HANNIBAL'S  MARCH 


SPENSER  WILKINSON 


/ 
Price  Seven  SUlUngs  and  Sixpence  net 


/ 


\\  ■ 


OXFORD 

AT  THE  CLARENDON  PRESS 
1911 


vtf 


HANNIBAL'S  MARCH 
THROUGH  THE  ALPS 


BY 

SPENSER  WILKINSON 

CHICHELE     PROFESSOR    OF    MILITARY     HISTORy 


WITH  TWO  FIGURES  AND  FOUR  MAPS 


OXFORD 

AT  THE   CLARENDON    PRESS 

1911 


HENRY  FROWDE,  M.A. 

PTJBLISHEE  TO   THE   UNIVEESITY   OF   OXFORD 

LONDON,   EDINBTIEIGH,   NEW   YORK 

TORONTO   AND  MELBOURNE 


HANNIBAL'S  MARCH  THROUGH 

THE  ALPS 


THE  PROBLEM  AND  THE  CLUE  TO  ITS  SOLUTION 

From  the  age  of  Augustus  until  the  present  day 
historians  have  been  unable  to  agree  with  one 
another  concerning  the  route  which  was  followed 
by  Hannibal  in  the  year  218  B.C.,  in  his  march 
from  the  P3n['enees  through  Southern  France  and 
the  Alps  to  the  plain  of  the  Po.  The  purpose  of 
this  essay  is  to  show  that,  chiefly  in  consequence 
of  the  researches  of  a  group  of  French  officers, 
that  route  can  now  be  traced  as  regards  its  main 
points  with  reasonable  certainty,  and  as  regards 
the  incidents  of  the  march,  with  a  fair  degree  of 
probability.  The  sources  of  our  information  are 
the  third  book  of  the  History  of  Polybius  and  the 
twenty-first  book  of  that  of  Livy.  Polybius  was 
born  at  Megalopolis  in  Arcadia  between  208  and 
198  B.  c.^  In  166  or  167  b.  c.  he  was  taken  as  a  hos- 
tage to  Rome,  where  he  remained  continuously  for 
seventeen  years.  He  was  then  at  liberty  to  return 
to  Greece  and  to  travel,  but  it  was  in  Roman 
rather  than  in  Greek  society  that  he  was  thence- 
forth at  home.     He  had  not  been  long  at  Rome 

1  208,  Mommsen,  R.  G.,  ii.  448 ;   198,  Polybius  und  Sein  Werk, 
von  0.  Cuntz,  p.  77. 


4  THE  PROBLEM  AND  THE 

before  he  acquired  the  friendship  of  the  Scipio 
who  afterwards  became  famous  as  the  conqueror 
of  Carthage  in  the  third  Punic  War.  Polybius  had 
received  the  education  of  a  statesman  and  a  soldier, 
and  his  intimacy  with  Scipio  led  to  his  accompany- 
ing that  general  in  the  campaigns  against  Carthage 
and  against  Numantia,  in  which  it  is  believed  that 
he  rendered  some  service  as  an  engineer.  He 
wrote  his  history  for  the  purpose  of  explaining  to 
his  Greek  countrymen  how  Rome  had  come  to  be 
the  mistress  of  the  world.  In  pursuit  of  the 
information  which  he  required  he  travelled  widely, 
and  in  particular  made  the  journey  across  the 
Alps  in  order  to  follow  and  to  see  for  himself  the 
route  that  had  been  taken  by  Hannibal.  He  says 
that  he  had  had  the  opportunity  of  inquiring 
about  the  events  of  the  march  from  men  who  had 
taken  part  in  it.  His  residence  in  Rome  began 
half  a  century  after  the  march,  and  he  was  then 
between  thirty-one  and  forty-one  years  old.  He 
may  very  well  have  had  the  opportunity  of  dis- 
cussing the  events  of  218  with  Carthaginians  who 
had  accompanied  Hannibal,  with  Greek  officers 
who  had  been  on  his  staff,  and  with  Romans  who 
had  taken  part  in  the  operations  against  him. 
As  he  wrote  he  had  before  him  the  accounts  of 
earlier  authors,  some  of  whom  were  Roman,  and 
one  probably  a  Greek.  In  point  of  time  he  stood 
to  the  events  of  218  as  a  man  who  was  born  in  1871 
would  stand  to  the  Crimean  War. 

Livy  wrote  a  century  and  a  quarter  after  Polybius. 


CLUE  TO  ITS  SOLUTION  5 

He  was  a  literary  historian  who  sat  in  his  study 
and  compiled.  He  lacked  the  contact  with  men 
and  with  affairs,  and  the  habits  of  travel  and 
observation,  which  specially  qualified  Polybius 
to  be  a  military  historian.  His  practice  was  to 
translate  or  paraphrase  the  works  of  his  pre- 
decessors, relying  for  each  period  mainly  upon 
some  writer  whom  he  regarded  as  for  that  period 
the  principal  source,  and  to  interpolate  and  incor- 
porate into  his  story  such  passages  from  other 
writers  as  seemed  to  him  to  supplement  the 
narrative  or  to  be  of  special  interest.  The  inter- 
polated passages  are  not  always  consistent  with 
the  main  text,  and  are  thought  sometimes  to  have 
been  inserted  at  points  where  they  do  not  properly 
belong.  One  of  the  chief  difficulties  which  besets 
the  interpreter  of  Livy  is  this  habit  of  compiling 
what  purports  to  be  a  continuous  story  out  of 
passages  which  have  not  always  been  judiciously 
fitted  together.  In  his  account  of  Hannibal's 
march,  Livy's  main  story  so  closely  corresponds 
with  that  of  Polybius  that  it  might  be  taken  for 
a  free  translation  from  the  text  of  the  Greek  his- 
torian. But  here  and  there  occurs  a  passage  not 
to  be  found  in  Polybius,  and  one  of  these  passages 
is  hardly  capable,  as  it  stands,  of  reconciliation 
either  with  that  author  or  with  the  general  trend 
of  Livy's  own  narrative. 

The  attempt  will  here  be  made  to  show  that  the 
text  of  Polybius  indicates  a  specific  route,  and  that 
the  acceptance  of  this  route  leads  to  a  simple  and 


6  THE  PROBLEM  AND  THE 

probable  explanation  of  the  apparent  discrepancies 
between  the  text  of  Polybius  and  that  of  Livy. 
The  right  way  to  begin  is  to  take  those  points  as 
to  which  the  two  historians  upon  whom  we  have 
to  rely  are  in  absolute  accord,  and  if  in  regard  to 
certain  main  points  of  the  route  we  can  thus  reach 
reasonable  certainty,  we  shall  gain  an  outline 
which  it  may  be  possible  to  some  extent  to  fill  in. 
Polybius  (xxxiv.  10),  describing  the  Alps,  says 
that  the  pass  used  by  Hannibal  led  to  the  territory 
of  the  Taurini,^  and  the  text  of  his  story  leaves  no 
doubt  that  Hannibal's  descent  from  the  pass 
brought  him  directly  into  the  territory  of  that 
tribe.  Livy  says  that  the  Taurini  were  the  first 
tribe  that  Hannibal  met  when  he  came  down  into 
Italy.  The  district  of  the  Taurini  is  represented 
by  the  modern  city  of  Turin.  We  may  therefore 
take  the  neighbourhood  of  Turin  as  a  point  fixed 
by  the  agreement  of  Livy  and  Polybius,  and  may 
work  back  from  that  fixed  point.  The  plain  of  the 
Po  in  the  neighbourhood  of  Turin  was  reached  in 
three  days'  march  from  the  great  pass  over  the 
Alps,  of  which  a  very  striking  description  is  given. 
We  are  told  that  the  army,  consisting  at  this  time 
of  not  less  than  26,000  men,  was  halted  for  two 
days  at  the  summit   of  the  pass,  and  Polybius 

1  In  iii.  56,  Polybius  says,  '  he  came  down  boldly  to  the  plain 
of  the  Po  and  the  tribe  of  the  Insubres.'  This  is  one  of  the 
preliminary  summaries  of  which  Polybius  is  fond.  Chap.  Ix, 
however,  shows  that  the  first  tribe  he  reached  were  the  Taurini. 
The  Insubres  lay  to  the  east  of  the  Ticino  about  Milan. 


CLUE  TO  ITS  SOLUTION  V 

says  that  as  the  men  were  greatly  disheartened 
'  Hannibal  assembled  them  and  tried  to  cheer 
them  up.  The  one  thing  that  he  had  to  rely  on 
for  this  purpose  was  the  clear  view  of  Italy,  for 
Italy  lay  at  the  foot  of  the  mountains  in  such 
fashion  that  when  both  were  seen  in  one  view  the 
Alps  appeared  to  constitute  an  Acropolis  of  all 
Italy.  Therefore  showing  his  men  the  plains  of 
the  Po,  reminding  them  of  the  goodwill  of  the 
Gauls  who  inhabited  those  plains,  and  also  indi- 
cating the  position  of  Rome  herself,  he  put  them 
for  the  time  into  good  spirits.' 

Livy  says  that  when  a  start  was  made  from  the 
pass  at  dawn  '  Hannibal  called  a  halt  on  a  pro- 
montory from  which  the  view  extended  far  and 
wide,  and  showed  his  soldiers  Italy  and  the  plains 
of  the  Po  lying  at  the  foot  of  the  Alps.  He  told 
them  that  they  were  then  crossing  the  battlements 
not  only  of  Italy  but  of  the  city  of  Rome.'  Thus 
Polybius  and  Livy  are  in  accord  in  representing 
Hannibal  as  crossing  the  Alps  by  a  pass  from  which 
the  plain  of  the  Po  near  Turin  could  be  seen  by 
a  large  number  of  men  at  the  same  time,  and  from 
which  the  view  conveyed  the  impression  that  the 
Alps  were  the  Acropolis  of  Italy.  Until  1884  the 
historians  were  agreed  in  rejecting  these  passages 
as  mythical.  From  none  of  the  passes  which  they 
discussed  is  it  possible  to  see  the  plain  of  Italy, 
though  the  advocates  of  the  Mont  Cenis  route 
suggest  that  a  glimpse  of  the  plain  can  be  seen 
by  the  tourist  who  undertakes  a  laborious  ascent 


8  THE  PROBLEM  AND  THE 

up  the  side  of  one  of  the  mountains  which  border 
the  pass.^  In  1884  Colonel  Perrin  published  his 
account  of  the  Col  du  Clapier,  from  which  in  clear 
weather  are  visible  the  plain  of  Turin,  the  whole 
range  of  the  hills  of  Montferrat  behind  Turin,  and 
a  great  sweep  of  the  plain  to  north  and  south  of 
that  range.  It  is  just  such  a  view  as  would  suggest 
the  words  quoted  from  Polybius,  and  the  impression 
that  the  Alps  form  the  Acropolis  of  Italy  is  pre- 
cisely the  description  which  a  Greek  would  be 
likely  to  give  either  of  the  view  from  the  Clapier 
looking  towards  Turin,  or  of  the  view  from  Turin 
looking  towards  the  Clapier.  Both  of  these  views 
Polybius  must  have  seen,  and  that  is  probably 
why  he  used  the  expression  '  when  the  Alps  and 
Italy  are  seen  in  one  view '.  A  rough  idea  of 
the  view  from  the  Clapier  is  conveyed  by  Fig.  1, 
and  a  rough  idea  of  the  view  looking  from  Turin 
towards  the  Clapier  by  Fig.  2.^ 

1  My  own  observation  and  that  of  my  companions  has  failed 
to  find  a  point  on  or  near  the  Mont  Cenis  route  from  which  the 
plain  of  Italy  can  be  seen. 

2  It  is  not  practicable  to  take  a  photograph  from  the  Clapier 
which  will  include  both  the  Alps  and  the  plain.  I  have  made 
three  visits  to  the  pass,  and  on  each  occasion  one  at  least  of  the 
party  has  carried  a  hand  camera.  A  telephotographic  lens  is  not 
permitted  on  the  pass.  In  August,  1909,  we  had  a  magnificent 
view,  but  the  plates  taken  record  only  a  fraction  of  it.  They 
dimly  show  the  Monte  Musine  on  the  left  side  of  the  valley  of  the 
Dora  Riparia,  where  that  river  enters  the  plain,  and  the  sanctuary 
of  San  Michele  on  the  opposite  side  of  the  valley.  But  as  regards 
all  beyond  those  points  the  plates  are  a  blank.  Monte  Musine  is 
twenty-six  miles  from  where  the  camera  stood,  and  this  seems 
to  be  the  limit  of  its  power.    The  Superga  which  was  visible  above 


Fii^.  1 


Fig.  2 


p.  8 


CLUE  TO  ITS  SOLUTION  9 

The  Col  du  Clapier  is  an  almost  level  depression 
between  two  peaks,  its  floor  being  about  half 
a  mile  wide  and  two  or  three  miles  long.  At  the 
Italian  end  this  floor  forms  a  promontory  which 
would  hold  several  thousand  men,  aU  of  whom 
would  at  the  same  time  see  the  view.  There  is  no 
other  pass  between  the  Lake  of  Geneva  and  the 
sea  from  which  it  would  be  possible  for  an  army 
or  any  considerable  portion  of  an  army  to  see  the 
plain  of  Italy.  There  is  said  to  be  a  fine  view 
of  Italy  from  the  Col  de  Malaure  near  Abries,  but 
no  one  has  ever  suggested  that  Hannibal  used 
that  pass,  of  which  the  peculiar  structure  is  incon- 
sistent with  the  story,  because  the  view  is  to  be 
seen  only  from  a  narrow  ridge  on  which  no  body 
of  troops  could  be  stationed.  The  nature  of  the 
view  from  the  Col  du  Clapier  is  the  crucial  evidence 
that  this  was  Hannibal's  pass,  and  what  is  even 
more  important,  that  neither  Polybius  nor  Livy 
draws  upon  his  imagination  for  his  statements. 

Assuming  the  Clapier  to  be  the  pass,  we  have 
to  find  the  route  by  which  Hannibal  reached  it. 
The  natural  route  would  be  by  the  river  valleys. 
Tracing  it  backwards,  it  would  follow  the  course  of 

the  Monte  Musine  was  forty-one  miles  from  the  camera,  and  there 
was  no  trace  of  it  on  the  plate.  Accordingly  Fig.  1  has  been 
prepared  by  tracing  from  an  enlarged  photograph  the  outlines 
of  the  hills  up  to  and  including  Monte  Musine,  and  filling  in  from 
sketches  taken  on  the  spot  the  position  of  the  hills  of  Montferrat 
in  the  view. 

Fig.  2  is  a  reproduction  from  a  portion  of  a  rough  Hthographed 
panorama  of  the  Alps  seen  from  the  Capucin  Monastery  at  Turin. 

B 


10  THE  PROBLEM  AND  THE 

the  Arc,  then  of  the  Isere,  and  then  of  the  Rhone» 
This  route  accords  with  the  distances  given  by 
Polybius,  who  divides  the  journey  into  stages  for 
each  of  which  he  gives  the  distance  in  hundreds 
of  stadia,  a  hundred  stadia  being  eleven  miles. 
The  stages  which  concern  us  are  three :  first,  from 
Emporion  (Ampurias)  on  the  Spanish  side  of  the 
Pyrenees  to  the  crossing  of  the  Rhone,  about 
1,600  stadia;  secondly,  from  the  crossing  of  the 
Rhone .  along  the  river  itself  until  the  ascent  of 
the  Alps  leading  towards  Italy,  1,400  stadia;  and 
thirdly,  the  passage  through  the  Alps,  about 
1,200  stadia.  Polybius  refers  to  '  the  ascent  to 
the  Alps '  as  a  specific  point,  the  place  where  the 
first  battle  took  place,  and  his  choice  of  this  point 
as  the  end  of  one  stage  and  the  beginning  of  another 
shows  that  he  has  in  view  some  particular  place 
where  the  transition  from  the  region  of  plain  to 
the  region  of  mountains  is  unmistakable.  If  we 
measure  back  from  Avigliana,  where  the  plain  of 
Turin  begins,  taking  as  our  route  the  Col  du 
Clapier,  the  valley  of  the  Arc,  and  the  valley  of 
the  Isere,  1,200  stadia  or  132  miles  will  bring  us 
within  five  miles  of  the  Bee  d'Echaillon,  where  the 
Isere  emerges  from  the  Alpine  region  and  enters 
the  plain  of  the  Rhone  valley.  If  from  the  Bee 
d'Echaillon  we  measure  down  by  the  bank  of  the 
Isere  and  of  the  Rhone  a  distance  of  1,400  stadia, 
154  miles,  we  shall  reach  a  point  two  miles  north 
of  Fourques,  the  town  which  marks  the  head  of  the 
delta  of  the  Rhone.     The  distance  from  Fourques 


CLUE  TO  ITS  SOLUTION  11 

to  Ampurias  measured  along  the  shortest  possible 
route  is  179  miles,  or  three  miles  more  than  the 
1,600  stadia,  176  miles,  given  by  Polybius  for 
the  march  from  Ampurias  to  the  crossing  of  the 
Rhone.  ^  Assuming  the  itinerary  thus  produced 
to  be  that  indicated  by  Polybius,  we  may  now 
examine  it  in  detail  with  the  text  of  Polybius  and 
of  Livy  before  us. 

1  The  text  of  Polybius  is  met  by  distances  not  exceeding  or 
falling  short  of  bis  measurements  by  100  stadia  or  eleven  miles. 
A  greater  discrepancy  than  eleven  miles  would  be  discordant  with 
his  measurements. 


II 

THE  PASSAGE  OF  THE  RHONE 

According  to  Polybius,  Hannibal  moved  quickly 
from  the  Pyrenees  towards  the  Rhone,  keeping  the 
Sardinian  Sea  on  his  right,  and  when  he  reached  the 
river  '  immediately  took  in  hand  its  passage  by 
the  single  stream,  his  army  being  distant  about 
four  days'  march  from  the  sea '.  He  at  once 
bought  up  from  the  natives  all  their  boats  and 
canoes,  and  gathered  timber  for  the  construction  of 
further  canoes,  so  that  in  two  days  he  had  enough 
craft  for  his  purpose.  A  great  number  of  boats 
were  available,  because  the  people  who  lived 
on  the  banks  of  the  Rhone  were  accustomed 
to  maritime  trade.  Meantime  a  Gallic  army 
collected  on  the  opposite  banks  to  oppose  the 
passage.  Accordingly,  when  the  third  evening 
came,  Hannibal  sent  a  detachment  under  Hanno 
to  cross  the  stream  higher  up,  in  order  to  surprise 
the  enemy  by  a  flank  attack.  Hanno' s  party 
marched  200  stadia,  22  miles,  up  the  river,  and  there 
finding  an  island,  crossed  the  river  in  rafts  and 
canoes  and  spent  the  fifth  night  on  the  left  bank, 
starting  at  dawn  next  day  to  march  down  that 
bank  against  the  enemy.  As  soon  as  Hannibal 
received  Hanno's  signal  that  he  was  at  hand,  he 
ordered  the  crossing  of  the  main  force  to  begin. 


PASSAGE  OF  THE  RHONE  13 

The  natives  were  defeated,  and  during  that  day 
the  whole  of  the  infantry  and  cavaby  crossed  the 
river  and  were  encamped  for  the  sixth  night  on 
its  left  or  eastern  bank.  Next  day  Hannibal 
heard  that  a  Roman  fleet  was  anchored  off  the 
mouths  of  the  Rhone.  He  therefore  sent  off 
500  Numidian  horse  to  reconnoitre  the  Romans, 
and  then  arranged  for  the  transport  of  his  thirty- 
seven  elephants  across  the  river.  Meantime,  in  the 
camp  Hannibal  assembled  the  army,  which  was 
addressed  first  by  certain  Gallic  chiefs  from  the 
valley  of  the  Po  and  then  by  Hannibal  himself. 
No  sooner  were  the  speeches  over  and  the  troops 
dismissed  than  the  Numidian  cavalry  returned. 
They  had  been  defeated  by  a  party  of  Roman  and 
Celtic  cavalry  that  had  been  sent  to  reconnoitre 
by  the  Roman  commander  Scipio,  who  had  dis- 
embarked his  army  to  the  east  of  the  eastern  mouth 
of  the  Rhone,  that  is  near  the  modern  Eos.  The 
engagement  had  taken  place  not  far  from  the 
Carthaginian  camp,  which  the  Roman  cavalry, 
being  successful,  reconnoitred,  and  then  hurried 
back  to  report.  Scipio  at  once  re-embarked  the 
baggage  of  his  army  and  set  off  with  his  troops 
beside  the  river,  being  eager  to  engage  the  Cartha- 
ginians. These  events  appear  to  have  all  taken 
place  on  the  day  following  the  sixth  night,  that 
is,  the  seventh  day  after  Hannibal's  reaching  the 
river.  Next  day  (the  eighth)  in  the  morning 
Hannibal  sent  the  whole  of  his  cavalry  towards 
the  sea  to  act  as  a  rear-guard,  and  set  his  infantry 


14  PASSAGE  OF  THE  RHONE 

in  march  towards  the  north.  He  then  completed 
the  transport  of  the  elephants.  A  number  of 
large  rafts  were  made  and  fastened  one  to  another 
until  a  landing-stage  had  been  formed  which 
projected  200  feet  from  the  bank  towards  mid- 
stream. To  the  extremity  of  this  platform  was 
tied  a  large  double  raft/  on  to  which  the  elephants 
were  brought.  This  was  then  cut  loose  from 
the  landing-stage  and  towed  by  boats  across  the 
river.  Some  of  the  elephants  were  frightened 
and  fell  off  the  raft,  but  even  these  got  safely 
across,  some  of  them  having  to  swim  a  short 
distance,  but  most  of  them  finding  that  they  were 
not  out  of  their  depth.  As  soon  as  the  elephants 
were  across,  stiU  on  the  eighth  day,  Hannibal 
recalled  the  cavalry  and  set  off,  with  the  cavalry 
and  the  elephants  as  a  rear-guard,  to  follow  the 
infantry  towards  the  north. 

Scipio  reached  the  point  where  the  river  had 
been  crossed  on  the  third  day  from  the  departure 
of  the  Carthaginians. 

Almost  all  the  early  writers  have  supposed  that 
Hannibal  crossed  the  Rhone  at  a  point  much 
higher  up  the  stream  than  that  just  below  which 
it  bifurcates  to  form  the  delta.  Napoleon,  con- 
sidering the  abstract  probabilities,  thought  that 
Hannibal  would  wish  to  cross  the  river  above  its 
confluence  with  the  Durance  and  below  its  con- 

^  It  may  be  inferred  from  chap,  xlvi,  par.  1,  that  the  double 
raft  was  at  least  50  feet  broad.  Livy  makes  the  landing-stage 
project  200  feet,  and  the  raft  100  feet  broad. 


PASSAGE  OF  THE  RHONE  15 

fluence  with  the  Ardeche,  and  accordmgly  the 
favourite  sites  are  at  Roquemaure  and  Pont 
St.  Esprit.  Captain  Colin,  by  suggesting  that  the 
point  of  passage  was  at  the  head  of  the  delta  just 
above  Fourques,  has  contributed  almost  as  much 
to  the  elucidation  of  the  story  told  by  Polybius  as 
Colonel  Perrin  by  indicating  the  Col  du  Clapier  as 
Hannibal's  pass.  The  measurements  of  distance 
given  by  Polybius  point  to  a  crossing  at  the 
lowest  possible  point  above  the  delta.  This  seems 
to  be  the  natural  meaning  of  the  words  of  Polybius, 
that  Hannibal,  when  he  reached  the  river, '  immedi- 
ately took  in  hand  its  passage  by  the  single  stream.'^ 
The  details  of  the  story  suggest  a  point  as  near 
the  sea  as  possible.  The  statement  of  Polybius 
that  boats  were  numerous  because  the  natives  were 
addicted  to  maritime  trade  is  hardly  consistent 
with  any  place  north  of  Tarascon.  The  landing- 
stage  and  the  raft  tied  together  projected,  accord- 
ing to  Polybius,  250,  and  according  to  Livy,  300 
feet  into  the  stream.  This  space  of  300  feet  could 
only  be  a  fraction  of  the  breadth  of  the  river,  for 
otherwise  Hannibal  would  have  prolonged  his 
landing-stage  until  it  formed  a  bridge.  The 
necessary  breadth  will  not  be  found  above  the 
confluence  with  the  Durance.  The  use  of  the 
landing-stage  and  of  the  raft,  as  well  as  the  fact 
that  the  horses  swam  the  river,  point  to  its  more 

1  xlii.  1  'AvvtySas  8e  Trpocr/Ai^as  tois  Trept  rbv  iroTafjLov  tottois 
€v6ew<;  kv^^upu  Troieiadai  t^v  8ta/3acrtv  Kara  rrjv  a.TrXrjv  pvcriv,  (T)(e86v 
r]fi€p<j}v  TCTTapuv  68ov  aTri)((jiV  crTpaTOTreSo)  t'^?  6a\dTTr]<;. 


16  PASSAGE  OF  THE  RHONE 

sluggish  course  in  the  region  below  Tarascon,  and 
are  inconsistent  with  the  swifter  stream  near 
Roquemaure  or  Pont  St.  Esprit. 

Some  writers  have  been  puzzled  by  the  state- 
ment of  Polybius  that   Hannibal's  camp  at  the 
point  of  passage  was  about  four  days'  march  from 
the  sea,  Eos  being  little  more  than  thirty  miles 
distant  from  the  Rhone  above  Fourques.     They 
suppose  that  four  days'  march  ought  to  be  con- 
siderably more  than  thirty  miles.     But  a  day's 
march   is   not    a   fixed  distance.     Polybius   says 
'  about  four  days'  march ',  because,  according  to 
his  account,  Scipio  reached  the  Carthaginian  camp 
on  the  fourth  day  from  his  leaving  the  shore. 
Scipio  started  in  the  afternoon  or  evening  of  the 
seventh   day.      Hannibal   marched   off   with   the 
elephants  on  the  eighth  day.     On  the  tenth  day 
Scipio  arrived  at  the  camp,  to  find  that  the  Cartha- 
ginians   were    gone.      The    Roman    general    was 
marching  to  meet  the  most  dangerous  enemy  in  the 
world.    Livy  distinctly  states  that  he  marched  in 
order  of  battle,  a  procedure  which  is  inconsistent 
with  rapidity  of  movement.     Hannibal's  average 
rate  of  march  up  the  vaUey  of  the  Rhone  was 
rather  less  than  nine  miles  a  day.    Scipio,  marching 
with  his  army  drawn  out  for  battle  and  carefully 
fortifying  his  camp  each  night,   would  be  very 
unlikely  to  move  faster  than  this.    But  on  the  first 
day,  setting  off  in  the  late  afternoon,  for  he  could 
not  start  until  the  cavalry  had  reported  and  his 
baggage  had  been  shipped,  he  would  hardly  do 


PASSAGE  OF  THE  RHONE  17 

more  than  four  or  five  miles,  and  there  is  no  reason 
to  suppose  that  on  the  fourth  day,  when  he  reached 
the  site  of  the  Carthaginian  camp,  he  had  made 
a  full  day's  march.  Accordingly  the  distance  from 
Fos  to  a  point  on  the  river  a  mile  or  two  north  of 
Fourques  perfectly  accords  with  the  text  of  Poly- 
bius,  with  which,  in  all  that  concerns  the  passage 
of  the  Rhone,  that  of  Livy  precisely  concurs. 


Ill 

FROM  THE  RHONE  TO  THE  BEGINNING  OF  THE  ALPS 

After  crossing  the  Rhone,  Hannibal  marched 
up  that  river  away  from  the  sea  '  towards  the  centre 
of  Europe '.  On  the  fourth  day's  march  from 
the  point  of  passage  he  reached  a  district  called 
the  Island,  populous  and  rich  in  corn,  resembling 
in  size  and  shape  the  delta  of  Egypt,  being  a 
triangle  of  which  two  sides  were  formed  by  the 
Rhone  and  its  tributary,  the  Skaras,  and  the  third 
side,  not  as  in  Egypt  by  the  sea,  but  by  mountains 
hard  to  approach  and  hard  to  enter,  and  therefore 
impassable.  In  the  Island  two  brothers  were 
rivals  for  the  chieftainship.  The  elder  brother 
appealed  to  Hannibal,  with  whose  support  he 
established  his  supremacy  and  expelled  his  rival. 
The  successful  chief  showed  his  gratitude  by  supply- 
ing Hannibal's  army  with  provisions,  new  weapons, 
clothing,  and  shoes.  Moreover,  as  Hannibal  was 
anxious  about  a  possible  encounter  with  the  tribe 
of  the  AUobroges,  the  friendly  chief  accompanied 
him  with  his  own  army  on  his  further  march,  until 
the  ascent  of  the  Alps  was  approached.  This  took 
place  after  ten  days'  march  by  the  river  from  the 
Island,  during  which  a  distance  of  800  stadia, 
eighty-eight  miles,  was  covered.  Then  began  the 
ascent  towards  the  Alps,  and  the  difficulties 
presented  by  the  country  itself. 


FROM  RHONE  TO  ALPS  19 

The  identification  of  the  Island  described  by 
Polybius  has  embarrassed  those  historians  who 
place  the  crossing  of  the  Rhone  north  of  its  con- 
fluence with  the  Durance.  If  the  crossing  was 
near  Fourques  the  problem  of  the  island  is  sim- 
plified. Starting  from  Fourques,  Hannibal  during 
the  second  day  would  reach  the  Durance.  This 
river  would  be  forded  by  the  army,  not  without 
difficulty,  and  the  operation  would  certainly  not 
be  completed  before  the  evening  of  that  day,  if  so 
soon.  It  might  very  well  have  been  a  whole  day's 
work.  The  third  day's  march  would  take  him 
from  the  Durance  to  the  Sorgues  near  its  con- 
fluence with  the  Rhone,  and  on  the  fourth  day 
he  would  be  in  the  region  between  the  Sorgues 
and  the  Rhone.  This  region  offers  the  site  for 
the  Island  which  best  accords  with  the  text  of 
Polybius.  The  Rhone  with  almost  any  of  its 
tributaries  wiU  make  the  necessary  two  sides  of 
a  triangle,  and  the  mountains  can  in  most  cases 
be  construed  as  forming  the  third  side.  Livy 
describes  the  Island  in  much  the  same  terms  as 
Polybius,  but  names  the  tributary  river  Ibisarar 
or  Ibisaras,  Bisarar  or  Bisaras,  Sarar  or  Saras, 
according  to  the  reading.  None  of  the  possible 
regions  are  as  large  as  the  delta  of  Eg3rpt,  but 
several  of  them  have  the  triangular  shape  required. 
The  most  probable  tributary  is  the  Sorgues,  on 
which  there  is  a  town  which  still  bears  the  name 
of  the  Island  (Isle-sur-Sorgues). 

Polybius  makes  no  mention  whatever  of  the 


20  FROM  THE  RHONE 

Durance.  Livy  repeats  exactly  the  story  of  Poly- 
bius  until  he  has  concluded  the  account  of  the 
assistance  given  to  Hannibal  by  the  chief  in  the 
Island,  though  after  saying  that  the  Allobroges 
lived  '  near '  the  Island,  he  writes  as  though  the 
rival  chiefs  were  themselves  AUobroges.  But 
Livy  goes  on  to  say  that  after  settling  the 
dispute,  as  he  went  on  towards  the  Alps,  Hannibal 
did  not  go  the  straight  road,  but  turned  to  the 
left  to  the  Tricastini,  that  thence  he  moved  round 
the  edge  of  the  territory  of  the  Vocontii  into  that 
of  the  Tricorii,  his  march  being  quite  unhindered 
until  he  reached  the  river  Durance,  which  his 
army  forded  with  considerable  difficulty.  From 
the  Durance,  Livy  says  a  little  later,  Hannibal 
marched  as  far  as  the  Alps  through  level  country, 
without  molestation  from  the  natives. 

This  portion  of  Livy's  story  has  been  thought 
irreconcilable  with  the  narrative  of  Polybius,  and 
has  led  a  number  of  writers,  who  suppose  Hannibal 
to  have  crossed  the  Rhone  above  its  confluence 
with  the  Durance,  to  take  Hannibal  by  difficult, 
circuitous,  and  impracticable  routes  to  the  Upper 
Durance.  But  the  text  of  Livy  is  perfectly 
explicable  and  intelligible  if  we  assume  that  he 
has  compiled  it  from  different  accounts,  and 
has  made  a  slight  slip  in  putting  them  together. 
He  first  follows  an  authority  who  must  be  either 
Polybius  or  the  original  of  Polybius,  until  he  is 
done  with  the  Island  and  the  two  brothers ;  this 
version  begins  by  saying  that  Hannibal  marched 


TO  THE  ALPS  21 

up  the  left  bank  of  the  Rhone,  not  because  it  was 
the  more  direct  way  to  the  Alps  but  because  he 
wished  to  avoid  Scipio's  army.  But  when  Livy 
has  finished  with  the  Island  he  begins  again,  and 
again  says  that  Hannibal  going  on  towards  the 
Alps  did  not  follow  the  direct  way,  but  turned  to 
his  left.  This  new  beginning  is  a  repetition,  and 
probably  represents  a  second  version  of  Hannibal's 
decision  at  the  moment  when  he  had  crossed  the 
Rhone.  At  that  moment,  when  Hannibal  was  on 
the  left  bank  of  the  Rhone  south  of  Tarascon,  the 
direct  route  to  the  Alps  would  have  been  up  the 
valley  of  the  Durance,  but  instead  of  that  Hannibal 
went  to  his  left  up  the  Rhone  through  the  Tricas- 
tini  (either  near  St.-Paul-Trois-Chateaux  or  a  few 
marches  further  north),  round  the  Vocontii  (Vercors) 
to  the  Tricorii  (between  Grenoble  and  Vizille). 
This  is  exactly  the  route  that  has  been  suggested 
as  accordant  with  the  text  of  Polybius.  Setting 
out  in  this  direction  from  his  crossing  of  the 
Rhone,  Hannibal  first  came  to  the  Durance,  which 
he  forded  near  its  mouth,  and  Livy  gives  a  vivid 
description  of  the  fording.  Livy's  third  passage, 
saying  that  the  march  to  the  Alps  was  made 
through  level  country  without  molestation  from 
any  of  the  inhabitants,  is  a  condensation  of  what 
Polybius  has  to  say  about  the  same  march,^  which 
we  shall  examine  in  the  next  chapter. 

1  I  have  little  doubt  that  Livy  had  three  separate  texts  before 
him.  From  the  first  of  these  he  copied  chap,  xxxi,  pars.  1-8,  from 
the  second  he  copied  chap,  xxxi,  pars.  9-12,  and  wrote  himself. 


22  FROM  RHONE  TO  ALPS 

The  objection  that  Livy  was  incapable  of  con- 
fusing the  story  so  as  to  bring  in  at  the  wrong 
places  the  passages  which  he  took  from  a  second 
and  third  source,  can  hardly  be  sustained  by  those 
who  infer  from  Livy's  narrative  that  Hannibal 
crossed  the  Durance  in  its  upper  course,  for  they 
impute  to  Livy  a  very  much  greater  derangement 
of  the  proper  order  of  his  narrative. 

to  join  the  two  extracts,  the  words  :  sedatis  Hannibal  certaminibus 
Allobrogum  cum  iam  Alpes  peteret.  From  the  third  he  copied 
chap,  xxxii,  par.  6. 


IV 

THE  FIRST  BATTLE 

There  were  two  ways  in  which,  in  the  mountains, 
the  native  tribes  could  oppose  Hannibal.  They 
might  choose  a  strong  position  athwart  his  line 
of  march,  thus  challenging  him  to  a  pitched 
battle.  In  such  a  case  he  would  have  no  reason 
for  attacking  them  if  he  could  move  round  the 
position,  and  the  position  would  be  of  no  use  to 
them  if  he  could  possibly  turn  it.  But  if  he  must 
attack,  it  was  probable  that  with  his  highly  trained 
and  disciplined  army  he  would  make  short  work 
of  any  Gallic  tribes  attempting  to  hold  a  position 
against  him.  The  Gauls  blocking  his  advance 
would  be  as  easily  brushed  away  by  Hannibal  as 
were  the  Afghans  blocking  the  advance  of  Sir 
Frederick  Roberts  on  the  Peiwar  Kotal. 

But  another  plan  was  open  to  the  Gauls.  When 
the  Afghans  destroyed  an  Anglo-Indian  army  in 
1842  they  assailed  it  where  it  was  stretched  out 
in  a  long  column  in  the  gorges  of  the  Kurd  Kabul 
pass.  It  is  in  the  defiles  that  an  army  on  the  march 
can  most  effectively  be  assailed. 

Hannibal  started  from  the  Rhone  with  38,000 
infantry ;  that  means  a  column  seven  miles  long 
when  in  perfect  order  on  a  good  road,  and  ten  miles 
long   when  slightly  disordered  by  its  movement. 


24  THE  FIRST  BATTLE 

In  single  file  on  a  mountain  path  this  mass  of 
38,000  men  would  fiU  twenty-one  miles  of  the  path. 
Hannibal  had  8,000  horse,  and  must  have  had 
a  couple  of  thousand  mules  to  carry  a  week's 
provisions  for  his  army.  These  10,000  horses  and 
mules  marching  two  and  two  would  make  a  column 
eleven  miles  long ;  and  in  single  file  the  column 
would  trail  out  to  a  length  of  not  less  than  twenty- 
two  miles.  These  figures  show  the  dangers  to 
which  an  army  is  exposed  if  it  is  liable  to  attack 
when  engaged  in  a  defile. 

According  to  Polybius,  after  ten  days'  march 
from  the  Island  along  the  Rhone,  during  which 
Hannibal  was  escorted  by  the  chief  from  the 
Island,  he  began  to  approach  the  hilly  country, 
and  almost  immediately  learned  that  the  Allobroges 
were  occupying  a  position  which  he  must  needs 
pass;  in  other  words,  they  were  holding  against 
him  a  line  of  heights  which  he  could  not  turn. 

Suppose  the  Island  to  be  the  district  between 
Rhone  and  Sorgues.  If  we  measure  from  Bedar- 
rides,  a  point  in  that  district,  a  distance  of  eighty- 
eight  miles  along  the  vaUey  of  the  Rhone  and  up 
the  left  bank  of  the  Isere,  we  shall  reach  St.  Nazaire, 
where  the  hiUs  of  the  Vercors  approach  the  Isere, 
and  where  the  difficulties  of  a  further  march  up 
the  Isere  would  begin.  From  St.  Nazaire  the 
road  follows  the  narrow  strip  of  flat  ground  which 
lies  between  the  mountains  of  Vercors  and  the 
swift  and  deep  Isere.  At  a  distance  of  twenty-five 
miles  from  St.   Nazaire  the  Isere  emerges  from 


THE  FIRST  BATTLE  25 

the  wide  gorge  between  the  mountains  of  Vercors 
and  those  of  the  Grande  Chartreuse.  Coming  out 
of  the  gorge,  the  river  skirts  the  foot  of  the  Bee 
d'Echaillon,  a  precipitous  cUff  surmounted  by  the 
plateau  of  St.  Ours,  about  1,700  feet  high.  This 
is  surmounted  by  a  second  precipice  of  which  the 
summit,  3,000  feet  high,  is  agaui  a  plateau  sloping 
gently  to  the  south  for  about  a  mile  and  a  half, 
and  then  rising  rapidly  to  the  ridge  of  the  Vercors, 
which  is  here  about  4,800  feet  high.  The  lowest 
point  of  this  second  plateau  is  about  2,300  feet 
above  the  sea  and  1,700  feet  above  the  Isere.  Until 
the  last  century  there  was  no  practicable  road 
round  the  Bee  d'Echaillon.  The  road  went  over 
the  Col  of  the  second  plateau  from  St.  Quentin 
on  the  west  to  Veurey  on  the  east,  or  to  Noyarey, 
a  mile  and  a  half  nearer  to  Grenoble.  The  old 
course  of  the  Isere  skirts  the  foot  of  the  cliffs 
between  Noyarey  and  Veurey,  so  that  it  is  probable 
that  in  ancient  times  the  only  practicable  road  was 
from  St.  Quentin  to  Noyarey.  At  the  present  day, 
besides  the  modern  high  road  which  foUows  the 
left  bank  of  the  Isere  and  skirts  the  foot  of  the 
Bee  d'Echaillon,  there  is  a  good  cart-road  which 
ascends  in  long  zigzags  the  steep  hill  from  St. 
Quentin  to  Montaud,  a  village  on  the  plateau,  and 
then  descends  a  ravine  to  Veurey.  A  mule-track 
and  several  footpaths  lead  from  the  plateau  near 
Montaud  over  one  of  the  spurs  running  down 
from  the  Vercors,  and  descends  a  second  gorge 
to  Noyarey.     The  gorge  ending   at   Noyarey  is 


26  THE  FIRST  BATTLE 

very  difficult;  rugged  and  precipitous  paths  lead 
down  on  either  side  of  it,  and  either  path  is  of 
course  commanded  from  the  high  ground.  In 
this  region  Captain  Colin  finds  the  site  of  Han- 
nibal's battle  with  the  AUobroges.  The  position 
they  held  was  the  plateau  of  Montaud;  the 
defile  which  followed  was  one  of  the  gorges  leading 
from  Montaud  to  Veurey  or  Noyarey.  No  other 
site  which  has  been  suggested  for  this  battle  so 
well  accords  with  the  account  given  by  Polybius. 
Hannibal  learned  after  the  friendly  chief  had  left 
him  that  the  enemy  was  holding  a  position  which 
he  could  not  evade.  He  then  sent  forward  Gallic 
scouts  to  reconnoitre  the  enemy's  arrangements 
and  intentions,  and  on  receipt  of  the  information 
marched  his  army  ostentatiously  to  the  foot  of 
the  position  and  there  encamped.  From  St.  Nazaire 
to  St.  Quentin  is  twenty-one  miles,  two  marches 
for  Hannibal's  army.  After  passing  St.  Nazaire 
he  would  hear  that  the  Gauls  were  occupying  the 
plateau  of  Montaud  which  he  must  cross,  for  he 
could  neither  take  his  army  over  the  mountains 
of  Vercors,  4,500  feet  high,  nor  go  round  or  across 
the  Bee  d'Echaillon.  He  therefore  pitched  his 
camp  probably  near  St.  Quentin.  Being  informed 
that  the  AUobroges  dispersed  to  their  homes  at 
night,  he  took  a  body  of  light  armed  troops  and 
made  his  way  in  the  night  on  to  the  ground  which 
the  Gauls  thus  left  clear  for  him.  This  would  be 
the  plateau  near  Montaud,  of  which  the  capture 
was   exactly    analogous   to   that   of    the   Peiwar 


THE  FIRST  BATTLE  27 

Kotal.  At  dawn  the  army  moved  out  of  the  camp, 
ascended  the  plateau,  and  entered  the  defile  beyond. 
Then  the  returnuig  Gauls  saw  their  opportunity, 
and  attacked  the  column  when  a  large  part  of  it 
was  aheady  committed  to  the  defile.  An  examina- 
tion of  the  ground,  especially  between  Montaud  and 
Noyarey,  leaves  no  doubt  that  an  army  in  the 
situation  described  might  very  well  have  lost  4,000 
or  5,000  men  and  a  large  number  of  horses  and 
mules.  Even  the  easier  descent  to  Veurey  might 
also  have  been  disastrous.  The  essential  point  in 
any  attempt  to  identify  the  site  of  this  first  battle 
is  that  we  must  distinguish  between  the  frontal 
position  seized  by  Hannibal  in  the  night  and  the 
subsequent  defile  in  which  the  disaster  occurred. 
The  position  of  Montaud  with  the  defile  in  rear  is 
the  only  site  which  has  been  suggested  that 
complies  with  these  essential  conditions.  The 
words  of  Livy,  '  praecipites  deruptae  utrimque 
angustiae,'  have  been  taken  to  describe  a  path 
along  the  top  of  a  ridge  precipitous  on  each  side. 
But  this  in  the  conditions  is  a  military  absurdity. 
The  road  must  have  been  en  corniche  with  a  preci- 
pice falling  down  on  one  side  and  a  precipice  rising 
up  on  the  other. 

When  he  had  dispersed  the  Allobroges  and 
restored  order  in  his  column,  Hannibal  pushed  on 
with  such  men  as  he  could  collect  and  seized  the 
enemy's  town,  according  to  Livy  their  capital, 
where  he  obtained  fresh  horses  and  mules,  as  well 
as  corn  and  cattle  enough  to  supply  his  army 


28  THE  FIRST  BATTLE 

for  two  or  three  days.  This  would  be  possible 
only  at  a  considerable  town  such  as  Grenoble, 
which  is  first  heard  of  as  one  of  the  capitals  of  the 
Allobroges,  and  would  be  reached  from  Noyarey 
by  cavalry  in  less  .than  an  hour,  and  by  infantry 
in  a  couple  of  hours. 


THE  SECOND  BATTLE 

From   the  town    of   the    Allobroges    Hannibal 

marched    three    days    unmolested.      During    the 

fourth  day  he  was  met  by  a  fresh  set  of  native 

chiefs  offering  friendship  but  planning  treachery, 

which  resulted  in  a  great  attack  upon  his  army 

two  days  later,  when  engaged  in  a  defile.     The 

probable  place  of  this  meeting  with  the  chiefs  seems 

to  be  the  entrance  to  the  vaUey  of  the  Arc,  which 

Hannibal  would  easily  reach  on  the  fourth  day 

from  Grenoble.     The  Arc  here  skirts  the  foot  of 

the  cliffs  on  its  left  bank,  and  the  road  may  very 

well  have  followed  a  pass  over  the  hills  instead  of 

the  bank  of  the  river.      In  that  case  Hannibal 

probably    left    the    Isere    near    Pontcharra    and 

marched    by  La   Rochette   and  the   Col  of    the 

Grand  or  Petit  Cucheron.       Either  here,  or  near 

Aiguebelle  if  he  foUowed  the  course  of  the  Arc, 

he  was  met  by  the  chiefs  who  offered  to  guide  him.^ 

1  Colonel  Perrin  proposes  as  the  site  of  tlie  first  battle  the 
ascent  from  Chamoux  to  Montandry.  Lieutenant  Azan  would 
place  it  at  the  Col  of  the  Grand  Cucheron.  Neither,  of  these  sites 
fulfils  the  conditions.  Either  of  the  positions  proposed  could 
have  been  turned  by  Hannibal,  and  in  neither  case  is  there  a  subse- 
quent defile  of  the  kind  required.  Colonel  Perrin  and  Lieutenant 
Azan  are  compelled  to  look  for  a  site  for  the  first  battle  further 
north  than  the  Bee  d'Echaillon,  because  they  make  Hannibal 
cross  the  Khone  north  of  Avignon  and  are  compelled  to  measure 
1,400  stadia  from  the  point  of  passage. 


30  THE  SECOND  BATTLE 

He  accepted  their  company  without  dismissing  his 
suspicions. 

He  probably  advanced  some  miles  further  during 
this  fourth  day.  After  two  further  marches  came 
the  great  attack  when  the  army  was  in  the 
defile.  The  ground  that  best  corresponds  to  the 
story  is  near  St.  Michel.  Here  the  valley  is  almost 
entirely  blocked  by  the  Rocher  de  la  Porte,  which 
has  a  precipitous  edge  several  hundred  feet  high, 
and  is  nearly  a  mile  long  from  flank  to  flank  of 
the  valley,  having  at  its  south  end  a  narrow  gully 
in  which  there  is  room  only  for  the  river,  the  road, 
and  the  railway.  Above  this  rock  the  valley  is 
for  many  miles  so  narrow  as  to  have  no  flat  ground 
at  all.  Below  the  rock  it  is  wide  enough  for  the 
army  to  have  marched  in  several  parallel  columns. 
This  is  just  the  place  where  the  tribes  would  com- 
bine their  attack,  for  at  this  point  descend  the 
principal  side-routes  into  the  valley — the  one 
from  the  north  over  the  Col  des  Encombres,  the 
other  from  the  south  over  the  Galibier.  Here 
then  the  tribes  could  arrange  to  concentrate 
suddenly.  Hannibal's  cavalry  column  being  in 
front  and  the  infantry  coming  up  behind,  the 
Gauls  would  cut  the  column  in  two.  Hannibal 
would  storm  the  rock  at  its  north  end,  where  it 
is  accessible,  and  would  then  deploy  his  army 
against  the  crowds  of  natives.  This  would  occupy 
the  natives  and  give  the  mounted  troops  time  to 
escape.  The  natives  would  be  unable  to  stand 
against  the  Carthaginians  once  they  had  gained 


THE  SECOND  BATTLE  31 

the  level  top  of  the  rock.  The  fortress-like  nature 
of  this  rock,  which  is  as  white  as  any  that  can  be 
found  in  the  Alps,  absolutely  corresponds  with 
the  words  of  Polybius,  '  a  strong  (or  fortress-like) 
white  rock.'  There  is  nothing  resembling  it 
either  on  the  route  of  the  Little  St.  Bernard,  on 
that  of  the  Mont  Genevre,  or  on  that  of  the  Col 
de  Larche. 

Next  day  Hannibal  pushed  on  with  his  infantry 
and  joined  the  mules  and  the  cavalry,  and  on  the 
ninth  day,  apparently  the  third  day  after  the 
battle,  encamped  on  the  summit  of  the  pass. 


VI 

THE  PASS  AND  THE  DESCENT 

The  Col  du  Clapier  is  like  a  trough  three  or  four 
miles  long  and  half  a  mile  wide  which  has  been 
planed  out  in  a  straight  line  across  the  main  ridge 
of  the  Alps.  The  promontory  on  which  the  trough 
ends  is  8,000  feet  above  the  sea.  The  peaks  on 
each  side  rise  up  nearly  3,000  feet  above  the  bottom 
of  the  trough,  that  on  the  east  being  Mont  Giusalet, 
that  on  the  west  the  Aiguille  de  Savuie,  of  which 
the  ridge  is  prolonged  for  some  miles  to  the  south 
beyond  the  point  where  the  pass  ends  in  the 
promontory  from  which  the  view  of  Italy  is 
obtained.  On  this  ridge  about  a  mile  from  the 
promontory  lies  the  sloping  glacier  Del  Agnello. 
The  descent  from  the  pass  is  into  a  great  amphi- 
theatre, of  which  the  sides  are  so  steep  as  to  be 
for  the  most  part  precipitous.  The  streams  that 
run  down  these  sides  collect  into  the  Clarea,  which 
rises  between  the  promontory  and  the  glacier, 
drains  the  amphitheatre,  and  runs  through  a  deep 
gorge  into  the  Dora  Riparia.  There  are  two  ways 
down  from  the  Col.  The  ordinary  path  descends  into 
the  amphitheatre,  follows  the  left  bank  of  the  Clarea 
and  leads  through  the  gorge  to  Giaglione.  But  there 
is  an  alternative  route  which  skirts  the  top  of  the 
amphitheatre  on  the  right-hand  side,  keeping  pretty 


THE  PASS  AND  DESCENT  33 

nearly  level  till  it  reaches  the  huts  of  Tuglia,  where 
it  crosses  a  ridge  and  descends  into  the  valley  of  the 
Dora.  It  passes  some  distance  below  the  glacier  Del 
Agnello.  The  path  leading  to  Tuglia  is  thus  a  path  en 
corniche,  and  there  is  a  point  at  which  it  has  in  quite 
recent  years  been  carried  away  by  landslides.  The 
direct  descent  into  the  amphitheatre  is  after  the 
first  quarter  of  a  mile  exceedingly  steep,  and  if  the 
traveller  strays  a  few  yards  from  the  path  he  finds 
himself  looking  down  a  precipice.  In  1690  the 
Vaudois  on  their  march  from  Switzerland,  known 
as  '  the  glorious  return ',  encamped  on  the  Col  du 
Clapier,  and  next  morning  continued  their  progress 
in  a  mist  by  Tuglia  to  Exilles.  But  a  portion  of 
them  by  mistake  descended  the  ordinary  path,  to 
find  the  gorge  which  is  the  exit  from  the  amphi- 
theatre blocked  by  the  Piedmontese  troops.  Some 
of  them  escaped  by  a  terrible  climb  from  the  bottom 
of  the  valley  up  to  Tuglia.  The  corniche  path 
by  Tuglia  was  followed  in  1794  by  a  smaU  French 
column  destined  to  attack  Exilles. 

While  Hannibal  was  on  the  pass  there  was  a  fall 
of  snow,  which  was  renewed  during  his  descent. 
Polybius  says  that  the  descent  was  steep  and 
narrow,  and  the  path  hidden  by  the  snow,  and  that 
those  who  missed  it  fell  over  the  precipices.  The 
head  of  the  column  shortly  came  to  a  place  which, 
in  consequence  of  a  recent  landslide  following 
upon  an  earlier  one,  was  for  a  space  of  300  yards 
too  narrow  for  the  mules  or  the  elephants  to  be 
able  to  pass.     Thereupon  Hannibal  attempted  to 

E 


34  THE  PASS  AND  DESCENT 

find  a  way  round.  But  a  fresh  fall  of  snow  made 
the  roundabout  way  impossible,  and  he  gave  it  up. 
'  What  occurred,'  says  Polybius,  '  was  peculiar  and 
extraordinary.  For  the  new  snow  having  just 
fallen  upon  that  which  was  there  before  and  had 
remained  from  the  previous  winter,  was  easy  to 
penetrate,  being  fresh,  still  soft,  and  not  yet  deep. 
The  men  therefore  trod  through  it  on  to  the 
hard  snow  below,  which  their  feet  did  not  penetrate 
but  slipped  from  under  them.  Then  when  they 
tried  to  rise,  propping  themselves  up  with  their 
knees  or  their  hands  to  stand  up  again,  they  slid 
along  all  the  more,  as  the  place  was  very  steep. 
The  mules,  however,  penetrated  the  frozen  snow 
bejow  and  stuck  fast  in  it.'  This  seems  to  be 
a  description  of  an  attempt  to  cross  a  glacier. 
With  it  may  be  compared  Mr.  Ellsworth  Hunting- 
ton's account  of  his  crossing  the  Chang  La  pass 
in  Tibet  {The  Pulse  of  Asia,  p.  71) : — 

*  The  ascent  was  not  specially  difficult,  as  we 
started  at  night  when  there  was  a  stiff  crust  on 
the  snow.  On  the  farther  side  the  coolies  made 
good  progress,  although  the  new  snow,  which  had 
fallen  to  a  depth  of  eight  or  ten  inches  on  the  old 
crust  since  last  I  crossed,  became  soft  almost  as  soon 
as  the  sun  rose.  The  animals,  however,  even 
though  none  of  them  carried  loads,  broke  through 
and  floundered  and  struggled  pitifully,  scarcely 
able  to  draw  their  bleeding  legs  out  of  the  deep 
holes  in  the  icy  crust.' 

Hannibal  gave  up  the  attempt  to  find  a  way 
round,  and  proceeded  to  repair  the  broken  path, 


THE  PASS  AND  DESCENT  35 

which  after  one  day's  work  was  practicable  for 
the  mules  and  horses,  which  were  at  once  sent 
down  to  a  camp  below  the  snow-line.  Two  more 
days'  building  at  the  path  made  it  possible  to  lead 
down  the  elephants. 

The  story  as  told  by  Polybius  suggests  that 
the  path  which  Hannibal  thus  repaired  was 
the  corniche  leading  from  the  Col  to  the  huts 
of  Tuglia,  and  that  the  circuitous  route  which 
he  attempted,  but  had  to  give  up,  led  over 
the  glacier  Del  Agnello,  by  crossing  which  he 
could  have  reached  the  Col  d'Ambin.  But  Livy 
has  a  different  version.  According  to  Livy,  the 
broken  road  was  a  landslip  which  had  transformed 
the  path  into  a  precipice  1,000  feet  deep,  and  the 
road  which  Hannibal  quarried  led  down  the  face 
of  this  precipice.  According  to  this  version, 
Hannibal  must  have  descended  by  something  like 
the  existing  path  into  the  amphitheatre,  which 
exactly  corresponds  with  the  first  few  sentences 
of  the  narrative  of  Polybius  where  he  says  that 
all  who  missed  the  path  fell  over  precipices,  and 
which  by  a  slight  landslip  could  at  any  time  be 
transformed  into  just  such  a  precipice  as  Livy 
describes.  Livy  repeats  precisely  the  story  told  by 
Polybius  of  the  attempt  to  cross  the  glacier  and 
of  its  failure. 

The  explanation  suggested  by  several  visits  to 
the  pass,  and  repeated  study  of  the  text  of  both 
writers,  is  that  Hannibal  made  use  both  of  the 
descent  into  the  amphitheatre  and  of  the  corniche. 


36  THE  PASS  AND  DESCENT 

The  infantry  would  start  down  the  ordinary  path, 
and  in  spite  of  the  accidents  many  of  them  would 
go  that  way.  Hannibal  would  think  it  too  danger- 
ous for  his  horses,  mules,  and  elephants,  and  would 
start  with  them  along  the  corniche.  Then  he 
would  try  the  glacier,  and  having  given  that  up 
would  repair  the  broken  corniche  for  the  animals.^ 

1  I  have  made  only  the  descent  by  the  Granges  de  Savine  to 
Giaglione,  and  have  neither  followed  the  corniche  nor  examined 
the  glacier,  both  of  which  I  suspect  would  repay  the  scrutiny 
of  an  expert  cragsman  interested  in  Hannibal. 


VII 

REJECTED  HYPOTHESES 

A  SERIES  of  visits  to  the  Alps,  and  especially  the 
examination  of  all  the  principal  passes  (repeated 
again  and  again  for  the  Mont  Genevre  and  the 
Col  de  Larche)  as  well  as  of  the  valleys  leading  to 
the  passes,  have  convinced  me  that  the  route  which 
has  here  been  traced  fuUy  accords  with  the  text 
of  Polybius,  and  offers  a  satisfactory  explanation 
of  the  apparent  discrepancies  between  Livy  and 
his  Greek  predecessor.  The  acceptance  of  this 
route  clears  Polybius  of  the  charges  of  ambiguity 
and  obscurity  brought  against  him  by  modern 
writers  who  have  lacked  the  diligence  or  the 
opportunity  to  interpret  him. 

Mr.  Heitland,  in  his  Roman  Eepuhlic,  published 
five  years  after  the  researches  of  Captain  Colin, 
says  that  '  no  certain  inferences  can  be  drawn  from 
the  loose  descriptions  of  Livy  or  even  Polybius  '.^ 

The  account  of  the  march  which  has  been  given 
may  perhaps  carry  conviction  to  some  minds.  But 
much  authority  is  commonly  attached  to  dis- 
tinguished names,  and  the  weight  which  is  rightly 
carried,  for  example,  by  the  opinion  of  a  scholar 
like  Mommsen  on  the  meaning  of  an  inscription, 
is  transferred  to  his  opinion  on  a  problem  of 
1  The  Roman  Refuhlic,  i,  p.  233. 


38  REJECTED  HYPOTHESES 

military  geography,  for  the  solution  of  which  he  had 
no  special  qualification.  It  therefore  seems  desir- 
able not  only  to  give  here  the  reasons  why  the 
passes  other  than  the  Clapier  must  be  uncon- 
ditionally rejected,  but  also  to  show  how  and  why, 
before  the  Clapier  had  been  suggested,  an  Arnold 
and  a  Mommsen  could  accept  the  Little  St.  Bernard, 
an  Ellis  the  Mont  Cenis,  a  Hennebert  the  Mont 
Genevre,  and  a  Freshfield  the  Col  de  Larche. 

The  first  witness  against  the  commonly  accepted 
passes  is  Thomas  Arnold  himself,  who  wrote:  'The 
account  of  Polybius  is  so  unscientific  and  so 
deficient  in  truth  and  liveliness  of  painting,  that 
persons  who  have  gone  over  the  several  Alpine 
passes  for  the  very  purpose  of  identifying  his 
descriptions  can  still  reasonably  doubt  whether 
they  were  meant  to  apply  to  Mont  Genevre  or 
Mont  Cenis  or  to  the  Little  St.  Bernard.'  ^ 

It  is  remarkable  that  Arnold  should  have  inferred 
from  these  doubts  that  Polybius  was  careless  or 
inaccurate,  instead  of  reaching  the  obvious  con- 
clusion that  no  pass  which  did  not  correspond 
to  the  account  of  Polybius  could  reasonably  be 
supposed  to  be  the  one  that  he  was  describing. 

The  principal  sources  of  error  arise  from  the 
neglect  to  ascertain  exactly  where  Polybius  indi- 
cates that  Hannibal  crossed  the  Rhone,  and  where 
he  expressly  says  that  he  came  down  on  to  the 
plain  of  Italy.  Polybius  says  in  a  passage  sum- 
marizing the  march,  that  Hannibal  having  made 
1  History  of  Rome,  iii,  p.  484. 


REJECTED  HYPOTHESES  39 

the  whole  journey  from  New  Carthage  in  five 
months,  and  the  passage  of  the  Alps  in  fifteen 
days,  came  down  boldly  to  the  plain  of  the  Po 
and  the  tribe  of  the  Insubres.  The  Insubres  lived 
in  the  neighbourhood  of  Milan,  east  of  the  Ticino, 
and  those  who  think  that  Polybius  means  that 
Hannibal  reached  the  plain  in  the  territory  of 
the  Insubres  imagine  that  such  a  description 
would  be  met  by  his  reaching  the  plain  at  Ivrea, 
where  the  route  of  the  Little  St.  Bernard  ends, 
although  Ivrea  was  not  in  the  territory  of  the 
Insubres  and  is  fifty  miles  west  of  the  Ticino. 
Mr.  Freshfield  thinks  he  has  complied  with  the 
text  of  Polybius  just  quoted  by  bringing  Hannibal 
down  among  the  Vagienni,  although  they  were 
a  hundred  miles  from  the  Insubres,  because  the 
Vagienni  were  of  Insubrian  descent.  But  the  pas- 
sage of  Polybius  is  a  summary,  and  in  it  the  Insubres 
are  mentioned  because  they  were  the  allied  tribe 
which  Hannibal  was  to  join.  In  the  sixtieth 
chapter  Polybius  clearly  and  distinctly  says  that 
as  soon  as  Hannibal  reached  the  plain  he  en- 
camped under  the  very  slope  of  the  Alps,  and 
that  the  Taurini  were  the  tribe  who  lived  beside 
this  slope.  The  interpretation  that  Hannibal  first 
touched  the  plain  in  the  territory  of  the  Insubres  is 
untenable,  because  no  pass  leading  from  the  Rhone 
valley  to  the  Alps  could  possibly  comply  with  it. 

The  evidence  against  the  Little  St.  Bernard 
is  sufficiently  accumulated  in  Thomas  Arnold's 
own  notes,  lor  Arnold  points  out  the  probability 


40  REJECTED  HYPOTHESES 

that  '  Hannibal  passed  by  some  much  higher 
point  than  the  present  roads  over  the  Little  St. 
Bernard  and  Mont  Cenis.'^  Again  Arnold  says: 
'I  lay  no  stress  on  the  "roche  blanche";  it  did 
not  strike  me  when  I  saw  it  as  at  all  conspicuous; 
nor  does  the  X^vKonerpov  mean  any  remarkably 
white  cliff,  but  simply  one  of  those  bare  limestone 
cliffs  which  are  so  common  both  in  the  Alps  and 
the  Apennines.'  Arnold  had  not  seen  the  Rocher 
de  la  Porte,  and  ignored  Polybius's  word  6xvp6t/, 
*like  a  fortress.'  The  Little  St.  Bernard  has  no 
view  of  Italy,  and  for  the  difficult  place  on  the 
descent  Arnold  fixes  on  the  river  gorge  below 
La  Thuile,  where  the  road  or  path  follows  the  side 
of  a  steep  cliff  at  a  height  of  twenty  or  thirty 
feet  above  the  level  of  the  stream.  This  gorge 
would  not  have  long  delayed  an  army  of  20,000 
men,  who  would  easily  have  made  a  causeway, 
if  need  be,  on  one  side  or  the  other  in  the  bed  of 
the  stream  itself.  No  army  could  march  from 
the  gorge  of  La  Thuile  to  the  plain  at  Ivrea  in 
three  days. 

As  regards  the  crossing  of  the  Rhone,  the  modern 
historians  have  been  put  on  the  wrong  track  by 
Napoleon,  who  wrote  at  St.  Helena,  '  Hannibal 
crossed  the  Rhone  above  its  confluence  with  the 
Durance  and  below  its  confluence  with  the 
Ardeche.  He  passed  above  the  confluence  of  the 
Durance  because  he  did  not  wish  to  direct  his 
march  upon  the  Var  ;    he  passed  below  the  con- 

1  Vol.  iii,  p.  485. 


REJECTED  HYPOTHESES  41 

fluence  of  the  Ardeche  because  there  commences 
the  chain  of  mountains  which  almost  precipitously 
commands  the  right  bank  of  the  Rhone  as  far  as 
Lyons,  while  the  valley  on  the  left  bank  is  several 
leagues  wide.  From  the  mouth  of  the  Rhone  to 
its  confluence  with  the  Ardeche  is  twenty-eight 
leagues ;  it  is  probable  that  Hannibal  crossed 
four  leagues  lower  down  at  the  level  of  Orange, 
twenty-four  leagues  or  four  days'  march  from  the 
sea.' 

This  is  Napoleon's  theory  of  what  Hannibal 
would  be  likely  to  do,  and  Napoleon  further  relied 
on  the  four  days'  march  from  the  sea.  Napoleon 
did  not  know  Greek,  and  can  hardly  have  had 
before  him  a  translation  of  Polybius  close  enough 
to  show  that  that  historian  represented  Hannibal  as 
crossing  at  the  head  of  the  delta ;  nor  did  he 
observe  that  the  four  days'  march  was  that  of 
Scipio,  and  if  it  had  taken  him  to  Orange,  must 
have  taken  him  across  the  Durance,  which  he 
dared  not  have  attempted  to  pass  in  the  neigh- 
bourhood of  a  superior  enemy.  Hannibal  crossed 
the  Rhone  with  something  like  38,000  infantry  and 
8,000  cavalry  {Polybius^  iii.  60).  Scipio's  army 
was  22,000  infantry  and  2,000  cavalry. 

All  the  writers  from  Napoleon  onwards  until  Colin 

have  made  Hannibal  cross  the  Rhone  above  the 

confluence  of  the  Durance.     They  fail  to  see  that 

Livy's  description  of  the  passage  of  the  Durance  is 

perfectly  appropriate  though  slightly  in  the  wrong 

place,    and    are    therefore    compelled    to    make 

s 


42  REJECTED  HYPOTHESES 

Hannibal  cross  the  Durance  in  its  upper  course. 
By  fixing  the  crossing  of  the  Rhone  too  far  north 
they  are  led  to  look  for  the  Island  much  too  high 
up  the  Rhone  valley.  This  accounts  for  the  con- 
jecture that  the  river  whose  confluence  with  the 
Rhone  formed  the  Island  was  the  Isere,  and  for  the 
emendation  of  Livy's  text  to  suit  that  conjecture. 

Napoleon  thought  that  Hannibal  would  take 
the  shortest  route  through  the  Alps,  which  was 
either  that  up  the  vaUey  of  the  Arc,  eventually 
leading  to  Susa,  or  that  following  the  road  which 
Napoleon  himself  rebuilt,  leading  from  Grenoble 
by  Bourg  d'Oisans,  the  Lauteret,  BrianQon,  and  the 
Mont  Genevre  to  Susa. 

In  point  of  distance  there  is  little  to  choose 
between  these  two  routes.  But  that  by  the 
Lauteret  could  hardly  have  been  possible  for  an 
army  before  the  days  of  road-making.  It  threads 
a  number  of  exceedingly  difficult  gorges  and 
crosses  the  high  pass  of  the  Lauteret..  The 
Mont  Genevre  has  no  view  of  Italy,  no  glacier 
near  it,  no  such  difficulties  in  the  descent  as  are 
described  by  the  ancient  authors,  and  is  more 
than  three  days'  march  from  the  plain.  The  Mont 
Cenis  route  is  identical  with  that  of  the  Clapier 
as  far  as  Bramans,  but  from  Bramans  to  Susa 
it  is  considerably  longer,  and  nowhere  offers  such 
a  view  of  the  plain  of  Italy  as  is  described  by 
Polybius  and  Livy. 

The  hypothesis  that  the  Island  was  the  Rhone 
valley  above  the  Isere  is  inconsistent  with  most 


REJECTED  HYPOTHESES  43 

of  the  statements  made  by  Polybius  about  the 
Island.  The  Island  was  especially  fertile,  which 
is  not  the  quality  of  the  land  north  of  the  Isere. 
The  mountains  do  not  even  in  appearance  form 
the  base  of  a  triangle,  of  which  the  two  rivers  are 
the  two  other  sides.  Polybius  distinctly  says  that 
the  mountains  forming  the  base  of  the  triangle 
were  impassable,  and  yet  many  modern  writers 
make  Hannibal  cross  these  very  mountains. 
Mommsen  and  Arnold  put  the  first  battle  at  the 
Mont  du  Chat,  but  this  is  a  single  ridge  which 
would  be  perpendicular  to  Hannibal's  route.  Once 
Hannibal  by  surprise  had  gained  possession  of  any 
part  of  the  ridge,  he  would  have  been  able  to 
drive  off  the  enemy  from  any  point  of  command, 
so  that  the  subsequent  disaster  could  not  have 
occurred  here. 

Mr.  Freshfield  ^  proposes  as  Hannibal's  pass  the 
Col  d'Argentiere  or  Col  de  Larche.  He  assumes 
the  Island  to  be  the  country  between  the  Rhone 
and  Isere.  Apparently  he  thinks  that  Hannibal 
did  not  enter  this  Island,  but  started  from  Valence 
and  went  by  Vizille  and  Corps  to  the  Col  Bayard, 
then  crossed  the  Durance  and  marched  up  its  left 
b^nk  to  Guillestre,  then  crossed  the  Col  de  Vars 
to  St.  Paul,  and  so  made  his  way  up  to  the  Col  de 
Larche.  Mr.  Freshfield  apparently  thinks  that 
Livy  and  Polybius  are  describing  different  routes. 
He  rejects  Polybius,  and  professes  to  follow  Livy. 
When  Livy  says  that  after  settling  the  dispute 
1  The  Alpine  Journal,  vol.  xi,  1884,  p.  267. 


44  REJECTED  HYPOTHESES 

in  the  Island,  Hannibal  turned  to  his  left,  Mr.  Fresh- 
field,  who  has  made  Hannibal  march  up  the  Rhone 
to  Valence,  makes  him  go  off  towards  Grenoble  or 
Vizille.  As  to  the  direction,  Mr.  Freshfield  is  satis- 
fied if  Hannibal  had  on  his  right  '  the  direct  road 
to  Italy ',  and  it  is  quite  clear  to  him  that  this 
road  went  up  the  valley  of  the  Drome.  Mr.  Fresh- 
field  puts  the  first  battle  at  the  Col  Bayard.  He 
therefore  rejects  the  statement  of  Livy  that  Hanni- 
bal met  with  no  opposition  before  he  reached  the 
Durance.  Livy  further  says  that  from  the 
Durance  onwards  Hannibal  marched  across  a  plain 
to  the  Alps,  but  there  is  no  plain  whatever  in  the 
upper  vaUey  of  the  Durance  or  between  it  and 
the  Alps,  nor  is  it  intelligible  that  on  this  theory 
Hannibal  is  not  supposed  to  reach  the  Alps  until 
aiter  he  has  crossed  the  Durance.  Surely  Grenoble, 
Vizille,  and  Corps  are  all  of  them  in  the  Alps. 
Mr.  Freshfield,  like  Arnold,  thinks  that  a  '  white 
rock '  can  be  found  anywhere.  He  cannot  bring 
himself  to  include  among  the  requirements  a  pano- 
rama of  Italy.  He  places  the  difficulty  of  the 
descent  at  the  Barricades,  a  defile  through  which 
the  Stura  flows,  surmounted  on  one  side  by 
a  high  perpendicular  cliff  and  on  the  other  by 
the  gentle  spur  of  a  hill  which  falls  precipitously 
to  the  stream.  The  old  road  went  over  the  spur 
of  the  hill,  which  presents  no  difficulties  whatever, 
nor  would  there  be  any  difficulty  in  marching  an 
army  through  the  defile  itself  beside  the  bed  of  the 
stream.     The  sole  difficulty  of  the  defile  of  the 


REJECTED  HYPOTHESES  45 

Barricades  in  war  is  that  which  has  sometimes 
been  caused  by  an  enemy's  resistance.  But  no 
resistance  was  offered  to  Hannibal  on  his  descent. 
There  is  on  the  descent  from  the  Col  de  Larche 
no  possibility  of  men  or  mules  falling  over  a 
precipice.  There  is  no  glacier  near  it,  and  no 
place  whatever  where  snow  can  remain  from  the 
year  before. 

The  truth  is  that  Mr.  Freshfield  has  not  been 
guided  either  by  Polybius  or  Livy,  the  statements 
of  both  of  whom  he  brushes  aside  wholesale.  His 
case  rests  entirely  on  a  passage  quoted  from  Varro 
by  a  fii th-century  grammarian  as  follows  :  '  Varro 
says  that  the  Alps  can  be  crossed  by  five  routes  : 
one  by  the  sea  through  the  Ligurians  ;  a  second 
by  which  Hannibal  crossed ;  a  third  by  which 
Pompeius  set  out  to  his  war  in  Spain  ;  a  fourth  by 
which  Hasdrubal  came  from  Gaul  into  Italy ; 
a  fifth  which  was  formerly  held  by  Greeks,  so  that 
the  Alps  through  which  it  passes  are  called  Greek  ' 
(i.  e.  the  Graian  Alps).  Mr.  Freshfield  assumes 
that  these  five  passes  correspond  to  the  five 
modern  high-road  passes :  the  Corniche,  the  Col  de 
Larche,  the  Mont  Genevre,  the  Mont  Cenis,  and 
the  Little  St.  Bernard.  He  assumes  that  Varro 
enumerates  them  in  order  from  south  to  north, 
that  Hasdrubal  made  use  of  the  Mont  Cenis  and 
Pompeius  of  the  Mont  Genevre,  and  that  the 
first  is  the  Roman  road  along  the  coast.  Mr. 
Freshfield  therefore  infers  that  Varro  believed 
Hannibal  to  have  crossed  by  the  Col  de  Larche. 


46  REJECTED  HYPOTHESES 

Between  Mont   Blanc   and  the   sea  there    are 
more  than  thirty  passes  crossed  by  regular  mule- 
tracks  or  by  footpaths.     During  the  eighteenth  and 
nineteenth  centuries  high  roads  have  been  made 
over  four  of  them,  the  Little  St.   Bernard,  the 
Mont  Cenis,  the  Mont  Genevre,  and  the  Col   de 
Larche.     In  Hannibal's  time  there  was  of  course 
no  high  road.    He  had  to  choose  one  of  the  numer- 
ous  mule-tracks.     The    early   nineteenth-century 
historians,  however,  assumed  that  he  must  have 
gone  by  one  of  the  high-road  routes,  and  confined 
their  attention  to  these.     This  is  the  explanation 
of  the  various  conflicting  hypotheses  which  they 
set  up.     The   high-road  routes  before   the   high 
roads  were  made  were  not  necessarily  more  con- 
venient than  the  others  for  armies  on  the  march. 
To  the  pedestrian  it  is  more  important  that  the 
length  of  his  pass  should  be  comprised  within 
a  single  day's  walk,  than  that  it  should  be  of 
moderate  height.     One  of  the  most  frequented 
passes  for  pedestrians,  especially  for  Italian  work- 
men seeking  employment  in  France,  is  the   Col 
d'Agnel,    which   is   9,000   feet   high.      The   road 
engineer  is  indifferent  to  distance,  and  seeks  a  gentle 
slope,  which,  if  it  is  not  offered  him  by  nature,  he 
creates  by  endless  zigzags. 

In  the  exposition  here  given  care  has  been  taken 
to  avoid  a  priori  theories  of  what  Hannibal  would 
be  likely  to  do,  but  it  may  now,  in  conclusion,  be 
pointed  out  that  the  route  which  has  been  shown 
to  conform  to  the  accounts  of  the  ancient  historians 


HEJECTED  HYPOTHESES  47 

is  also  that  which  would  be  commended  to  Hamiibal 
by  purely  military  considerations. 

The  effect  of  mountainous  country  on  the  move- 
ment of  an  army  is  of  three  kinds.  It  adds  to  the 
fatigue  of  the  troops,  it  adds  to  the  difficulties  of 
supply,  and  it  adds  to  the  dangers  which  may  be 
caused  by  armed  opposition.  In  a  populous  and 
fertile  plain  such  as  the  valley  of  the  lower  Rhone 
an  army  of  50,000  men  on  the  march  would  find 
plenty  of  food,  though  if  it  remained  long  in  one 
place  it  might  produce  a  famine.  But  in  the 
Alpine  region  the  population  is  sparse,  food  scarce, 
and  the  stores  of  it  limited.  No  general  would 
plunge  into  such  a  country  unless  he  expected  to 
carry  with  him  supplies  for  his  army  during  the 
whole  of  its  passage  through  the  mountain  region. 
An  army  on  the  march  always  loses  men  from 
fatigue,  and  every  increment  of  fatigue  causes  the 
loss  of  an  increasing  proportion  of  its  men.  The 
best  way  to  keep  down  losses  from  this  cause  would 
be  to  make  the  fatiguing  marches  as  few  as  possible; 
in  other  words,  to  choose  the  shortest  way  through 
the  mountain  region.  If  the  army  must  carry  its 
food  through  the  mountains,  the  shorter  the  route 
through  them  the  smaller  the  number  of  days 
for  which  the  supply  will  be  needed,  and  the  less, 
therefore,  the  burden  to  be  carried.  Moreover,  the 
shorter  the  line  taken  through  the  mountains 
the  shorter  the  time  during  which  the  army  would 
be  exposed  to  the  peculiar  dangers  which  might  be 
caused  by  armed  opposition. 


f 


48  REJECTED  HYPOTHESES 

The  route  from  St.  Nazaire  byMontaud,Noyarey, 
Grenoble,  the  valley  of  the  Arc,  and  the  Col  du 
Clapier  to  Susa  and  Avigliana,  is  both  the  shortest 
route  through  the  Alps  from  France  to  Italy  and 
the  route  which  offers  the  fewest  and  least  for- 
midable defiles.  It  coincides,  except  in  the  actual 
pass,  with  the  route  which  commended  itself  more 
to  Napoleon's  military  judgement  than  any  other 
except  one,  along  which  he  had  himseK  found  it 
necessary  to  reconstruct  the  military  road. 


Oxford :  Printed  at  the  Clarendon  Press  by  Horace  Hart,  M.A. 


[ 


^ 


^ 




1 

UNIVERSITY  OF  TORONTO 

! 

^  > 

LIBRARY 

Lithor 
tie 

/ 

r 

:^ 

^^v 

t» 

H« 

^^^ 

iS 

;M 

>^> 

F 

^' 

\  \     1^  P    not 

'  ~- 

H 
m 

is 

W    remove 

u 

i>    V 

1 

3 

-, 

the    card 

09 

;■/   // 

D 

//  from    this 

— ^ 

- 

£± 

00 

'    // 

5 

tr 

o> 

^/   Pocket. 

i 

i 

on 

--- 

^        I 

j^ 

^ 

ts 

ij 

^/ 

• 

! 

Acme  Library  Card  Pocket 

Under  Pat.  "Ref.  Index  File." 

Made  by  LIBEARY  BUEEAU,  Boston 

i 

/ 

Is 

I 

I 

t 

i 

I Ji