456 



PHAGOCYTOSIS OF SOLID PARTICLES. I 



In Table III is also given the average per cent of error which was 



assumed in the experimental points in order to fit them to the equation 



and the average probable error involved in counting the particles 



for the experimental points. The latter value was calculated by the 



1.05 

 expression ~/^' 



= . where A^ was the number of particles counted.^ 



The assumption that the points He on a straight line is justified by 

 Log A-x 



1 



3 Hours 



Fig. 6. Experiment 2. Legend as in Fig. 5. Experimental ratios, 1: 0.87: 0.58; 

 theoretical ratios, 1:0.41:0.51. if's determined graphically. See Table III. 



^ The error of a set of observations varies inversely as the square root of the 

 number of observations (Tuttle, L., The theory of measurements, Philadelphia, 

 1916, 219). The constant 1.05 ± 0.049 was determined for counting particles 

 from calculations of the dispersion of extensive counts made on fourteen differ- 

 ent suspensions. The dispersion of the fourteen determinations was ±0.19. The 

 constant was also determined from counts made on forty-eight different leucocyte 

 suspensions and gave 1.07 ± 0.05, the dispersion of the individual determinations 

 being ±0.34. 



