17 



killed many of the homologous and other organisms as long as there 

 was a high leucocytosis. 



We know that there is an intense local reaction in the localized 

 streptococcus infections, and it has been thought by some that there may 

 be lysis of cocci by the inflammatory serum in these areas. It is difficult 

 to confirm or refute this theory on account of the difficulty of obtaining 

 inflammatory serum in the same condition as it is found in the tissues. 

 As it is not uncommon to find blebs of considerable size on the affected 

 parts of erysipelas patients, the fluid from these blebs was taken as the 

 nearest approximation to the inflammatory serum. This blister fluid 

 from several cases of erysipelas was tested for streptococcidal proper- 

 ties but gave negative results. 



The importance of a high leucocyte count in the destruction of 

 streptococci by blood is clearly shown by the following experiment: 



Experiment i. Ten cubic centimeters of blood were drawn from 

 the vein at the elbow of an erysipelas patient and carefully defibrinated. 

 One cubic centimeter of the defibrinated blood, which contained 9,800 

 white corpuscles per cubic millimeter, was put into a small test tube, 

 inoculated with one loop f til of virulent streptococcus culture, and two 

 loopfuls of the inoculated blood were plated at intervals. The remaining 

 eight cubic centimeters were centrifugated and the serum drawn off. 

 We know that the uppermost stratum of centrifugated corpuscles con- 

 tains a high percentage of leucocytes, because they are thrown down less 

 easily than the red corpuscles. This stratum was therefore drawn off 

 with a sterile pipette and mixed with a small quantity of serum. The 

 resultant mixture contained 17,200 leucocytes per cubic millimeter. One 

 cubic centimeter of this "suspension of leucocytes" was introduced into 

 a small test tube, inoculated and plates made as before. To complete 

 the experiment one cubic centimeter of the clear serum was put into a 

 small tube, which was likewise inoculated, and plates made at intervals. 

 This experiment was also performed with normal blood and a non-viru- 

 lent streptococcus. The plates were incubated for twenty-four hours, 



TABLE 8. 



Colonies on Agar Plates. 



Strepto- Leucocyte 



cocci. count. Immed. 2 to 3 lirs. 5 hrs. 



B104 Erysipelas Wood 9800 390 108 350 



B104 Suspension erysipelas leuc 17200 260 58 21 



B104 Erysipelas serum * 200 360 1100 



298 Normal blood 4400 1650 600 315 



298 Suspension normal leuc 6600 1600 420 62 



298 Normal serum . 1700 2500 10000 



