P R E F A C E

H AV E r i di u r I hea d this quest on sc ssed eve since was a child , r u but have never yet w itten anything abo t it except, in my

Sma ll Ta lk a t Wre lc md . r I y In the First Se ies , page 75 , was talking r n I : about t avelli g on the Continent , and said Plenty of people went to Switzerland at the time when I first — — went 1 869 far more than when my father went there thi rty

but i r r . years before , nothing l ke the c owds that go the e now They kept more to peaks and passes then ; and they were always talking

’ o : of s passage of the . 'unius was talked ut Tich borne and Dreyfus were yet to come ; and Hannibal filled the gap . I used to hear them at home as well as there ; and they all had their ’ u al — d Ar entiere r pet ro tes for Hannib Col g , Mont Genev e , Mont

ni r Ce s , Little Mont Cenis , Little St Be nard and Great St Bernard , m and even Si plon and St Gothard . In 1 87 1 I went looking for traces of the vinegar on the Great St Bernard . My father upheld the Cenis routes as the onl y passes from which you can look down al upon the plains of . I doubt if Hannib did look down . I think he may have shown his men their line of march up on a u Ari sta ora s u ShOW r r map , j st as g sed a map to the Spa tans thei line of march 282 years earlier . I A for Arista oras wrote naxagoras by mistake g , and passed it in the proofs ; and it was printed in the first impression of the r I r . First Se ies , though co rected in the second impression men tioned r 1 02 my mistake in the Second Se ies , page , and this and other instances led me on to say “ I fancy that the Greek and authors wrote the wrong

r . vi wo d now and then , and never noticed it That is not the ew of textual critics and editors : they ascri be all errors to the men who

t i rr copied out the manuscri pts . But his l mits them to e ors that r t u r might a ise in copying , and h s rest icts the choice of emendations

I s a f u . w or Amy far too m ch Take such an emendation as in ,

xxi . 1 . r 3 This makes Livy say that the iver was the Isere, not the o ui u Sa ne ; but the context req res him to say it was the D rance ,

‘ ’ ‘ ’ otherwise he would be saying ri ght instead of left a few lines

r A r a rm for is a u . i m f rthe on copy st might easily w ite , so this

u u . emendation is accepted , altho gh it does not s it

Such emendations are deceitful things . In this case they make i an d Pol b os 1 . Livy say the Isere , make y say it also , i i 49 , though he says something else ; and then Members of the Al pine Club go r r u saying that the ive m st have been the Isere , since Livy and

Pol bios r y agree in saying that it was . Othe folk may say it does not matter what the ri ver was ; but that is a reason for leaving the o for . u u whole thing alone, not getting it wrong If y take it p at ou ul ri r W ur a all , y sho d not sk the so t of snubbing that estb y g ve

r r - ou l : the he ald afte cross examination Go away, y sil y man you ’ ’ don t even understand your own sill y science .

( v i ) That brought me letters from Members of the Alpine Club and

from a former President who is a champion of the Isere route .

And this is my reply .

CE CI L T O R R .

Y O WR E Y L AND N D E R , U I G L ST LE H ,

D E V O N .

l POSTSCRIPT . I fear there is much repetition in the fol owing

I for pages , but have a reason it . The same facts recur in different contexts ; and I have sometimes thought it better to re - state a

i - fact than merely g ve cross references . SIMPLON G REAT SAINT BERNARD LITTLE SAINT BERNARD MONT CENIS MONT GENEVRE Proposed route COL DE LA TRAVERSETI' E COL D’ARGENTIERE f 1 Pol bios u a r ut r . r . y , of co rse , is the best a ho ity He was bo n ’ 1 in Hannibal s lifetime ; and he mentions (iii . 48 . 2) that before ’ r u A he w ote his acco nt of Hannibal s passage of the lps , he went over the ground himself to make qui te sure . A di — 2 . s he writes in Greek , he gives the stances in stades nine

— 6—10 r 600 stades make an English mile and (iii . 39 . ) he eckons 2 1 600 r A ur stades from Cartagena to the Ebro, from the e to mp ias , 1 600 at the Mediterranean end of the Pyrenees , from there to the

’ di a ba szs 1 00 al r crossing ( ) of the Rhone , 4 ong the iver from its

ana bolé A 1 200 r crossing to the ascent ( ) of the lps , and ac oss the

Al ps into Italy .

ul 0 A ri 3 . He says (iii . 39 . 8) that he calc ated the 16 0 from mpu as to the crossing of the Rhone by the milestones on the Roman road , reckoning eight stades to a Roman mile . Thus , in Roman

di u 2 r 200 A ur miles his stances wo ld be 3 5 to the Eb o , to mp ias ,

00 n 1 ri 1 0 Al 2 . to the Rho e , 75 along the ver, and 5 across the ps

6 . He also says (iii . 5 3) that Hannibal took fifteen days in marching 1 800 200 . 0 . 1 the stades , and (iii 5 ) took ten days in marching , or 10 part of the 1400 . Both cases give an average of 80 stades Roman mil es a day ; and this looks as if he knew the time employed

di d di c al c u here but not know the stance covered , and therefore l lated the di stance from the time . He certain y knew the time

as employed upon the march of fifteen days , he elsewhere gives

1 . 0 8 2 . I 2 . 6 . . 6 . 8 the days in detail , ii 5 5 , , 5 , , 53 5 , , 9 , 54 4 , 55 , , l 1 . n u 56 . But Han ibal wo d not really have gone at the same pace on the fifteen days when he was fighting his way through the mountains as on the ten days when he was marching up the river

Pol bios r u . nopposed y must have taken a standard ate, and used it indiscriminately when he had no help from milestones .

A hi am 8 00 1 0 0 4 . fter giving distances w ch ount to 4 stades or 5 — . . 6 1 0 . . 1 1 Roman miles in iii 39 , he goes on in iii 39 to give their

a t 1 1 total as about (p r) 9000 stades or 25 Roman mil es . There must be an error in the total or the items . I fancy the total should be

8000 1 000 was l stades or Roman miles , as he more ikely to reckon

iii . . 1050 as about 1000 than as about 1 1 25 . He mentions ( 56 3) that Hannibal took five months on the march ; and 1 50 days for the 1 050 miles gives an average of 7 Roman miles a day . Large forces coul d not move fast . The column woul d be some miles in

- u length , and the advance g ard might be close to the new camping ground before the rear- guard left the previous camp ; and time

- would have to be allowed for the rear guard to come up .

1 0 Pol bios but 5 . Strabo was 5 years j unior to y ; the Roman

r . 1 . oad from Spain would not have moved , and he says (iv 3) that it crossed the Rhone at Tarascon . He goes on to say that it

r u Aix A bifurcated there , one b anch going thro gh to ntibes on the

( 2 ) rr t r ai Medite anean coast , while the other went h ough Cav llon and along the to the beginning of the ascent (a na ba sis) of the A 6 ru r r lps , 3 miles from Tarascon , thence to Emb n , 99 miles fu the r a di ri 1 on , and thence through the B i ncon st ct in 7 miles to

in . u Cesanne , the first town Italy (The road m st thus have crossed the Al ps by the pass of Mont Genevre . ) Strabo here treats the n A beginni g of the ascent of the lps as a definite point , marked by

Pol bios A a milestone, just as y treats the ascent of the lps and the crossing of the Rhone as defini te points from which measurements coul d be made . r r u 6 . Roman oads were not built cap icio sly ; and the road from Spain woul d not have crossed the Rhone at Tarascon unl ess there was some substantial reason for crossing it j ust there rather than

r a little higher up o a little lower down . This reason must have existed at the time when Hannibal crossed the Rhone a few years

’ before the road was made ; and by reckoning Hannibal s march

di Pol bios r r m accor ng to the Roman milestones , y athe i plies that al Hannibal came that way . In any case Hannib cannot have devi

r r r ated greatly f om the Roman oad , as that ran nea the coast , 1 and Polybios says (iii . 4 . 7) that Hannibal kept the sea upon hi s ri de ion ech ght hand ( z én) as he marched . A the 7 . fter mentioning (iii . 37 . 8) that the Rhone entered sea Pol io i b s . 2 . 1 by several mouths , y says (iii 4 ) that Hann bal crossed

’ r r it whe e it was a single stream , about fou days march from the di . A sea It ceases to be a single stream at rles , and vides there into branches leading to the mouths ; an d is eight English

( 3 ) - — r miles below Tarascon and fiv e and twenty f om the sea . Four — — days at ten miles a day see paragraph 3 woul d give 40 Roman

m di r r . iles , or 37 English , for the stance f om Ta ascon to the sea

n u ur But as Ha nibal was coming from the west , he wo ld have to t n rn inland at the weste most mouth of the Rhone, somewhere near

A u r u u r . ig es Mo tes , whence the distance wo ld be f lly fo ty miles

u r r r - Rh No do bt, the e we e many c ossing places on the one , and — any of them might be known as a Diab asis see paragraph 2 r as this is merely the Greek for crossing . But there we e many

r - R Tra ec tum c ossing places on the hine, yet there was only one j , r but nl ' u Pol bios and many on the Euph ates , o y one e gma ; and y f al ways speaks of the Diab asis of the Rhone as i there were no other . — — P l io 1 . 8 . A o b s . . 2 . 1 6 . 1 ccording to y (iii 4 , 4 9 , 44 3 , 45 , 47 ,

. 1 r 49 . ) Scipio was taking a Roman a my from Italy to Spain by r u di s sea , but stopped at the easte nmost mo th of the Rhone and r ri al embarked the a my there , on hea ng that Hannibal had ready

o al in . ut r reached the cross g of the Rhone He sent cav y to scout , ni al out rr and Han b likewise sent cavalry, having heard of the a ival

’ : r of the Roman ships . These two forces met Scipio s d ove Hanni ’ a u b l s back and p shed on far enough to see his camp , and then

' a mutzc a t e returned with information . Scipio forthwith (p ) embarked the baggage and then marched up al ong the river with

hi s as r s eudén . whole force , he was eage ( p ) to attack the enemy r n r u But on eachi g the c ossing , he fo nd that the enemy had left r r r thei camp three days before , having sta ted the mo ning after

r the encounter of the caval y .

( 4 ) ’ m r i r ri . a al 9 If the ca p was at Ta scon , Sc pio s cav y had a de of or i m sixty seventy Engl sh iles , there and back , besides some fighting on the way ; and after their return there was the embarca tion of the baggage ; so that Scipio coul d hardly have begun his march till the next morning or next afternoon . He may have been

’ r n fu r ma chi g on three days , though not a ll day s ma ch on the first day and perhaps not on the third ; and he woul d entrench his

i r fa r off. camp each n ght with extra care, as the enemy we e not r but m Even so , his prog ess seems a little slow ; had he gone ten iles

u r ul ri u . f rthe in the time , there wo d have been a se o s check

ri 10 . Ten miles above Tarascon the ver Durance flows into the

1 al l r r ul Rhone . Livy (xxi . 3 ) says that , of the ive s of Ga , the ur fa r u lon e dz ic illimus D ance was the most diffic lt to cross , g fi tra nsitu and he mentions that it happened to be swollen by rains

’ or e i bmb (f t m us a uc tus) at the time when Hannibal was there . If Hannibal had crossed the Rhone more than ten miles above ’ r u ur tw r Ta ascon , he wo ld have had the D ance be een Scipio s fo ces and his own ; and Polybios says nothing to suggest that there was any obstacle between . ’ A ri 1 1 . t thi s pe od the Romans had only a militia ; and Scipio s marches must not be calcul ated on the scale of marches by the u rm ll reg lar a y that Marius created a century later on . Sti less

ul ul Ve etius sho d they be calc ated from statements by g , who lived six centuri es later and dealt with marches along Roman roads to

At r en established forts and towns . this pe iod a Roman force

r m . t enched itself each night , and the entrenching took some ti e

( 5 ) A r — 1 2 . Pol bi . ni cco ding to y os (iii 49 . 5 7) Han bal marched up al r r ong the iver afte leaving his camp at the crossing of the Rhone , a nd rri ur u u r l a ved in fo days at the Island , a pop lo s and fe ti e place with the Rhone running along one side of it and the Sc ora s along

. ri but Pol bios the other It was not st ctly an island , was ( y says) r r of the shape and size of the Delta of the Nile , with ive on two sides and a range of mountains instead of the sea for the third

. 1 l P l side Livy (xxi . 3 ) fo lows oybios almost word for word in saying that Hannibal marched up along the Rhone to a di stri ct r r called the Island , where another ive flowed into the Rhone . But

r r Ar hi o ri he calls the othe iver the ar, w ch is the Sa ne , and thus b ngs

n al m r 200 . u Han ib up to Lyons , iles f om the sea This wo ld mean

2 u r an average of 5 miles a day , the confl ence being fou days from r u the crossing and the c ossing being fo r days from the sea , whereas Polybios gives an average of only 7 miles a day for the whole march and 10 miles a day for 250 miles out of the final 3 25 after r in : the c oss g of the Rhone see paragraphs 3 , 4 .

’ 1 As . i 3 a reason for Hannibal s going so far nland , Livy says 1 (xxi . 3 ) that Hannibal did not wish to fight a battle against the

ur r Romans till he was actually in Italy, and the f the he kept away r hi from the sea , the less was the isk of meeting the enemy . But t s 2 is not borne out by what he says just afterwards (xxi . 3 ) where

i P l i . o b os . 2 . he is copy ng from y , iii 49 , 3 He says there that Scipio marched back to his ships on finding that the enemy had quitted

r am t e - r r thei c p at the crossing of the Rhone, emba ked the a my a nd ul it r ur him sent the b k of on to Spain , and et ned to Italy self ( 6 ) to take command there if Hannibal got through the Al ps . The march - back and the embarcation must have been as bri sk as the

- u Pol bios u s eudén iii . . march p , y sing the same term for both , p , 45 4 , : rri 49 . 4 . Hannibal had good information he soon knew of the a val — of the Roman ships at the mouth of the Rhone see paragraph 8 and woul d soon know of their departure ; and he woul d not go 200 miles inland to avoid the enemy if he knew they were not there .

1 . 4 In order to reduce the distance , the editors of Livy have “ ri tampered with the text , and have p nted Isara in place of

A r o r . ra , thus changing the Sa ne into the Ise e They seem to have

I alic u — m 6 . t s . 2 forgotten Silius , iii 44 47 He was al ost a contem — ’ — p ora ry of Livy born onl y eight years after Livy s death and in

’ his descri ption of Hannibal s march he not only speaks of the

fl A r con uence of the ra and the Rhone , but contrasts the rapidity of the Rhone wi th the quiet flow of the Arar; and that is j ust the di o di f r fference between the Rhone and the Sa ne at Lyons , a fe ence i r U U r r . emarked by Seneca (L D S , 7) and othe ancient w ters Silius goes on to speak of the Tri c a stini and the Voc ontii and the ri ver u 1 D rance exactly as Livy (xxi . 3 ) goes on to speak of them , and

’ i L v in almost the same words . It seems clear that Sil us had i y s him r u A ar words before when he w ote , and that he fo nd r there , “ i r o . not Isara , and had no doubts about the ver being the Sa ne al nl That re ly is decisive for the text of Livy . The o y pretext of the editors is that there is a mark like an 3 between the words ibi and

Ara r ri in one of the manusc pts of Livy .

( 7 ) 1 . i r r r u wi in r 5 These ed to s we e ve y n se making Livy say Isa a , as he woul d not have mentioned the confluence of the Isere and the Rhone without saying a good deal more . The monuments of

r 1 B C. the the victo y of 8 August 1 2 . were at confluence of the

r iv . 1 . 1 1 Isere and the Rhone . St abo mentions them ( ) and Livy, his r u k n who was contempora y, m st have now that they were

. ri r hi s there (Livy desc bed the victo y in one of lost books , lxi , the contents of which are known from the epitome . ) Livy says

(xxi . 3 1) that as soon as Hannibal reached the confluence of the o r Allobri es . Sa ne and the Rhone, he secured the suppo t of the g

u I s é re If this had happened at the confl ence of the and the Rhone , Livy would have made a telling point of the contras t between ’ Hannibal s securing the support of the Allobriges for hi s onslaught r an r on Rome , and thei subjugation by the Rom s at the ve y same

r place a centu y later on .

ri 16 . No doubt the Saone ses in the Vosges and not in what we

A . ri c 1 . 1 1 . 10 all the lps ; but Strabo (iv . ) and Ptolemy (ii 3) desc be as ri A i it sing in the lps , and Livy is probably tak ng the same View in u u saying that it rose there . It seems unwise to arg e that he m st have meant the Isere because the Isere rises in the Alps .

1 . Am 1 1 . 1 A r al 7 mianus says (xv . 7) that the ra was so known

Amrim uem Sa uc ona m a ella nt. as the Saucona , , q pp He quotes Tirn a enes a xv . . 2 g by n me in 9 , and probably quotes him here, as

- FLUVI I S 6 r the pseudo Plutarch (DE , ) quotes him for info mation u A hi m c abo t the rar, w ch infor ation (it says) he opied from

alli h Tirna enes r I C st enes . g was a contempora y of Livy ; and ( 8 ) u i Pol bios Sc ora s s spect that when Livy was copy ng y , he took for u or r a am but r it A r Sa cona a va i nt of that n e , t anslated as ra because this name was better known .

1 8 r Pol bios r . The edito s of y have also tampered with thei text “ f r ri I saras o Sc oras . . 6 ri r was and p nted in iii 49 , as if the ve clearly the Isere . They seem to have forgotten what he says in As al r . r 1 2 the next sentence eady mentioned in parag aph , he says that the so-called island between the Sc 0ras and the Rh one

m as was of the sa e shape the Delta of the Nile , with these two ri vers as the sides and a range of mountains (instead of sea) as

B i u r r he r . ut t t the base is a q ad angle , not a t iangle , between Ise e

: the d a and the Rhone the Isere on south , the Mont Chat on the

Rh r east, and the one on the no th as well as on the west , as its

- h Pol ios course turns round a ri ght angle at Lyons . W en yb wanted

r ul so—he tetra to say that a place was quad ang ar, he said says ' o — onos to s vi . 2 . 2 ul g p in 7 and he wo d have said so here, if that had been his meaning .

u ri c ora r 1 9 . Some of the man sc pts have S s and othe s have Soaras and this discrepancy is not uncommon in manuscripts

Pol bios r ri i of y , as if they all were copied f om the w t ng of a man Al who made his phas and Omegas very much alike . Casaubon ” “ “ al tered Sc oras or Soaras into Ara ros to make it agree

u I sa ra s r with Livy . Cl ver altered it into to make it ag ee wi th the alteration in Livy . Neither of them had any better reason W u for the change . hen editors doubt a reading they o ght to “ ul r Ar r query it . These editors sho d have p inted a Isara) and

( 9 ) c ora s I sa ra s ri A S ) , or p nted rar in the text and Isara ’ Drakenborc h s di in a footnote , as in e tion of Livy ; but they have “ I sa ra s l printed Isara and in the text itse f, and in many

‘ editions they have not even added footnotes . Readers are thereby

the ri Pol bios misled , and think they have autho ty of Livy and y for saying that the Island was at the confluence of the I sé re and

nl r di r the Rhone , when in reality they have o y the autho ity of e to s

r k who knew no mo e about the matter than they now themselves .

Af u o 20 . ter taking Hannibal p to the confluence of the Sa ne

. 1 i ai and the Rhone , Livy says (xxi 3 ) that instead of mak ng str ght

ad l a m A a v . for the lps , Hannibal then turned to the left , Strabo

. 6 . 1 1 twice says (iv 7 , ) that there were two roads from Lyons to

al r n It y, one ove the Pe nine pass , and the second through the

rr r Centrones r rr te ito y of the , meeting the fi st in the te itory of the

al . s S assi These clearly were the Great and Little St Bernard route , meeting at ; and if Hannibal was near Lyons and took the

s r n Great instead of the Little , he might fairly be de c ibed as turni g n A to hi s left i stead of making straight for the lps . Livy says 8 ni (xxi . 3 ) that many people thought that Han bal crossed the Pennine pass (the Great St Bernard) as they fancied that the “ “ ” name Pennine was deri ved from Punic . He also says that ” Coelius al Cremonis u um thought that Hannib had crossed the j g ,

s and he a sumes that this must be the Little St Bernard , as he says that both these routes would have brought Hannibal into A rri . 1 . 0 2 the te tory of the Libici , and Ptolemy (iii 3 , 3 ) fixes osta

( I O )

uri i r rr r belonged to the Ta ni , and m ght be desc ibed as te ito y of

the by writers of a later age .

r . 6 . 1 2 ur r A 23 . St abo says (iv ) that only fo passes ac oss the lps

r were mentioned by Polybios . The nea est to the Medi terranean “ u rri r u went thro gh the te to y of the Lig res ; the next , which Han

al r u rri r uri i nib crossed , th o gh the te to y of the Ta n ; the next , through the territory of the ; and the fourth through the

m r r rr r Rhaeti . te ito y of the But the i po tant wo ds, which Han ” a re u r r nibal crossed , not in all the man sc ipts of St abo , and there ol bio fore are suspected as interpolations . P y s does not mention the Salassi or the Rhaeti in the extant portion of his work ; but a pass through the territory of the Rhae ti might have brought

ri r I nsubres r . 1 . Hannibal into the ter to y of the , as St abo (vii 5)

rr ri c onterrninous Rhaeti says that these te ito es were , the having some terri tory on the south side of the Alps as well as on the north

a m bu side . The Rh etian pass might thus have been the Si plon ; t it is incredible that Hannibal should have crossed a pass so far

eastward as the Simplon or even the Great St Bernard .

2 . r E NEI DEM . 1 4 Varro is quoted by Se vius (AD , x 3) as men tionin A : r g five passes across the lps one alongside the sea , th ough

rri r ur the te to y of the Lig es ; a second , which Hannibal crossed ; u a third , by which went to Spain ; a fo rth , by which — Hasdrubal came into Italy ; and a fifth in the Graian Al ps pre

r sumably the Little St Bernard : see paragraph 58 . This ag ees with ’ Strabo s quotation from Polybios in making Hannibal cross the ’ next pass to the coast - road ; and if it were Varro s own statement

( 12 ) u u r . B ut r at first hand, it wo ld have high a tho ity simila inter

olations u p may be s spected here , as it makes Hannibal and Has A a l i r . drub cross d ffe ent passes , whereas Livy (xxvii 39) and ppian

B 2 r i ru a (HANNI AL , 5 ) ag ee in say ng that Hasd bal crossed the s me pass that Hannibal had crossed twelve years before .

25 . Pompey crossed another pass . He wrote a letter to the — Senate after he had gone to Spain the letter has been preserved

r di s — by Sallust , and is p inted in most e tions of his work and in this letter Pompey says that he has made a new road across the ’ Al u a ps , taking another and more convenient ro te than Hannib l s , i h e iter a liud a l a e H a nniba l nobis o ortun a s . t g , pp In mentioning

i A B CI VI LI B US 1. 10 mak ng of this road ppian says (DE ELLIS , 9) the i that it passed near the sources of the Rhone and Po , wh ch

. 6 . ur were not far apart . Strabo (iv 5) calls them the so ces of two ributa n es ur t of the Rhone and the Po , namely, the D ance and the ’ Dora ; and this is enough to show that Pompey s pass was Mont

G ené v re .

’ m r 26 . 6 B . C. This was in 7 , therefore in Varro s lifeti e , but afte

Pol bios r y was dead , and a few yea s before Strabo and Livy were

iv l . 1 As a . born . ready mentioned in paragraph 5 , Strabo says ( 3) that the Roman road from Spain crossed the Rhone at Tarascon

ur Aix A and bif cated there , one branch going through to ntibes

di rr r n r il on the Me te anean coast , and the othe goi g th ough Cava lon “ and along the Durance to the begi nning of the ascent of the ” A 6 ru lps , 3 Roman miles from Tarascon , thence to Emb n , 99 i ur ri ri m les f ther on , and thence through the B ancon dist ct in

( 1 3 ) 1 mi . r u u 7 les to Cesanne , the first town in Italy The oad m st th s A have crossed the lps by Mont Genevre , and presumably was ’ Pompey s .

l b e i nnin of 27 . Strabo here speaks of the g g the ascent of the

A fin r lps as a de ite point , and eckons distances from there along “

Pol io . . . b s 10 the Roman road y (iii 39 9 , ) likewise treats the A ascent of the lps as a definite point , and reckons distances

’ from there in both di rections along Hanni bal s line of march . Up

r r . . to that point he eckons the distance along the iver, iii 39 9 ;

hi ri and he says that after marc ng along the ver , Hannibal began ” A Al . 0 1 . s ur the ascent of the ps , iii 5 . he makes no f ther mention

r r n u ri r of the ive , the inference is that Han ibal q itted the ve at

Pol bios ut that point ; and if y p the point where Strabo puts it , this woul d mean that Hanni bal turned off up the Verdon valley .

th e u u a In that case act al ascent wo ld not be on the Rom n road , — but on an older track di verging from it at the sixty third mil e

i . stone , the point which Strabo calls the beg nning of the ascent l ul 2 8 . a If Hannib went up the Verdon valley, he wo d be head

’ ing for a pass between Pompey s and the coast- road ; and this is ’ the r in conformity with Varro s statements . I imagine that oute

ul r u l Colrna rs wo d be f om Mirabea to Castel ane and , thence to

’ Barcelonnette and across the Col d Argentié re to Borgo San al B t 1 0 D mazzo near Cuneo . u the di stance would exceed the 5 Roman mil es (1 200 stades) which Polybios has allotted to the “ ” s 1 0 i march from the as cent to Ital y . He compute it as m les

1 r u 1 a day for 5 days , and the e is no doubt abo t the 5 days , as he

( 14 ) B sets them out in detail : see paragraph 3 . ut the usual average coul d hardly be maintained in such wild country amongst hostile

ri a . t bes , and the distance re lly covered may have been much less

i . 0 . 1 u 1 Pol b o . 2 . s 9 y (iii 39 9 , 5 ) p ts the ascent 75 Roman miles 1 00 s n ( 4 stade ) from the crossi g of the Rhone, and says that the 1 r last 1 00 of the 1 75 were marched in 0 days . The fi st 75 woul d ’ answer to Strabo s 63 from the crossing to the point he calls the i n f beg nni g of the ascent , the di ference being that Strabo is reckon ing along the Roman road whereas Polybios woul d be reckoning r h along the iver , and the R one and Durance make a bend between

1 00 Pol bios Tarascon and Cavaillon . The other in y woul d answer

’ to Strabo s 99 to Embru n by the road along the Durance from the beginni ng of the ascent near Mirabeau and Polybios would thus be speaking of some place near Embrun when he says that Hannibal began the ascent after this march of 1 00 miles along the ri ver .

0 . u 3 Hannibal might have q itted the Durance at La Breole, U twelve miles below Embrun , gone up the baye valley to Barce

’ lonnette d Ar entié re and across the Col g to Borgo San Dalmazzo , a di stance of about 80 Roman miles ; or he might have quitted the

i ru Durance at Mont Dauphin , ten m les above Emb n , gone up the Guil val ley to Abri es and across the Col de la Traversette to l di a u 0 . S zzo , a stance of about 7 Roman miles (The Col de la

’ n d Ar Traversette is k own also as the Col de Viso , and the Col

As gentiere as the Col della Maddal ena or Col de Larche . ) the u u O G il is merely a mo ntain torrent , that route is not pen to the

( 15 ) objection that Hannibal woul d still be marching along the ”— river an obj ection that may be urged against the Ubaye and

r u . A u the Ve don ro tes lso , by going on past Embr n to Mont

u r rr r Tric orii Dauphin , Hannibal wo ld be ente ing the te ito y of the ,

r as Livy says he did : see parag aph 39 .

’ 1 Pol bios ri r 3 . y does not give the ve s name when the says that “ al ri r Hannib marched up along the ve as far as the ascent . He but i often mentions the Rhone , does not g ve the name of any

r r Sc Oras r u r other iver in those pa ts except the , a t ib ta y of the

r r Sc 0ra s Rhone . The e is no mention of a iver in any ancient author but Polybios ; and the presumption is that although the name was

r r out . As o used in his time , it afte wa ds went of use the Sa ne had

A ur al two names , rar and Saucona , the D ance might so have two hi Druentia Sc 0ra s . u c on ec names , and T s , of co rse , is merely a j ture ; but it seems to meet the case .

A c Ora s — 2 . S 0 3 pparently, Livy thought the was the Sa ne see — ’ p aragraph 1 7 in which case there is no question of Hannibal s

i r . r march ng up any other iver than the Rhone Livy, howeve ,

s . 1 contradi cts hi mself . He ays (xxi 3 ) that instead of going Al ur straight from the confluence to the ps , Hannibal t ned to the

a d la va m rri r Tri c a stini d left ( ) into the te to y of the , skirte the terri tory of the Voc ontii and entered the territory of the Tric orii . — Whatever their exact boundari es may have been see paragraph

— rri ri 39 these te to es were east of the Rhone , north of the Durance l u I sé re . a ul and so th of the Consequently, Hannib co d not have

( 1 6 ) r ur le t u r r eached them by t ning to the f , nless he was sta ting f om the confluence of the Durance and the Rhone . If he had been as

u o an d the Rh far north as the confl ence of the Sa ne one , or even i u r ht. of the Isere and the Rhone , he m st have turned to the g This shoul d have struck the edi tors who tampered wi th the text of Livy and changed the Saone to the I sé re they shoul d have changed “

u r . it to the D rance , or else put ight for left i . r di 33 Livy ve y often contra cts himself, as he comp led his history out of older histori es and did not always take the trouble

Pol bios hi s u r to reconcile them . y was us al sou ce for this part of Coelius Cinc ius his work , but he quotes on one point and on another, Pol i u xxi . 8 b os 3 , and when he quotes y , he does not always q ote

him accurately .

. 1 R 34 Livy says (xxi . 3 ) that after getting across the hone , al r ul Hannib marched up the other bank towards the inte ior of Ga ,

P 1 r hos whereas olybios says (iii . 47 . ) that he ma ched eastward (

e i ten heo ri - p ) going along the ver up stream . (His phrase for

u - r r p stream here is away from the sea , towa ds the inte ior of

Europe , and in iii . 39 . 9 , speaking of the same march , his phrase ” s ul is towards the source ) Livy has put Ga for Europe , and

r . al omitted eastwa d But if Hannib marched eastward here , he R must have marched along the Durance , not the hone , as the Durance here runs from east to west and the Rhone from north

. Pol io b s iii . . 2 to south No doubt , y says ( 47 , 3) that the Rhone r A had its sources on the no th side of the lps , its sources facing

ros ten hes era n r - 9103 west (p p ) , and that its cou se was south west (1

( 1 7 ) ' ' c men a s d el s ta s hez n s . And a r y ) to the sea as gene al statement , that is true ; but if he had imagined that the Rhone was muning

u - a t r u ul so th west the point where Hannibal c ossed it , he s rely wo d — have sai d that Hannibal marched north - east he would not dis tinguish west and south -west in one sentence and confound east

- and north east in the next . — i s . . . A din Pol b o 8 1 35 ccor g to y (iii 49 5 , 3) Hannibal marched in four days from the crossing of the Rhone to the so- called Island

h Sc 01as — between the R one and the , found a civil war in progress r the e, joined one faction in crushing the other , and obtained u u ri s pplies from the successf l faction as the p ce of his support . u ores r in He m st have crossed the Sc and ma ched to the Island , as Polybios says joined in attacking and expelling -synepitke ‘ — menos kai synekba lon which certainly implies that he di d some h r r t ing more than make a demonst ation from the othe bank .

And r the ur if the crossing over was comp ised in fo days , the march may have been less than 40 Roman miles at 10 Roman

r miles a day : see parag aph 3 . 6 nni m r u r r s 3 . Suppose that Ha bal a ched p f om Ta a con , going first along the Rhone and then al ong the Durance from the con fluenc e of the two : he woul d reach a point just Opposite Cavail lon

r l in 30 Roman miles . The obvious route from Ta ascon to Cavail on

r r u r the is st aight ac oss co nt y, on the line of Roman road ; and in saying so emphaticall y that Hannibal marched along the river ’ ton a ra ton otamon . . 1 r a r au (p p , iii 47 ) or close along the iver (p to l i s i n otamon . . Po b o ma i p , iii 39 9) y y be say ng it to negat ve the

( 18 )

All obri es but i Allobri es the g ; if both the fact ons had been g , as

P l bios ul Livy says , oy wo d have spoken of attacks by the defeated

nl Al lob ri es . faction o y, not attacks by the g , ’ Li s 8 . . 1 . 1 1 v 3 Strabo says (iv ) that in his time , which also was y Allob ri es u m r r time, the g were merely h sband en , though in fo me ages they had armies of tens of thousands in the field . He men n but u tions Vienne, on the Rhone , as their chief tow ; Dion Cassi s

(xlvi . 50) shows clearly that they did not take Vienne till a century ’ — and a half after Hannibal s time the city of Lyons was founded

B C. h All ob ri es in 43 . to house the in abitants of Vienne after the g r And r had d iven them out . other t ibes may likewise have shifted ’ their position between Hannibal s time and the time of Polybios or Ptolemy or Strabo . e n 39 . In d ali g with the part of Provence on the east side of

. 1 0 . 8 the Rhone , Ptolemy (ii 7 , ) fixes Vienne as the city of the Allobri es Se all auni r A g Valence as the city of the g O ange , vignon and Cavaillon as cities of the ; and Vaison as the city of

Voc ontii Tric a s tini Se alla uni the ; and he places the east of the g , north of the Voc ontii and south of the Allobri ges . He does not

the Tric orii . 1 . 1 1 6 . mention , but Strabo says (iv , 5) that the “ Voc ontii r r Tric orii the we e above the Cava i , and the above

i - Voc ontii . Tr c orii u or ru The wo ld thus be east north east of Emb n ,

1 . rr r oc on as he states (iv . 3) that the te ito y of the V tii extended al ong the Durance up to Embrun .

i r r r v . 1 . r 40 . St abo says he e ( 3) that the Roman oad f om Tarascon to the Alps entered the territory of the Voc 0ntii at the

( 20 ) i 6 beginn ng of the ascent, 3 Roman miles from Tarascon and

r r r u u ru the efore nea Mi abea , and q itted it again at Emb n , 99 miles And r al Dur further on . as he says he e that the road ran ong the

1 1 1 6 . ur ance , and elsewhere (iv . . , cf . 3 , 4) treats the D ance as Sal es r r ru the frontier of the y , the oad might be desc ibed as nning — al ong the extreme edge of the territory of the Voc ontii p er — extrema m 07 a m Voc ontiorum agri which is the phrase employed ’

1 for . di by Livy (xxi . 3 ) Hannibal s line of march The stances , the ’ 6 Pol bios 3 and 99 miles , agree with what y says of Hannibal s — — march see paragraph 29 and the digression would not be in ” cluded in the distances he gives as marched along the river , a . 0 . 1 . ul u r iii . 39 9 , 5 It wo d th s appea that after Hannib l had ur r Tri c a stini am t ned to the left into the territo y of the , he c e back r on to the line of what was afte wards the Roman road, and

r followed it from a point near Mirabeau to some point nea Embrun .

1 n 4 . Han ibal must have reached the Island the day after h l t e Po bios i . . Scipio reached crossing of the Rhone , as y says ( 49 1 r , 5) that Hannibal was four days on the march and had sta ted Pol io from the crossing three days before Scipio got there . yb s also says (iii . 49 . that Scipio went back as fast as he came ,

- re off. u embarked his forces and sailed Thus , by abo t the fifth r day after Hannibal eached the Island , there was nothing to hi s r r u A prevent retu ning to the direct o te to the lps , supposing u r i that he had q itted it in orde to avoid a battle w th the Romans . B ut rr r Tri c as tini r l this deviation , into the te ito y of the , may eal y

’ have been part of Hannibal s movements in the civil-wa r in the

( 21 ) for Pol bios . . 1 0 a nl e Island , y says (iii 49 ) that Hannib l not o y join d

but r out. one faction in attacking the other, joined in d iving it

As r . 1 a matte of fact , Livy (xxi 3 ) does not exactly say that Han nibal entered the territory of the Tric a s tini : he merely says in

Tric a stlnos an d rw in Tric orios afte ards says , whereas in the inter ” v enin r r u ri r g wo ds , al eady q oted , he speaks of the ter to y of the Voc ontii and the difference may be more than merely verbal . o n ii r 42 . Supposing that the V c o t had the same bounda ies in

’ ’ ’ m r Liv s ul Hannibal s ti e as in St abo s time and y , Hannibal wo d thus have qui tted their terri tory at Embrun and therefore crossed

. 1 u a the Durance higher up Livy (xxi . 3 ) mentions that the D r nce o happened to be swollen by rains ; and when in flo d , it may be

f u r r . di fic lt to c oss , even in that early po tion of its course But when ” he says that of all the rivers of it was fa r the most di fficul t u ri ur to cross , he m st be thinking of the ver a long way f ther down, u nearer to its confl ence with the Rhone .

— 43 . In the next part of his narrative (xxi . 32 35) Livy says P l i 0— exactly what oyb os says (iii . 5 53) though he says it more i r r r l but Pol bios . 0 . 1 r rheto cal y ; where y (iii 5 ) me ely says the ive , “ ”

2 . Livy (xxi . 3 ) says the Durance

t ri nni ni . u i 44 On q it ing the ver and beg ng the ascent , Han bal r u Allobri es proceeded ve y ca tiously, as he suspected that the g and

t r ul r o he natives wo d attack him . On the fi st day he soon halted , and sent his gui des out for intelligence : on the second day he i for i merely moved nto position a night attack , and had no fight ng

( 22 ) ill r . t the thi d day, when he was going through a gorge He had occupied the summits on the previous night while the natives were off guard ; but the natives made their way al ong the s10pes

- - — and stampeded the caval ry horses and the baggage animal s the natives were too much fri ghtened of the elephants to go anywhere As but near them . the track was not only steep and narrow ran al u r ong a cliff, many of the animals were p shed ove and fell down the cliff ; and the stampede got worse still when Hannibal charged r ill down from the summits to d ive the natives off. He k ed most

Allobri es of the g , and then took the town that served them as a b for urt ase ; and he remained there a day, that is , the fo h day .

r r He made fair prog ess on the next th ee days , but on the fourth day (that is the eighth day of the march) he was attacked again l whi e going through a gorge . The natives were on the higher ground, and rolled rocks down or came down the slopes themselves and

u ai threw stones by hand , ca sing a stampede ag n with heavy loss of animals and men . Hannibal himself got through with the

- u r r advance g ard , but the remainde of the a my spent the whole u night in getting thro gh . The next day (the ninth) he reached the u m s m it of the pass .

45 . There is next to nothing in all this to indicate the route . AS Allobri l i es Po b os iii . 1 for the g whom y ( 5 . 9) mentions in the ul fight at the first gorge , they certainly were not the pop ation of — r 1 0 . the dist ict, as he says (iii . 49 . 3 5 3) they had been following ’ n al r u r r Han ib s a my for a h nd ed miles or more, not da ing to attack it till it was entangled in hill- country where its caval ry could not

( 23 ) . r u A u l act Narrow go ges can be fo nd on any lpine ro te , and a so rocks such as Polybios describes . He says (iii . 53 . 5) that Hannibal

al i e hi r ti lea ko etron h ted on the eighth n ght at a def nsible w te ock , p h “ “ oc ron. B ut r r for y white rock means no more than ba e ock ,

r . hi u he says elsewhe e (x 30 . 5) that the w te rocks themselves co ld m ur be cli bed up by active men ; and , clearly, the colo of the rock l r u di f . wo d make no ference in the climb There is , however , a ock nl r but hi that is not o y ba e w te , the Roche Blanche , on the Little

n r n r St Ber ard , another (nea St Michel) on the Mont Ce is , and othe s

but Pol bios iii . . on other passes , all identified as Hannibal s y ( 53

r far 4 , 5) places the ock at the end of a gorge into which Hannibal l hi s u . u un was led by the treachery of g ides Th s , ess the rock was

th e r r r u i at point whe e he regained the prope o te , it w ll not be found on the main road of any pass at all ; nor will the gorge be

u . fo nd on the main road of any pass , as it was on a deviation Yet people have been identifying gorges on the main roads of the different passes as the very gorge that Hannibal passed through by deviating from the proper route . i al 6 . Pol b os . . 6 r i ui u u u 4 y (iii 53 5 , ) emarks , as someth ng q te n s , that Hanni bal was separated from a large part of hi s force on the

ni r u r eighth ght of the march . If the whole force was b o ght togethe

r h n u r . on the other nights , the ate of marc i g wo ld be ve y slow Pol bio A iii . s 6 . in y ( 5 4) says that , spite of heavy losses in the lps , Hannibal arri ved in Ital y with infantry and caval ry ;

i - r u and such a force , w th elephants and baggage t ain , wo ld form a column of enormous length when marching along a narrow

( 24 ) h track . The head could not advan ce beyond a point w ich the rear could reach on the same day ; and the distance woul d be much — — less than the average march see paragraph 3 which Polybios

as reckons 10 Roman miles a day .

47 . It is about forty Roman miles from the Durance at Mont u um al Da phin to the s mit of the Col de la Traversette, and so about forty from the Durance at La Breole to the summit of the

’ C01 d Ar en ié re r r t . g Hannibal , however, deviated f om whateve i — u P . ol b os . 2 ro te he meant to take . y says (iii 5 3 8) that Hannibal was led into a dangerous gorge by the treachery of some guides

r r he had engaged two days before, thei object being to d aw him into a position where he could be attacked on every side from

l io r ui higher ground . (Poyb s has not made it clea why these g des

: . . 8 . 1 1 u were engaged he states in iii 44 5 , 7 , 4 that g ides had come over from Ital y to meet Hannibal at the crossing of the — v Rhone perhaps they were all killed at the first gorge . ) Li y

. u l i ui (xxi 35) s ggests that instead of fo low ng the g des , Hannibal took a line of his own as soon as he lost confidence in them , and

m er invia lera a e et thus went astray into i passable places , p p q errores r . His actual march must have been a good deal longe than

r his intended oute .

A — . 1 al 48 ccordi ng to Polybios (iii . 54 . 3) Hannib made a speech li i r at the summit of the pass , tel ng the sold ers that the worst pa t of their march was done ; and he not only pointed to the plain of r out but to os the Po sp ead before them , indicated the position ( p )

( 25 ) As r . 00 of Rome itself he was a good 3 miles f om Rome , he could

H ero oto not have managed this unless he used a map . d s says Ari sta oras v . . 1 r ( 49 , 5) that g used a map to show the oute , when he was trying to induce the Spartans to march through Asia

B C. e Minor in 500 . and Hannibal may likewise have us d a map

2 1 8 B . C. . u in , when maps were commoner No doubt , a map co ld but u not be seen except by men who were close by, a speech co ld not be heard by men who were far off ; and I conceive that when Hannibal (or any other general) addressed an army of or

men , he relied on those who heard him to convey the sub

of r im stance his speech to those who coul d not hea . The more

rt ofli c ers u di r po ant wo ld be stan ng near the gene al , and they ul wo d see the map and tell the others about it .

49 . He must at any rate have had a good View of the plain — from the summit of the pass he crossed else his speech woul d be — absurd and there are (I believe) onl y two passes with summits

i P G U . 1 . 2 command ng such a view . In the old AL INE IDE , pt , p 5 , “

. 1 86 all r . ed 3 , B desc ibes the Col de la Traversette To those who approach from the side of , the View suddenly unfolded at n the summit , extendi g (in clear weather) across the entire plain ”

dm r . . of Pie ont as far as Milan , is extremely st iking On p 55 he

’ endorses Bonney s descri ption of the Cenis passes . Between the plateau of the Littl e Mont Cenis and La Grande Croix 'on the ’ r al G eat Mont Cenis'a ridge can be gained by a few minutes w k ,

r u whence is seen the count y to the east of the Po , and the so th of ”

A n . n . the Tanaro , as far as the pen ines He says nothi g (p 57) of

( 26 )

’ r th ur B ut to g aze on e past es down below . three days repairs for were needed to make the track passable the elephants , who — meanwhile nearly died for want of food That is the substance

— - Pol bios . . . 6 8 him of what y says , iii 54 4 7 , 55 , and Livy follows .

52 . The complete skeleton of an elephant was found on the Little St Bernard two centuries ago ; It is mentioned by Saint OI G U P Simon in the preface to his HIST RE DE LA ERRE DES AL ES .

1 r (This was the campaign of 744 , in which he himself took pa t ,

’ out and his book came in He says , On s est encore plus ’ attaché de nos jours a soutenir q u Annibal a dfi passer par l e ’ ’ Petit St Bernard depuis qu on assure qu on a trouvé dans cette

’ ’ montagne tous les ossemens d un un pays qu on appelle dans plusieurs cartes La Grande Route des Romains . But

For Hannibal was not the only person who used elephants .

’ u O 2 - r instance , S etonius mentions (NER , ) that Nero s great g eat great - grandfather went riding about Provence on an elephant

1 2 1 B . C when he was there in . im d . di 53 The landslip that stopped the an als , not stop the ’ men : the infantry went on and reached the plai n in three days

6 6 . 1 r r Pol bios . . . . ma ch f om there . y reckons (iii 53 9 , 54 4 , 55 , 5 ) um that Hannibal reached the s mit on the ninth day, camped for there two days , that is , the tenth and eleventh , came to the landslip the next day, that is the twelfth , and reached the plain

r : hi on the third day from the e , that is , the fifteenth day w ch iii i fif r 6 . al ag ees with his statement ( . 5 3) that Hann b took teen r n r ai r the days to cross the Al ps . The cavalry we e left behi d to ep

( 28 ) r for Pol bios . . 8 t ack the elephants , y saying (iii 55 ) that it was

ai r the u di u l for G r rep ed by N mi ans , which is his us a name a tha

lr gini an cava y . A l r . alm 54 lands ip might occur on any pass , and ost eve y pass has places where the track is steep and narrow and running along

i — ri B Pol b os . . . ut a cliff, just as y (iii 54 5 7 ) desc bes the track here there are exceptional ly bad places just below the summits of the — Col du Clapier and of the Col de la Traversette the Clapier is

’ l Esc alier also known as , the descent being as steep as a staircase

r 000 r for the fi st 4 feet, and the Trave sette takes its name from a An tunnel buil t in 1 480 to avoid a precipitous bit near the top . d u i both these s mmits command a wide v ew of the plain . From each of these two summits the di stance to the plain is about 30 English — (or 33 Roman) miles measuring to Avi gliana in the one case and u the — hi u l ur to Sal zzo in other and t s wo ld fu ly occupy fo days , as there was a landslip to be passed and in the earlier part of the i — Pol b os . . . 1 descent there was the snow . y says (iii 54 5 , 55 5) that the new autumn snow was lying loose on the old winter snow which now was frozen hard ; and men and animal s slipped and rf fell , when they trod through the new snow on to the icy su ace of the old .

r 55 . From the summits of the othe passes the distance to the u 0 i Ar ur plain is abo t 4 Engl sh miles for the gentiere , meas ing to

r San m for r A l an Bo go Dal azzo ; about 45 the Genev e , to vig i a ; u for m Ar u 60 for r abo t 55 the Si plon , to ona ; abo t the G eat St

( 29 ) r r an d for r ar al Be nard , to Iv ea ; about 75 the Little St Be n d, so to Ivrea . These 75 miles (more than 80 Roman miles) woul d not be marched in the four days ; and even if the 55 or 60 could be im ul diffi u u marched in that t e , there wo d be a c lty abo t the total di ol bios 0 r r stance marched . P y makes it 26 0 stades f om the c ossing 8 u of the Rhone to Italy, and he reckons stades to the mile, th s r 2 or 00 : a riving at a total of 3 5 Roman miles , 3 English see para

2 . or 60 nl 2 0 or 2 graphs , 3 Subtracting the 55 , this leaves o y 4 45 English miles from the crossing of the Rhone to the summits of r m u hl 100 1 0 the Great St Be nard and the Si plon , or ro g y and 5 A i . nd miles too l ttle to reach either of those passes , Hannibal would have to march along the lake of Geneva from one end to

l ios r . P b di o dl . the othe , a stance of 45 miles y says repeate y (iii

. 1 0 . I r r 39 9 , 47 . , 5 ) that Hannibal ma ched along a iver, but says nothing of hi s marching al ong a lake . l . r o 6 A v xxi . 8 u va 5 cco ding to Li y, 3 , most people tho ght ( g c r e B eder ) that Hannibal had crossed the Great St Bernard . ut this was not unreasonable if they thought that he was coming And i u : 1 . from Lyons , as L vy and Sili s say see paragraph 4 if Pol bios u a y had not specified the distances , it might be arg ed th t

u r Sc ora s u or the trib ta y which he calls the really was the Sa cona ,

’ Saone ; and that when he speaks of Hannibal s marching along a

r r r B ut n al h all u . ive , he means the R one th o gh if Han ib had got fa r u ul r m r as as that , he s rely wo d have c ossed the Si plon rathe

r r r u ri n him than the G eat St Be na d , as the Simplon wo ld b g down

in rri r I ns ubres ain . to the te to y of the , his allies ag st Rome

( 30 ) r r 57 . Livy ejects the Little St Be nard as well as the Great on the ground that it woul d likewise bring Hanni bal down into the territory of the Libici and it cannot be the pas s of which Polybios i speaks , as the summit has no V ew towards the plain , and is too

’ fa r away for Hannibal to reach the plain in four days march from

B u xxi . Coelius there . t Livy ( 38) quotes as saying that Hannibal

liu P l s . Coe s r o bio went that way was a contempora y of y , though u r r r r DI VI NATI ONE perhaps a yo nge man ; and Cice o ema ks (DE , i .

2 Coelius r n Silenos 4) that copied from the w iti gs of , a Greek who

oelius ul i . C ru was w th Hannibal That being so , sho d be as t stworthy l Po bios . r as y himself , yet contradicts him here Livy , howeve , may be quoting Coelius quite correctly as saying that Hannibal — crossed the Alps p er Cremonis j a ga m a h expression that does not — occur elsewhere but may be wrong in thinkin g that Coelius k thereby meant the Little St Bernard . He may be ma ing a mis

r uri take that Strabo made . There were two ivers called D a in that

— mas d a s— r i . 1 6 1 1 8 Du a pa t of Italy Pl ny , iii , now distin

uished i r r g as the Dora R pa ia , which ises on the Mont Genevre

r B a ltea pass and joins the Po near Tu in , and the Dora , which ri fiv e - ses on the Little St Bernard pass , and joins the Po and

r n . 6 . r r twenty miles below Tu i . Strabo (iv 5) makes these two ive s

r i r u into one , with the sou ce of the Dora R pa ia and the co rse of the Dora B altea . Livy may have made the same mistake and fancied that Hannibal would go down the Dora B altea into the

rri r Coelius r te to y of the Libici , when eally meant the pass at the

ur r so ce of the Dora Ripa ia .

( 3 1 ) ’ r B u r r to g aze on the pastures down below . t th ee days epairs r for we e needed to make the track passable the elephants , who — meanwhile nearly di ed for want of food That is the substance — — P ios . o . l b iii . 6 8 him of what y says , 54 4 7 , 55 , and Livy follows .

52 . The complete skeleton of an elephant was found on the r Little St Berna d two centuri es ago . It is mentioned by Saint m O G U RE P Si on in the preface to his HIST IRE DE LA ER DES AL ES .

1 m r (This was the campaign of 744 , in which he hi self took pa t ,

’ u and his book came out in He says , On s est encore pl s ’ attaché de nos jours a souteni r q u Anniba l a dfi passer par l e ’ ’ Petit St Bernard depuis qu on assure qu on a trouvé dans cette

’ ’ montagne tous les ossemens d un umpays qu on u B appell e dans plusie rs cartes La Grande Route des Romains . ut

e For Hannibal was not th onl y person who used elephants .

’ i u O 2 r —r instance , Sueton s mentions (NER , ) that Nero s g eat g eat great - grandfather went riding about Provence on an elephant in 1 1 B C. when he was there 2 . a im di d 53 . The landslip that stopped the n als , not stop the ’ men : the infantry went on and reached the plain in three days

Pol bios r . . . . 6 6 . 1 march from there . y eckons (iii 53 9 , 54 4 , 55 , 5 ) al u that Hannib reached the s mmit on the ninth day, camped there for two days , that is , the tenth and eleventh , came to the r i landslip the next day, that is the twelfth , and eached the pla n fif : on the third day from there , that is , the teenth day which f r iii . al ag ees with his statement ( . 56 3) that Hannib took fi teen

r r in r ai r the days to cross the Alps . The cava l y we e left beh d to ep

( 28 ) P l i r o b os . . 8 t ack for the elephants , y saying (iii 55 ) that it was ir u di hi u ual Ca rtha repa ed by the N mi ans , w ch is his s name for ni gi an caval ry .

A r . u al m r 54 landslip might occ on any pass , and ost eve y pass has places where the track is steep and narrow and running along — f Pol bios . . r r . a cli f, just as y (iii 54 5 7 ) desc ibes the track he e But there are exceptionall y bad places j ust below the summits of the — Col du Clapier and of the Col de la Traversette the Clapier is

’ al l Esc a lier so known as , the descent being as steep as a staircase for r 000 r the fi st 4 feet , and the Traversette takes its name f om a 0 And tunnel built in 148 to avoid a precipitous bit near the top .

r both these summits command a wide view of the plain . F om each of these two summits the di stance to the plain is about 30 English — (or 33 Roman) miles measuring to Avigliana in the one case and u r— u u u u to Sal zzo in the othe and this wo ld f lly occ py fo r days , as there was a landslip to be passed and in the earlier part of the — P l ios . . . . o b 1 descent there was the snow y says (iii 54 5 , 55 5) that the new autumn snow was lying loose on the old winter snow which now was frozen hard ; and men and animals slipped and r ur fell , when they t od through the new snow on to the icy s face of the old .

r 55 . F om the summits of the other passes the di stance to the u 0 for A ur plain is abo t 4 English miles the rgentiere , meas ing to

r l u for r Av i Bo go San Da mazzo ; abo t 45 the Genev e, to igl ana ; u for m A u 60 r abo t 55 the Si plon , to rona ; abo t for the G eat St

( 29 ) r r r and u for r r al Be na d, to Iv ea ; abo t 75 the Little St Be na d , so r r ul to Iv ea . These 75 miles (mo e than 80 Roman miles) wo d not be marched in the four days ; and even if the 55 or 60 could be m r ul iffi ul u marched in that ti e , the e wo d be a d c ty abo t the total

ol i s r i distance marched . P yb o makes it 2 600 stades from the c oss ng h 8 u of the R one to Italy, and he reckons stades to the mile, th s rr 2 00 : r a iving at a total of 3 5 Roman miles , or 3 English see pa a

or 0 0 2 . u r 6 2 2 graphs , 3 S bt acting the 55 , this leaves only 4 or 45 English miles from the crossing of the Rhone to the summits of

r r u 1 00 1 0 the Great St Be nard and the Simplon, or o ghly and 5

An r . d miles too little to each either of those passes , Hannibal would have to march along the lake of Geneva from one end to l i Po b os . di . the other, a stance of 45 miles y says repeatedly (iii

. i r t . 1 0 . 1 r bu 39 9 , 47 , 5 ) that Hannibal marched along a ve , says hi r nothing of s ma ching along a lake . A u va l o 6 . di . 8 5 ccor ng to Livy , xxi 3 , most people tho ght ( g c redere) that Hannibal had crossed the Great St Bernard . B ut this was not unreasonable if they thought that he was coming A u : r 1 . nd from Lyons , as Livy and Sili s say see parag aph 4 if Pol bios u th y had not specified the distances , it might be arg ed at

r r Sc oras r u or the t ibuta y which he calls the eally was the Sa cona , ’ Saone ; and that when he speaks of Hannibal s marching along a

r r r u . B ut ive , he means the Rhone all th o gh if Hannibal had got u u r m r r as far as that , he s rely wo ld have c ossed the Si plon athe r r r im ul r him w than the G eat St Be na d , as the S plon wo d b ing do n

i rr r I nsubres . nto the te ito y of the , his allies against Rome

( 30 )

58 . There is also an ambiguity in the statement of Nepos B o I a r Al . n r (HANN AL , 3) that H nnibal crossed the G aian ps This mal ly woul d mean crossing by the Little St Bernard ; but it might

r iii . 1 . 6 also mean crossing by the Mont Genev e , as Ptolemy ( 35 , 3 ) Al r makes the Grai an ps include B iancon and Embrun .

66 i r r r O I'E . (HIST R , ii ) speaks of a leg on ma ching f om Tu in across Al r i the Graian ps , clearly meaning the Genev e or Cen s passes , as

the Little St Bernard is not accessible from Turi n . r u u 59 . Se vi s q otes Varro as mentioning five passes through the ’ ’ ’ A - r ru lps , the coast road , Hannibal s oad , Pompey s road , Hasd bal s ai A u road , and the road through the Gr an lps ; th s making Has

drubal ni al r f r r r and Han b c oss di ferent passes , though othe w ite s

r : r r 62 . u make them c oss the same pass see pa ag aph Th s , if r iu rr Se v s quotes co ectly, Varro seems to have assigned the Cenis

ru al . r r r u route to Hasd b This route , or the St Be na d outes , wo ld suit an army coming from Lyons ; and Hasdrubal most probably A A P 2 h di . I I 8 came from t at rection ccording to ppian (H S AN A , ) Hasdrubal came round the north - west corner of the Pyrenees

— - Hannibal came round the south eas t corner and marched al ong : l — Po bios iii . . 8 1 . ru u the coast see y , 39 7 , , 4 7 and Hasd bal m st

have kept a long way from the coast , as Livy (xxvii . 39) speaks

n Arv emi rr r l of his meeti g the , whose te ito y answered rough y to

Auvergne . i 60 . Possibly, and I th nk probably, some of the ancients con

u s of r rm u fused the ro te these two Ca thaginian a ies , and th s made Hannibal go to Lyons when in reality it was onl y Hasdrubal who

( 3 2 ) ’ r And u h u mi r went the e . s c conf sion ght easily a ise, as Hannibal s

’ was r ru r march celeb ated , whilst Hasd bal s was fo gotten or ignored

ur ri u . which is not s p sing, as his march had no res lts It was Tra sirnene and Cannae and their other di sasters that made the ’ r r ni r as Romans emembe Han bal s march so well , whe e they might ill nearly have forgotten it, if he had been defeated and k ed at the

Trebia dru al uru di , as Has b was at the Meta s , imme ately on enter ing Italy . i ’ 6 1 . ru . In speak ng of Hasd bal s march , Livy merely says (xxvii

r i Ar r r 39) that afte meet ng the ve ni , he met other Gallic t ibes and A ri n lpine t bes , and crossed the pass that Han ibal had opened up .

Ar r Au u If he met the ve ni in vergne , he might take some such ro te

. 1 n Al as Livy (xxi 3 ) has assig ed to Hannibal , first meeting the lo bri es Tric a stini Voc ontii Tric ori i g , then the , the and the , and then crossi ng the Durance on his way to the pass ; and Livy may have r al had this route in mind when he was w iting of Hannib , though he excludes it by hi s statement that Hanni bal turned to the left to reach the terri tory of the Tric astini .

al 62 . Livy says most distinctly that Hasdrub used the pass

l ler e u . er ma n ta a a which Hannibal had opened p , xxvii 39 , p p q

' tra nsita ra m e i vi a e A B s ua a nt a n a rant. 2 f , q f ppian (HANNI AL , 5 ) i in r hodo oiémena roteron h oAnn boa . says the same thing G eek , p p yp ru r Thus , if Hasd bal went over the Little St Be nard or Mont Cenis , u u Hannibal m st have gone that way , and Livy m st be wrong in

r ur r r Tri c orii saying that he c ossed the D ance in the ter ito y of the , ru —he u that is , above Emb n wo ld not cross it there , except in

( 33 ) f u r r going to passes urther so th than Mont Genevre , as the ive

ri . ses on that pass On the other hand , if Hannibal crossed it there ,

ru and Hasd bal went over the Little St Bernard or Mont Cenis , Livy (and Appian also) must be wrong in saying that they both

’ used the same pass ; and there is Varro s statement (though only at second hand) that they used di fferent passes : see para graph 24 .

6 . r ul u ru 3 The northe n passes wo d best s it Hasd bal , coming

r A f om the Bay of Biscay and uvergne , whereas the southern

ul r Mediter passes wo d be bette for Hannibal, coming from the ra nea n coast . Hasdrubal might gain by going to a southern pass i ul i that Hann bal had opened up , but Hannibal wo d gain noth ng

r . And by going to a northe n pass if he went up the Rhone to Lyons ,

Pol bios as Livy and Silius say, there is no sense in what y says

r r about the di stance ma ched . To each Lyons he would have to — average 25 Roman miles a day inste ad of his usual 1 0 miles see — paragraph 1 2 a nd he woul d then have to march 1 00 Roman — miles al ong the Rhone to the ascent of the Al ps see paragraph 2 — 1 00 ul r 9 and this miles wo d b ing him to Geneva , whereas the 0 Y u distance is only 5 Roman miles to enne , where he pres mably would quit the Rhone if he were making for the Little St Bernard or Mont Cenis .

r f ri r 64 . The e is no di ficulty about the di stance if the ve was the i . 2 r Durance , as Livy states explicitly in xxi 3 and ce tainly impl es

hi . 1 . 1 s 6 in xxi 3 . Excepting mention of the Sa ne in xxi 3 , his statements are consistent with the statements of Polybios as to

( 34 ) ’ — Hannibal s route he onl y supplements Polybios by quoting some one else about the Voc ontii and other tri bes and the crossing of ’ u Pol bios the Durance . No do bt, y says nothing of Hannibal s crossing any ri ver but the Rhone ; but he implies that Hannibal

r i — r r u r crossed a tributa y of t see pa agraph 35 . If the t ib ta y was u r the D rance , he must have c ossed it a second time a good deal

’ l bi s Liv s r . Po o further up , if y statement is co rect y might ignore u r r r a second crossing j st as he igno es the first , but Livy ep esents

r f r r this second c ossing as an operation of great di ficulty, the ive r in being then swollen by ain . His description , however , seems — — applicable to the Durance so far up see paragraph 42 and I suspect that he was copying a description of the Durance much further down its course .

Pol bio . r ffi u s 65 To ecapitulate al l this . The main di c lty is that y

r u r R Sc 01as r speaks of a t ib ta y of the hone as the , and no othe

u r r r o ancient a tho speaks of any ri ver of that name . The ive Sa ne

m u A — r 1 — had two na es , Sa cona and rar see pa agraph 7 and I con ec ture ri al Druentia j that the ver Durance so had two names , and

Sc Ora s Sc Ora s u Pol bios , and that the name of was c rrent when y

but . a re : wrote , obsolete when Livy wrote My reasons ”

Pol bios r . W 0 . 1 r 2 hile y (iii 5 ) me ely says the iver , Livy (xxi 3 ) says the Durance and in this part of his narrative he is copying Pol bios y almost word for word .

’ Polybios gi ves dista nces for Hannibal s march along the river which are curiously like the distances that Strabo gi ves for the

( 35 ) Roman road along the Durance from Taras con to the Alps : see

r parag aph 29 . Polybios says that Hanni bal marched eastward from the cross

- in ri u . As g of the Rhone , following the ver p stream the crossing was at Tarascon (or close by) an eas tward march woul d carry him

hi ru : along the Durance , w ch here ns from east to west see para graph 34 Livy says that on leaving the so- called Island at the confluence

ri u ur le t of the t b tary and the Rhone , Hannibal t ned to the f into the terri tory of the Tri c as tini instead of going straight on to the Alps ; and Hannibal coul d not have done this unless he was on the route along the Durance to the Al ps . He would turn to the right to reach their terri tory if he were going along the I sé re or

r r any other iver furthe north ; see paragraph 32 .

’ Li s 66 . r v On the othe hand , there is y statement that the

r r ri o di hi s t ibuta y ver was the Sa ne , contra cting own statement ’ ur n al r di i about Hannibal s t ni g to the left , and so cont a ct ng the

’ statements of Polybios about the length of Hannibal s march . But why shoul d Liv y contradict himself and al so contradict

Pol bios ' u y , whom he usually follows word for word My s ggestion “ ” is that he took Sc 0ra s in Polybios for Saucona or a variant of ri r o i u r that name , and called the t buta y the Sa ne w tho t conside ing i i ru al . r ul r what that implied The e ror wo d not be st k ng, if Hasd b u u i went that way, as people wo ld confuse his ro te w th Hanni

’ bal s .

i r 67 . There s really nothing in the a gument that Hannibal

( 36 ) would not venture to march along the Durance as the Romans r might attack him on the flank . In the first place , the e were no

As Romans there . soon as Scipio found he was too late to defend

r r re the c ossing of the Rhone , he went st aight back to the coast ,

r : for r embarked his a my, and sailed off he was not ashore mo e — r 1 . than about a week see parag aph 4 In the second place , ri Hannibal had no reason to fear the Romans . He was in supe or u ru as ru force , and co ld have c shed Scipio then as easily he c shed

Trebia but di d him at the two months later on , he not want to u ri u fight j st then , as he saw that victo es in Provence wo ld not produce the same effect at Rome as Victories in Ital y itself : see paragraph 13 . u 68 . Supposing that Hannibal followed the D rance to some

ru r point near Emb n , one wonde s why he did not follow it to its B u u source at Mont Genevre and cross that easy pass . t the s mmit il r Pol bios of the pass is less than ten m es from the iver, and y says that Hannibal took nine days in gorng from the river to the sum — u . mit Livy says nine days from the D rance Nine days , I think , a re not inconsistent with the distance from the Durance at Mont u r or ur Da phin to the Col de la Trave sette, from the D ance at ’ i r d Ar entié re r r 6 . La B eole to the Col g see pa ag aphs 4 , 47 I th nk u Mont Da phin the more likely point , partly because Hannibal would have to go through “ the territory of the Tricorn (as Livy

di d r r says he ) in orde to reach it , but not to reach La B eole ; and partly because there is only a mountain torrent (the Guil) coming r but ri Ub i down the valley the e , a ver (the aye) com ng down the

( 37 ) a t r Pol bios m r r a r valley La B eole , and y ight ega d ma ch along the ’ Ubaye as a march al ong the ri ver. l um 6 . W r was an a 9 hateve pass it that H nib crossed , the s mit u m st have commanded a wide View of the plain of the Po , else he r u would not have made his speech there . The e is s ch a view from but u the Col de la Traversette , not from the s mmit of the Mont

’ Genevre pass or from the Col d Argentié re ; or in fact from any other pass southward of the Little Mont Cenis and the adjacent d l u . And u Col Clapier those two passes are n ikely, as Hannibal had no motive for going so far north . a re nl i 70 . The points that I have mentioned hitherto o y deta ls u ai f n of the ro te, and are subordinate to one m n point a fecti g the u an d r entire route . The aut mn was advancing ; it was a matte of life or death for Hannibal to complete his march before the snows ul r r had made the Alps impassable . He wo d therefo e take the ve y shortest route with no more digression than was really needed for

’ getting supplies or avoiding Scipio s army duri ng the week it was An r rs ashore . d the shortest routes a re by the Col de la T ave ette ’ d Ar entié re and the Col g .

1 . t r ai 7 On questions of his sort no ce t nty can be attained , but I think the balance of probability inclines to some such route as I r sugges t for Hannibal . The earliest Roman oad from Spain to Ital y crossed the Rhone at Tarascon and crossed the Durance at or i l l ur u ( near) Cava l on , and then fo lowed the D ance to its so rce

r h on the pass of Mont Genev e . I t ink that Hannibal took this

( 38 )

n r A r Pol bios u comi g f om f ica, yet y speaks of the maho ts as ’

di . 0 . 1 . 6 . xi . 1 . 1 2 rm In ans (i 4 5 , iii 4 7 , ) not only in Hannibal s a y

but r r r . r in othe a mies coming from the e The elephants , howeve , are clearly Afri can (as shown by their large ears) on the coins the Carthagi nia ns struck at Cartagena while they were in possession

of i . r i are Spa n The elephant he e , and the other at the beg nning, taken from two of thes e coins which a re nearly contemporary i w th Hannibal .

CAMB RI D E RI NTE D B L WI AT TH UNI I RE G : P Y W. E S E VE RS TY P SS