340  Nitrogen  as  Nitric  Acid , in  the  Soils 
question  contained  little  or  no  nitrate,  and  that  from  this  time 
forward  the  formation  of  nitrate  was  more  or  less  active.  The 
first  running  of  the  drain-pipes  in  Broadbalk  Wheat-field 
informs  us  when  drainage  may  be  supposed  to  have  commenced 
in  the  soils  in  question.  The  composition  of  the  drainage- 
water  from  Plot  6 in  the  same  wheat-field  will  furnish  a very 
moderate  estimate  of  the  strength  of  this  drainage-water  through 
the  autumn  and  winter  months,  while  the  measured  drainage 
from  the  drain-gauges  will  give  a fair  estimate  of  its  quantity. 
After  March  1,  the  mean  amount  of  nitrate  passing  away  in  the 
drainage  from  the  20  and  40-inch  drain-gauge  soils  will  furnish 
a low  estimate  of  the  loss  suffered  during  the  spring  and  summer 
months.  These  estimates  lead  to  the  following  figures  : — 
Season 
1880-1. 
Season. 
1881-2. 
Nitrogen  as  nitrates  in  drainage,  August  to  February 
„ „ March  to  Sept,  or  Oct. 
„ in  soil,  September  or  October 
Lbs. 
17-0 
14-8 
57-7* 
Lbs. 
19-0 
7-6 
59-9 
Total  production  of  nitrogen  as  nitrates 
89-5 
86-5 
* Mean  of  the  quantities  found  in  Clay  Croft  and  Foster’s  fields. 
It  may  be  fairly  urged  that  these  estimates  of  the  quantity  of 
nitrates  formed  per  acre  during:  a season  of  bare  fallow  must  err 
on  the  side  of  excess,  as  they  assume  the  existence  of  absolutely  no 
nitrate  in  the  soil  at  the  commencement  of  the  fallow,  and  also 
take  no  account  of  any  drawing  up  of  nitrates  from  the  subsoil 
during  dry  weather,  by  which  some  of  the  previous  loss  to  the 
upper  layers  may  be  recovered.  On  the  other  hand,  the  esti- 
mates of  the  loss  by  drainage  are  intentionally  kept  extremely 
moderate.  On  the  whole,  the  evidence  at  our  command  leads 
to  the  conclusion  that  the  quantity  of  nitrates  produced  in  one 
of  the  Rothamsted  soils  in  good  agricultural  condition,  culti- 
vated as  bare  fallow,  will  be  equal  to  about  80  lbs.  of  nitrogen 
per  acre.  The  time  available  for  nitrification  in  a season  ot 
fallow  is,  however,  greater  than  a year  ; it  actually  amounted 
to  about  15  months  up  to  the  time  of  soil-sampling  in  the  season 
1880-1,  and  to  about  14  months  in  1881—2.  Of  the  quantity 
of  nitrates  produced  in  extremely  dry  seasons  we  have  as  yet  no 
experience. 
We  have  spoken  above  of  the  effect  of  bare  fallow  in  soils  in 
good  agricultural  condition.  In  Table  III.  (Nos.  5 and  9) 
will  be  found  determinations  of  nitrates  in  two  soils  of  very 
low  agricultural  condition,  which  had  been  cultivated  as  bare 
fallow  in  1878.  The  quantity  of  nitrogen  as  nitrates  found  was 
