Pedigree  Seed  Corn. 
J31 
The  multiplied  nursery  plot  experiment  was  repeated  in 
1909  with  substantially  the  same  result,  and  field  trials  also 
gave  confirmatory  results.^ 
The  explanation  of  the  apparently  contradictory  result  as 
between  alternate  row  and  alternate  plot  planting  in  this 
case  is  both  obvious  and  instructive,  as  bearing  on  the 
operation  of  natural  selection  in  cases  of  mixed  seeding. 
The  alternate-row  planting  in  1907  may  be  taken  as  a case 
of  mixed  seeding.  The  effect  evidently  was  to  crowd  down 
the  race,  which,  when  sown  unmixed,  was  the  most  prolific. 
“ Tystofte  ” has  about  a week  longer  growing  period  in  an 
average  season  than  the  wide  ear,  but  is  also  four  to  six  inches 
shorter  in  the  straw.  There  is  no  doubt  that  in  this  case  there 
was  crowding  down  of  the  slower  growing  and  ripening  variety 
by  the  taller-strawed,  quicker-growing  (though  normally  less 
productive)  variety,  which  over-topped  and  shaded  the  other 
in  the  grain-forming  stage. 
There  is,  in  fact,  afforded  an  illustration  in  respect  of 
varieties  of  what  would  be  generally  accepted  with  regard 
to  the  effect  of  natural  selection  in  the  case  of  species,  viz., 
that  in  a population  of  mixed  species  of  different  habits,  it 
will  not  necessarily  follow  that  the  species  which  reproduce 
most  abundantly  when  living  apart  will  survive  under 
conditions  of  free  competition  with  other  species. 
The  foregoing  illustrations  seem  to  warrant  the  proposition 
that  in  order  to  determine  an  “order  of  merit”  of  any  more 
than  a quite  temporary  value  as  between  aggregates  of  seed 
corn,  comparison  must  be  made  between,  at  least  approxi- 
mately, pure  races.  Aggregates  of  fiuctuating  composition  are 
in  a state  'of  unstable  equilibrium,  and  comparisons  between 
them  give  no  results  of  general  utility. 
We  may  conclude  this  part  of  the  subject  by  saying  that 
the  whole  problem  of  mixed  versus  pure  races  of  cereals  is 
very  much  on  a par  with  the  same  problem  in  respect  of  farm 
live-stock. 
Pedigree  stock  is  best  if  the  average  external  conditions 
of  the  locality  are  more  favourable  to  the  particular  race 
than  to  others,  but  except  under  these  conditions  may 
easily  be  less  profitable  than  an  aggregate  made  up  of  mixed 
breeds. 
* I have  some  evidence  to  show  that  series  of  small  scale  nursery  cultiva- 
tions under  precise  conditions,  supplemented  by  biometrical  methods  of 
interpreting  the  significance  of  the  results  obtained,  may  turn  out  to  be  of 
value  for  determining  the  relative  prolificacy  of  new  races  of  cereals.  The 
method  is  obviously  more  economical  than  that  of  field  plots  for  comparing 
large  numbers  of  new  rhces,  such  as  the  progeny  of  hybrids.  I hope  shortly 
to  publish  some  account  of  the  cultivations  carried  out  at  Warminster,  with 
a view  to  systematising  this  method. 
K 2 
