Am.  Jour.  Pharru. ) 
September,  1900.  j 
AtmospJieric  Ozone. 
425 
When  it  comes  to  actual  quantitative  work,  the  statistics  are 
even  more  bewildering.  Most  careful  examination  of  the  literature 
brought  to  light  but  three  writers  who  commit  themselves  to 
figures,  and  in  each  case  the  data  is  strikingly  dissimilar. 
Houzeau  [Ann.  Chim.  Phys.y  XXVII,  5)  states  that  the  maximum 
of  ozone  in  the  atmosphere  is  1  part  to  450,000  by  weight  (or  0-28 
milligramme  to  100  litres  air).  Schone  (Brochure,  Moscow,  »i897) 
gives  amount  as  varying  from  1  to  10  milligrammes  to  100  litres 
air;  while  H.  de  Varigny  (Smithson.  Miscell.  Coll.,  XXXIX,  27) 
says  that  the  average  is  1  milligramme  and  the  maximum  is  3^ 
milligrammes  to  100  cubic  metres  air.  The  latter  statement,  which 
means  1  to  3^  milligrammes  to  100,000  litres,  is  at  such  variance 
with  the  other  figures  that  we  can  only  consider  it  as  a  typographi- 
cal error. 
These  references  seem  to  confirm  us  in  the  belief  that  we  are 
among  the  pioneers  in  the  field  of  atmospheric  ozone  assay,  and, 
while  conscious  of  the  liability  to  err,  our  results  are  given  in  the 
hope  of  leading  to  further  observations  not  open  to  the  objections 
to  which  Schonbein's  paper  is  subject.  Most  strongly  is  it  hoped 
that  the  national  government  may  take  up  this  important  work,  as  it 
is  an  undertaking  almost  beyond  the  scope  of  a  single  observer. 
Two  methods  of  ozone  assay  were  employed  :  Hartley's,  in  which 
the  arsenite  is  oxidized  to  arsenate  by  the  following  reaction : 
KAs02  -j-  03  =  KAs03  -f  02 ;  second,  Houzeau's,  in  which  potas- 
sium iodide  is  oxidized  to  iodate,  by  the  following  method :  KI  -f 
303  =  KIO3  +  302. 
The  solution  of  potassium  arsenite  (corresponding  to  3  grammes 
arsenite  to  1  litre)  was  prepared  by  heating  in  water  1-966  grammes 
arsenious  acid  and  1-465  grammes  potassium  carbonate,  bringing  fin- 
ished solution  up  to  1,000  c.c.  Of  this,  portions  of  20  c.c.  were 
placed  in  glass  stoppered  bottles  for  the  ozone  test,  enough  space 
being  left  in  each  bottle  for  the  addition  of  wash  liquid. 
The  solution  of  potassium  iodide  was  made  by  dissolving  100 
grammes  of  the  iodide  in  enough  water  to  make  1,000  c.c,  and  por- 
tions of  20  c.c.  were  placed  in  bottles,  as  in  the  case  of  the  arsenite. 
These  bottled  solutions  were  sent  to  Covington,  La.,  and  the  passage 
of  air  through  each  was  performed  around  a  hotel  situated  on  the 
bank  of  a  small  stream,  about  a  mile  from  the  village.  A  signifi- 
cant difference,  however,  lay  in  the  fact  that,  in  February,  the 
