl«l in ConlltantinopleC. 673' Throughout ils hislory, however, il appe..... 10 h ..vebeen used entinly in naval and siege warfare. II w... firedfrom liiphon.. by h ....ting or by a jet of waler (opinion.. differ)and was ....In:mely difficult 10 extinguish. One of the .dvaneetlmade during thill era wa" the in"....tiOD of 'hand lCyrillg...•(mikroi "plo_~.) in Leo Vl'li reign, theM being fired frombehind iron IIhield¥. The pi<:t~ ben, from the MadridScyliues, show. the mOre cnnv.... lionat .hipboard >lH, withthe Fire being fired from copper, bronze or ironoCO'o'ered lUbes.

(Biblioteca Nacion.a.l, Madrid)

and sword-scabbard. On horseback he would alsocarry javelins, lasso and probably onc or two sparebows and quivers. TIle mace was also a popular\\·capon. Armour was mainly of lamellar con­struction, but here a captured mail eorselct is beingworn under thc topcoat. ~Iost Seljuks, however.were unannoured and would ha\·e carried forprotcction only the small shield. This, likc theirclothes, appears 10 have been brightly coloured.

H2.1-/3 Ilalo·XomulI/ merunaries, lale I lilt ctliluryFrom 1038 onwards, undcr such leaders as HerveFrankopoulos, Robcrt Crispin ane! Roussel deBaillcul, bands ofNorman adventurers had flockcdto take service with the Emperor, though il soonbecame apparent that their main ambition was tocarve out their own little ~ormand)' in the heart ofAnalOlia.

It seems almost superfluous to describe their armsand equipment here, but the Altxiad contains agood description which deserves 10 be qUOted:'Kcltic annour: Anna says, 'consists of a tunic ofinterwovcn iron rings linked one with another; theiron is of good quality, and being arrowproof

39

Page 47: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

protects the wearer's body. This armour is sup­plemented by a shield, not circular but long, broadat the top and tapering to a point. Inside it isslightly curved; tbe outside is smooth and shinywith a flashing, bronze boss. That shield couldrepel any arrow, whether Skythian or Persian or

40

even fired by the arms of a giam, and make itrebound against the firer.' She also says ofNorrnanarmour that it 'made them almost, if not com­pletely invulnerable'. The armour of H3, based ona chesspiece from Norman Italy, is less characteris­tic and betrays considerable Byzantine influence.

Page 48: Osprey+[Men+at+Arms+089]+Byzantine.armies.886 1118

=MILITARY 'IE:\'-\T-:\R'IS SERIES mm:JMllIT.... RY

An ........."lIcd >ou"", of inr ion on IIw ..nir~im.i~ni. and .PP'O"........... <A 1M .......!d•• fi,h,,"II:rnftI"rpuc ..... p•..,. ". 1M II '·.~ ",Ioco':n ...bjcn••• di ....".._.M Impnial R""",.. um~'.

thr,.~ awl Gennan .i' 'K'I"OOP> in .. populu 48,>00,,, fonna' ,nrlud,"tl_-Ill photosnopho and di.~and .i,h. fulkolour pblla.

CO.\II' \:"10:-': SER.n:.s nO.'1 OSI'R£\

,unO«J,i inr"...,....;.,., on 'IK unir_.nd ;"";pi. of ,hot .'orld', m..,," (arnow. mih••~ f",,",,-

Ea<h booL con"',,,,, .....,. 50 pho.OV"phs and di.JN...... and IlI"lICS of (.. Il-colour a.." ...".

",AiUtiORIkfini.;,·c • ..,,1,-",0 or ,he annou., """'" ';&cta and moli.".,...., of1M fi,l";,,, men ..rh;..~.f_h ......1"'•• b<>o~ ....nui.... "", )'. and exploded u,.-""k ..f,h. ,,'arT;or', ......_ and ",,,,,,,, •.

N.:W VANGUARllComp«hc".h-c hist","i". "r ,h. dui,n. de,-dup",,,,,, .nd OI",,,,,i,,,,.1 ..'" of lh. "odd'. um"..".,j

".hid.. and ... ill«y. Each ~S.I"'I' I>ool "",nai". d.h, pa,"" "f full...".,;"IO' .",,·o,k ;"dudi", " d.,.iltd...., ...... y "f ,I>< ,'chidc'. 'n'cr''''',

CA;\II'AIG~

Conrioc••u,hori, ••;.,c _n,. of<kcioin cncoun'cn; in mili,.I'}' hi.,.,..!. Ea<:h %-P~.c buol """•• ,....~ 'han"" illus<.......... indud'n. "...... orden "rba,,1c .nd colour pia."", plu........... or~I... ltlc"..... ,ha,,,..r\< d.c m.ialJ ....a of,hcnmpaill"

u,_a..-_

,....._----- -------.. , ,. ... .'. --

9 JII"'Ulll