86 The Opsonic Index— A Medico-Statistical Enquiry 
II. The case where couplets and triplets are counted as single ingestions. 
Frequency Distkibution 
Number of 
Leucocytes 
Bacteria 
Ingested 
Normal 
Test Serum 
Serum 
0 
32 
17 
1 
39 
87 
44 
3 
70 
41 
1. 
57 
48 
6 
53 
37 
7 
9? 
47 
8 
32 
0 
16 
10 
7 
25 
11 
5 
16 
12 
15 
13 
18 
U 
1 
8 
15 
0 
8 
16 
1 
6 
17 
7 
18 
- 
4 
19 
20 
2 
2 
21 
22 
2 
0 
23 
2 
2k 
0 
25 
0 
26 
27 
0 
2 
Means 
4-10±-09 
6-65 ±-14 
Standard 
Deviations 
2-86 
4-74 
Opsonic 
1-62 
Index 
So far as the actual index goes there would not appear to be any great 
difference produced on counting by either method. Probably a better way of 
determining the point would be to fit frequency curves to each of the distributions 
and then test for goodness of fit. The one in which the fit was worse — if there 
were a marked difference — would represent the worse method of counting. I have 
carried out this test in one case of a tuberculo-opsonic distribution which, though 
not conclusive even for the organism used, is at least suggestive. The frequency 
distributions and the corresponding curves are here given : — 
