WALLACE O. FENN 459 



Experiments 5 and 6. — In two more experiments on quartz particles the sus- 

 pensions used were less uniform because the large particles were removed by cen- 

 trifugalization. The diameters of these particles were obtained by direct micro- 

 scopic measurement. Frequency curves so obtained are plotted in Fig. 9. 

 Each frequency curve was then divided into a number of groups and the chances 

 of collision calculated between each one of the quartz groups and the three groups 

 of cells by equation (5). The chances of collision for any one suspension is, 

 therefore, equal to the sum of the products of the chances of collision of each 

 quartz group by the per cent of particles in that group, or 



/^total - jQQ 



where P is the per cent of cells or particles in the designated group, R the chance 

 of collision; the subscripts 1, 2, and 3 refer to groups of particles while a, b, 

 and c refer to groups of cells. 



Number of 3 2 1 



"Particles 

 30J 



QuarTz Suspenaions 



5 6 7 



Diajneters in Microns 



Fig. 9. Frequency curves where ordinates represent number of particles and 

 abscissae represent diameters of particles in microns. Data from microscopic 

 measurements of three quartz suspensions used in Experiments 5 and 6. Aver- 

 age diameters equal 2.44 /x, 4.08 p, and 4.63 /x. Points plotted are experimental 

 points smoothed by averaging each ordinate with the two adjacent ordinates. 

 Plotted to such a scale that the areas subtended by each graph are equal. See 

 Table V. 



The calculation in this case is much more laborious and the results 

 are not so satisfactory. Moreover, in these experiments, which were 

 the first comparisons attempted, larger test-tubes (8 mm. inside 



