allowed to suand for an hour. This is probably occasioned by 

 a disturbance in the osmotic pressure resulting from repeated 

 washings and centrifugation. 



Confusing and contradictory results invariably accom- 

 panied a careless or hurried preparation of the bacterial sus- 

 pension. The or.-anisms used in this work were grown, v/henever 

 possible, on moist plain agar for fifteen hours. The tube was 

 then washed out with saline to remove lint and debris. Five 

 cubic centimeters of salt solution were next added and the tube 

 gently agitated until the resulting turbidity was equal to or 

 slightly in excess of that desired. This was then transferred 

 to a second clean tube and vigorously shaken to break up pos- 

 sible clumps. Finally the tube was centrifuged at a speed 

 sufficient to remove from suspension all but the single bac- 

 teria. Two and five-tenths cubic centimeters were tnen drawn 

 from the upper portion and reserved for the bacterial suspen- 

 sion. All the samples for any experimental series were regu- 

 larly taken from Lhe same level in order to insure a more near- 

 ly uniform number of organisms, dacteria which do not grow 

 upon plain agar or which could not be uniform.ly suspended from 

 solid medium were grown in meat infusion broth. For growing 

 massive cultures of pneumococci, whole blood was added to the 

 broth. The bacteria v;ere sedimented, washed and suspended in 

 salt solution. A turbidity was selected which presented a 

 delicate opalescence in indirect light. 



Sera were taken in the usual manner except that those 

 (7) 



