402 Mr. G. Dean. Enquiry into the Nature of the [Apr. 23, 



are plotted, they are found to satisfy almost exactly the parabola whose 

 equation is y 2 = Kx, where K = 4*1 (mean of 41, 4'03, 4*4, 3*9). The points 

 (0*5, 1*16) and (1*1) obviously do not lie on the parabola. 



The curve shown in the figure has been drawn from the mean of the figures 

 for each dilution. 



I-4t 



1-3 



1-2 



I-I 



I-O 



•3 



•8 



'j. 



-6 



-o-3 



8)-I 



Co-icentra-bion 

 of Serum. 32 15 TT 















/"IContiinucLt 

 / \ lof 



on I 























> 

 / 



f 





























/ 



/ 

 / 



































































/ 

































f 



























































































































/ 

































/ 

































/ 

































X 

































































r 







■ 



























It will be seen that with the half concentration of serum there is a slight 

 rise in the curve. This occurs so consistently that it can hardly be due to 

 experimental error. From the quarter concentration onwards a fall is seen 

 in the curve. 



It is interesting to note that this square-root proportionality is analogous 

 to what was found by Schutz (1885), Volhard (1901) and others working 

 with albumin-pepsin and fat steapsine, etc. If these data are sufficiently 

 accurate to be relied upon, it means that the phagocytosis is proportional to 

 the square root of the dilution of serum for these concentrations, but that 

 another factor is probably at work to account for the upper part of the 

 curve. 



The importance of this question of dilution in relation to the estimation of 

 the amount of sensitising substances in the serum in normal, pathological, 

 and immune conditions is obvious. 



