252 
Hybridization  in  Plants. 
(  Am.  Jour.  Pharm. 
I     April,  1921. 
appear  on  both  epidermises.  The  leaves  of  hybrids  (M.  ceri- 
fera  x  M.  Carolinensis)  show  numerous  orange-red,  bowl-shaped 
and  golden-yellow,  balloon-shaped  glands  on  the  lower  epidermis, 
but  often  fewer  than  on  the  similar  epidermis  of  M.  cerifera.  The 
upper  epidermis  shows  a  scattering  of  orange-red,  bowl-shaped 
glands  and  golden-yellow,  balloon-shaped  glands,  while  the  number 
of  non-glandular  hairs  on  both  epidermises  are  fewer  than  on  M. 
Carolinensis  and  more  numerous  than  on  M.  cerifera.  The  ten- 
dency for  the  head  of  the  bowl-shaped  gland  in  the  hybrid  to  be- 
come saucer-shaped  is  very  striking. 
Further,  as  to  hardiness  of  hybrids  in  relation  to  parents, 
the  hybrid  is  very  evenly  intermediate  between  the  two  parent 
species.  For  instance,  our  Mountain  Rhododendron  (Rhododen- 
dron maximum)  resistant  to  frost,  when  crossed  by  the  scarlet 
Himalaya  Rhododendron  (Rhododendron  arboreum)  tender  to 
frost,  gives  a  hybrid  that  will  resist  15  deg.  to  20  deg.  above  zero. 
Again,  in  the  case  of  the  Myricas,  the  leaves  of  Myrica  ceri- 
fera (Wax  Myrtle)  are  evergreen  and  remain  green  on  the  stems 
over  winter  until  late  April  and  early  May  without  assuming  a 
coppery  color.  Those  of  Myrica  Carolinensis  (Bayberry)  are 
deciduous,  assuming  a  greenish-brown  hue  in  October  and  Novem- 
ber and  usually  completely  fallen  by  the  middle  of  December, 
while  the  leaves  of  hybrids  between  these  two  parents  are  semi- 
evergreen  and  usually  fall  in  February  and  March,  by  which  time 
they  have  often  assumed  a  slight  coppery  tint. 
The  plant  breeder,  in  starting  any  work  in  hybridization,  is 
obliged  to  choose  the  varieties  or  species  to  be  hybridized.  He 
must  have  some  definite  aim  in  view.  He  must  study  his  plants 
and  determine  their  good  and  weak  characters.  These  characters 
may  be  tallness,  dwarfness,  high  or  low  percentage  of  active  con- 
stituents, color  of  leaves,  flowers,  or  other  organs,  hairiness  or 
smoothness,  starch  or  sugar  content  of  seeds,  etc.  He  must  then 
use  his  judgment  in  deciding  what  combinations  of  these  characters 
would  form  the  best  variety. 
Unfortunately  the  reaction  between  different  characters  or  fac- 
tors on  one  another  in  various  combinations  has  not  always  been 
what  was  expected.  Up  to  the  present,  the  followers  of  the  Men- 
delian  theory  have  been  unable  to  make  a  complete  analysis  of  all 
the  unit  characters  or  factors  in  any  plant  with  which  they  have 
worked.  While  they  are  generally  able  to  predict  what  will  happen 
when  pure  strains  or  varieties  of  plants,  having  up  to  five  pairs  of 
