80 The Opsonic Index — ^1 Medico-Statistical Enquiry 
proportions; whether the serum agglutinates the red blood corpuscles; age of 
serum ; length of time of action ; variability of normal sera, either of one and tlie 
same individual at different times, or of different individuals at the same time ; 
heated or unheated. 
The factors here enumerated by no means exhanst the list which might be 
made. Nor are they in any sense classified; many are not mutually exclusive 
classes. Moreover, as it is the comparability of indices that we are now considering, 
we postulate throughout equality of conditions for individual observers. Thus, 
when we say that age of serum may be a factor affecting the opsonic index, it is 
not intended by this to raise the question as to whether an old serum might be 
compared with a fresh one — that, we know, it cannot be. But if we have test 
serum and normal serum taken at precisely the same time, will the index remain 
the same if estimated at once, and at varying intervals after clotting of the bloods ? 
Or take the case of length of time of action of the serum in the opsonic mixture. 
Is it all the same whether we give the mixture 15, 20 or 30 minutes in the 
opsoniser ? Will the ratio of the averages at the different periods remain the 
same ? Might we not get certain of the leucocytes more rapidly surfeited under 
the action of one serum than the other and so put out of action ? In such a case 
additional time for ingestion would not be of any use to these leucocytes. A great 
number of such questions arise in the mind of every observer, and it is cei'tain 
that they cannot be answered without set experiment. A most valuable paper 
which deals with this side of the subject is that by Dr A. Fleming in The 
Practitioner for May, 1908. But there are many problems still to be solved before 
the answer can be given as to the limits within which indices are comparable. 
We may, however, illustrate a few of the points, which have been raised by actual 
examples. 
Example 1. What is the bearing of concentration of emulsion upon degree 
of phagocytosis? The organism here used was the Bacillus coli and the emulsion 
a very thick one. The emulsion was diluted down in varying degrees, and tested 
at each dilution against the same serum and the same leucocytes with the following 
result (Table, p. 81). 
The result is a curious one. We see that the frequency distribution for 
a concentrated emulsion exhibits a smaller variation than that given by the same 
emulsion when considerably diluted — the standard deviations being as 1 to 2 for 
a tenfold dilution. Nor was there any agglutination of the bacteria obvious 
enough to account for this result. The probability is that we were dealing here 
with a toxic effect — the so-called aggressin effect. Again, when we reach the very 
high dilutions, we notice that there is no significant difference between the distri- 
butions given by the 80- and the 100-fold dilutions. Perhaps at this point we 
have reached a degree of separation of leucocytes and bacteria such that the 
opportunity for phagocytosis is so small that no difference can be brought out 
between different dilutions in the time allowed for action. It would be interesting 
