no  Methods  of  Water  Analysis.  {A^Sa&rm" 
Equally  marked  variations  were  obtained  by  Gage  and  Phelps 
[Centralbl.  f.  Bakt,  Paras,  u.  Infek.y  Abt.  I,  Bd.  xxxii,  No.  12,  1902  ; 
Irans.  of  the  Am.  Pub.  H.  Ass.  of  the  Twenty. ninth  Annual  Meeting, 
1 901).  They  experimented  with  thirteen  different  media  and  various 
waters.  Fuller  and  Johnson  (Trans.  Am.  Pub.  H.  Ass.,  Vol.  XXV, 
1899)  experimented  with  a  medium  composed  of  meat  infusion  and 
12  per  cent,  gelatin,  omitting  the  peptone  and  salt,  with  the  follow- 
ing comparative  results: 
NUMBER  OF  BACTERIA  PER  CUBIC  CENTIMETRE. 
Reaction  (per  cent  ). 
Regular  Nutrient  Gelatin. 
Meat  Infusion  and  Gelatin. 
00 
no 
200 
o"5 
no 
210 
I'o 
120 
100 
l"5 
80 
130 
2*0 
75 
70 
This  table  also  shows  the  effect  of  the  reaction.  The  optimum 
reaction  in  this  case  was  -J-  0  5  for  the  meat  intusion  gelatin  and  -f- 
i-o  for  the  regular  nutrient  gelatin.  The  optimum  reaction,  how- 
ever, will  differ  with  different  waters.  Generally  -f-  10  to  -j-  15 
(Fuller's  scale)  is  recommended.  In  my  own  experimental  work  I 
found  equally  striking  variations  as  shown  in  the  table,  page  in. 
The  gelatins  Nos.  I  and  9  were  prepared  according  to  the  direc- 
tions given  by  the  Laboratory  Committee  of  the  American  Public 
Health  Association  on  Standard  Methods.  Gelatin  No.  2  was  pre- 
pared in  accordance  with  the  same  method,  with  the  exception  that 
Armour's  extract  of  beef  was  used  instead  of  meat.  The  reaction  in 
each  case  was  -j-  15.  An  extended  series  of  observations  on  the 
Gelatins  Nos.  1  and  2  showed  that  the  latter  invariably  gave  twice 
the  number  of  bacteria. 
(4)  Condition  of  Cultivation. — An  atmosphere  saturated  with 
moisture,  as  shown  by  Whipple  {Technology  Quarterly ',  Vol.  XII,  No. 
4,  December,  1899),  favors  a  greater  development  of  bacteria.  The 
temperature  also  plays  an  important  role,  since  fewer  bacteria  will 
develop  at  iO°  C.  than  at  200  C.  The  difficulty  of  maintaining  a 
constant  low  temperature  is  well  known. 
(5)  Length  of  Cultivation. — The  day  on  which  the  colonies  are 
counted  influences  the  numerical  results,  perhaps,  more  than  any 
