WALLACE O, FENN 339 



curately selected. Unfortunately this is not possible. Moreover, 

 the activity of the cells cannot be quite the same during the periods 

 selected, since the higher temperature coefficient of the preparatory 

 reaction compared to the lethal reaction would make the concentra- 

 tion of Qctive P^ss through its maximum sooner at higher tempera- 

 tures. An improvement might be made by taking as corresponding 

 stages the times until the lethal reaction is one-half complete; i.e., 

 one-half the time necessary for cessation of phagocytosis instead of 

 the time for the ingestion of one-half the total number of bacteria. 

 The validity of this method depends upon the doubtful assumption 

 that cessation of phagocytosis is caused entirely by the lethal reaction. 

 It seems quite probable, however, particularly at the higher tempera- 

 tures, that the tilling up of the leucocytes with bacteria or the partial 

 exhaustion of the free bacteria is another factor of importance. There 

 is obviously a limit to the accuracy of interpretation which is possible. 

 On the whole, the corresponding stages in the reaction probably would 

 not differ so much from those which we have used that the average 

 rates of ingestion in those periods would be seriously affected. 



In conclusion, emphasis may be laid upon the central fact that the 

 phagocytic curves of Madsen and Watabiki represent a complex of at 

 least three reactions and consequently cannot be treated as a single 

 monomolecular reaction without serious error. Osterhout's concep- 

 tion of catenary and (we may add) collateral reactions is not only 

 applicable to the interpretation of these experiments, but obviously 

 essential. 



For comparison with the results of Madsen and Watabiki on the 

 temperature coefficient of the phagocytosis of bacteria, it seemed of 

 interest to determine the temperature coefficient of the phagocytosis 

 of solid particles of carbon and quartz. For this purpose leucocytes 

 obtained from a peritoneal exudate in the rat and particles of quartz 

 or carbon of uniform sizes (2 to 4^4 in diameter) were mixed in small 

 glass-stoppered vials which were rotated slowly about their horizontal 

 axes in water baths kept at the desired temperature. At frequent 

 intervals small samples were removed and counts made of the number 

 of particles not yet ingested. For the details of the procedure the 

 reader is referred to a previous paper (2). The difficulty with the 

 method is that botli particles and leucocytes are likely to agglutinate 



