AmApr5r'i9£f™'}^w  Agent,  in  Pulmonary  Tuberculosis.  175 
After  completing  a  second  sanatorium  course  returned  home  and 
for  fourteen  months  previous  to  treatment  had  been  confined  to  bed. 
During  treatment  toxaemia  was  reduced,  there  were  no  hyperidroses, 
and,  after  the  fourth  month,  no  rigors.  He  was  able  to  be  dressed 
and  about  his  rooms,  and  to  do  something  for  himself..  The  enormous 
cavities  continued  large  daily  secretions  and  were  a  constant  source 
of  exhaustion.  A  permanently  favorable  result  could  scarcely  be 
expected.  Tubercle  bacilli  greatly  diminished  in  number,  and  all 
secondary  forms  disappeared.  Improvement  continued  and  he  was 
able  to  return  to  business  for  part  of  the  day. 
The  results  of  treatment  may  be  indicated  as  follows : 
Stage  I.  Stage  II.  Stage  III. 
Unimproved   1 
(Case  III) 
Improved  1  2  9 
(Case  V)  (Cases  XVI,  XVII)  (Cases  I,  VII,  VIII,  IX,  XII, 
XV,  XVIII,  XIX,  XX) 
Arrested   ..  .  ;\v.  ......... .   2  1 
(Cases  XIII,  XVI)  (Case  VI) 
Died   ,   4 
(Cases  II,  IV,  X,  XI) 
Case  XXI  was  markedly  improved. 
This  classification  is  conservative,  as,  of  the  Stage  II  cases,  one 
(Case  XVII)  reported  as  improved  was  in  a  condition  approaching 
arrest,  but  would  not  remain  kmger  under  observation,  and  of  the 
Stage  III  cases,  three  (Cases  XII,  XV,  and  XIX)  reported  as 
improved  were  also  in  a;  condition  approaching  arrest,  and  two  of 
them  have  returned  to  employment. 
Accurate  studies  of  the  blood  of  tuberculous  patients  are  of 
comparatively  recent  date.  While  many  observations  are  recorded 
in  the  literature,  they  are  confused  by  a  lack  of  agreement  among 
the  reporters,  either  in  their  -technic  or  in  their  identification  and 
classification  of  the  several  cellular. constituents  of  the  blood,  or  in 
their  interpretation  of  the  frequent  and  Often  rapid  variations  noted 
in  their  numbers  and  forms.  Even  to-day,  with  our  fuller  knowl- 
edge of  hematology,  the  personal  equation  of  the  examiner  influences 
considerably  a  differential  count,  and  there  is  more  or  less  difference 
of  opinion  among  hsematologists  as  to  the  meaning  of  certain  findings. 
What  are  probably  the  first  entirely  reliable  studies  were  made  by 
Ullom  and  Craig.  From  observation  of  thirty-nine  patients  reported 
in  1906,  they  conclude : 
