W. F. Harvey and A. M'^Kendrick 
95 
The irregularity of the distributions is a noticeable feature in the case of the 
Streptococcus foecalis as compared with the Bacillus friedlander. The variability 
is also shown by the standard deviations to be much greater for the former than 
the latter. The irregularity here exemplified is also to be observed in the case of 
the Staphylococci. It may have something to do with the method of counting of 
pairs or threes. A pair of cocci ought very possibly to be counted neither as 1 nor 
as 2 but, as I have suggested, as 1'5, this last figure being the nearest approxima- 
tion to the degree of effort put forth in the phagocytosis. Again the irregularity 
may be indicative of nothing more than smallness of sample. 
Conclusions. 
1. The opsonic index is a valuable means of gauging the effect of therapeutic 
inoculation and is useful for diagnosis. 
2. The tuberculo-opsonic index is capable of giving in routine practice definite 
indications of the positive or negative opsonic activity of serum for values wiiich 
range outside 0'8 and 1"2. 
3. Opsonic sera give a much greater degree of differentiation amongst the 
leucocytes in the case of organisms other than the tubercle bacillus. It is impos- 
sible to give limits of normality and abnormality in these cases, because no inves- 
tigations have been made as regards the variability of normal serum action upon 
these bacteria. 
4. It is doubtful whether the opsonic indices as calculated by different 
observers are comparable with one another. 
I have to thank Sir Almroth Wright and his colleagues for having placed 
records and material freely at my disposal and for many suggestions as to some of 
the disturbing factors which enter into the estimation of the opsonic index. The 
experience of the Vaccino-Therapy Laboratory at St Mary's Hospital is now very 
large, and the technique of opsonic estimations has been reduced to a very fine 
art. Finally I owe a deep debt of gratitude to Prof. Karl Pearson for his constant 
supervision of the work, his suggestions of methods and controls and much help. 
