FORMATION  OF  GUM  TRAGACANTH. 
243 
INQUIRIES  INTO  THE  FORMATION  OF  GUM  TRAGACANTH. 
By  Hugo  von  Mohl. 
(Translated  from  the  Botanische  Zeitung,  Jahrg.  XIV.,  p.  33,  1857,  by  BerthoM 
Seeman,  Ph.  D.,  F.  L.  $.,  with  corrections  by  the  Author.) 
4  Concluded  from  page  165.) 
After  this  examination  of  the  tragacanth  gum,  I  proceeded 
to  that  of  the  stems,  extending  it  to  those  of  the  following  spe- 
cies belonging  to  the  section  Tragacanthce,  viz:  Astragalus  an- 
qustifoUus,  Lam.,  A.  aristatus,  L'Herit.,  A.  Anacantha,  M.  B., 
A.  aureus,  Willd.,  A.  Barba  Jovis,  De  Cand.,  A.  breviflorus, 
De  Cand.,  A.  bunophilus,  Boiss.,  A.  campylanthus,  Boiss.  A, 
Caucasicus,  M.  B.,  A.  cephalanthus,  De  Cand.,  A.  chromolepis, 
Boiss.,  A.  compactus,  Willd.,  A.  creticus,  Lam.,  A.  cyllenius, 
Boiss.  et  Heldr.,  A.  denudatus,  Stev.,  A.  echinoides,  L.  Herit., 
A.  echinus,  De  Cand.,  J.,  erianthus,  Willd. ,A.  gossyspinus  Fisch., 
A,  lagopodiodes,  Vahl.,  A.  leiocladus,  Boiss.,  A.  massiliensis, 
Linn.,  A.  mierophgsa,  Boiss.,  A.  murinus,  Boiss.,  A.  persicus, 
Fisch.  et  Meg.,  A:  plumosus,  Willd.,  A.  pseudotragacantha,  M. 
B.  ,  A.  ptychophyllus,  Boiss.,  A.  pycnocephalus,  Fisch.,  A.  turni- 
dus  Willd.  Amongst  these  there  were  only  four  species  in  the 
stem  of  which  no  tragacanth  formation  could  be  discovered,  viz., 
A.  aristatus,  L'Herit.  (from  the  Pyrenees,)  A.  massiliensis, 
Linn.,  A.  angustifolius,'Ln,m.,  and  A.  echionoides,  L'He'rit.  In 
all  the  others,  the  formation  of  gum  tragacanth,  in  a  more  or 
less  abundance,  was  apparent.* 
The  structure  of  the  stems  is  in  general  as  follows : — The 
wood  consists  of  thin  annual  layers,  and  is  extremely  tough,  easi- 
ly tearing  lengthways  into  thin  filaments.  It  incloses  a  thin 
column  of  pith  [em  enges  Marh^,  is  crossed  by  numerous  medul- 
lary rays,  and  does  not  exhibit  anything  unusual.  The  same 
may  be  said  of  the  bark  which  contains  perfectly  formed  liber, 
and  is  covered  by  a  strong  tough  periderm.  But  the  pith  and 
a  great  part  of  the  medullary  rays  are  very  curious  ;  for,  in- 
stead of  presenting  a  thin-walled  parenchymatous  texture,  they 
*  The  question  whether  the  transformation  of  the  cells  into  gum  tragacanth 
to  be  described  below,  also  takes  place  in  species  belonging  to  other  sections  of 
thegenus  I  have  not  specially  investigated;  but  I  may  remark  that  I  met  with 
the  same  transformation  in  two  species  belonging  to  the  section  Incani,  which 
I  selected  at  random,  viz.,  A.  brachycarpus,  M.  B.,  and  A.  a?iffulosus,  D.  C, 
