308  The  Wolntra  Pot-Culture  Kxperiments,  1003. 
were,  as  usual,  selected  so  as  to  be  as  nearly  as  possible  equal 
in  weight,  and  that  there  should  be  the  same  number  of 
“ starchy  ” and  “ glutinous  ” seeds  in  each  pot.  The  various 
ai)plications  were  manganese  ioilide,  potassium  iodide,  sodium 
iodide,  anti  lithium  iodide,  0'769  grammes  of  each  to  a pot — 
equivalent  to  1 cwd.  per  acre — and  of  the  oxides  of  the  same 
metals,  1‘538  grammes  of  each  to  a pot — equivalent  to  2 cwt. 
per  acre.  The  iodide  of  manganese  used  was  the  manganous 
salt  (Mnlo),  and  the  oxide,  the  red  oxide  (MuaOj). 
The  first  appearance  of  note  was  that  where  oxide  of  lithium 
had  been  applied,  the  surface  soil  became  hard  and  “ caked  ” ; 
still,  the  young  shoots,  though  kept  back  at  first,  forced  their 
way  quite  well  through  this,  and  ovei’  the  series  generally  there 
was  vei-y  fair  germination  and  no  very  marked  failure.  On 
summing  up  the  germination  it  was  found  that,  by  the  end  of 
January,  the  lowest  germination  (83‘3  per  cent.)  had  been  given 
by  the  iodide  of  manganese,  the  best  by  the  iodide  of  lithium 
(100  per  cent.).  The  “ caking  ” of  the  soil  with  oxide  of 
lithium  continued  to  be  noticed  throughout  April,  1003,  but 
the  plants,  though  retarded,  seemed  quite  healthy.  With 
iodide  of  manganese,  however,  and  iodide  of  potassium,  the 
plants  seem  to  have  been  affected,  they  not  l)eing  nearly  so 
robust  as  the  untreated  ones.  A month  later  the  whole  of  the 
iodide  series  showed  the  ill  effects  of  the  treatment,  the  plants 
turning  yellow  and  looking  poor  and  weakly.  The  contrast 
between  these  and  the  oxide-treated  ones  was  very  marked,  the 
latter  all  looking  well,  especially  the  lithium  oxide  ones,  these 
coming  into  ear  first.  Of  the  iodide  series,  the  worst  was 
manganese  iotlide,  this  lot  being  very  slow  in  ripening.  The 
measurements  of  straw  and  ear  were  taken  early  in  September, 
1003,  and  the  plants  were  photographed,  shortly  after  which 
the  corn  was  harvested.  The  appearances  of  the  plants  are 
indicated  by  Fig.  1,  page  307,  and  the  crop  residts  are  given 
in  Tal)le  I.,  page  310. 
The  differences  between  the  iodide  Series  and  the  oxide  series 
are  verj='  clearly  brought  out.  In  every  instance  of  the  ioilides 
the  results  in  corn  yield  came  below  those  of  the  untreated. 
As  had  been  indicated  during  the  period  of  growth,  manganese 
iodide  was  much  the  worst,  while  sodium  iodide  and  lithium 
iodide  were  the  best,  potassium  ioditle  giving  a better  return 
than  the  manganese  salt,  though  still  only  a half-crop.  The  only 
application  to  affect  the  straw  favourably  was  sodium  iodide, 
and  that  not  materially. 
It  may  here  be  recalled  that  in  the  experiments  of  1002, 
when  the  salts  were  used  at  the  rate  of  2 cwt.  per  aci’e 
in  solution  (and  not,  as  here,  in  the  solid  state),  manganese 
iodide  had  a similarly  bad  effect,  the  crop  only  being  a half 
