WALLACE 0. FENN 479 



ingested for which proteins (as ovalbumin) other than serum proteins 

 may be used. 



When cells are incubated in dilute serum with carbon they agglu- 

 tinate more rapidly than when carbon is omitted. A typical experi- 

 ment is plotted in Fig. 6. These data were taken from an early 

 experiment when the cells from the peritoneal exudate were not 

 washed in sodium chloride. Under these conditions agglutination 

 of the cells occurred to a considerable extent, since it was impossible 

 to add enough sodium citrate to prevent it without preventing phago- 

 cytosis as well. The degree of agglutination in this figure is, there- 

 fore, not characteristic of the other experiments in this paper. 



Three curves are plotted showing the number of cells still unclumped 

 without carbon, and with the addition of two concentrations of 

 carbon. More cells are always free in the control. The obvious 

 conclusion is that the process of ingesting carbon makes the cells 

 more sticky so that they agglutinate more readily. 



The result of the more rapid agglutination of cells in the presence 

 of carbon is that the percentage of cells containing carbon is higher 

 for the clumped cells than for the unclumped cells. This constitutes, 

 therefore, a serious objection to the use of the percentage of cells 

 containing carbon as a measure of phagocytosis (8). The reaction 

 may appear to have come to a standstill when the aggregation of car- 

 bon-containing cells is merely keeping pace with the ingestion. Kite 

 and Wherry (9) have also observed that clumped cells contain more 

 bacteria than unclumped cells. This might be due, however, to the 

 greater velocity of clumps of cells, the velocity increasing roughly as 

 the two-thirds power of the number of cells in the clump. Neufeld 

 and Rimpau (10), however, reported in 1905, from well controlled 

 experiments, that phagocytizing leucocytes clump more rapidly than 

 control leucocytes where phagocytosis is prevented by omission from 

 the mixture of either the immune serum or the bacteria. This seems, 

 therefore, to be a general phenomenon. 



In experiments with unwashed cells the behavior of the cells is 

 interesting though the results are worthless as accurate measures of 

 phagocytosis. A typical experiment is plotted in Figs. 7 and 8. In 

 Fig. 7 the number of cells still unclumped is plotted as ordinates 

 against time as abscissas. The percentage of these cells which con- 



