M. Greenwood and J. D. C. White 
379 
" In this way a double check on the accuracy of working is obtained, as not only is one 
control serum checked by the others, but one observer's count is at the same time checked by 
another's count of the same slide. 
" The figures in the table represent the opsonic indices which ai-e arrived at in exactly the 
same way as in deahng with a patient's blood, the actual phagocytic count of the slides being 
divided by the mean of the counts of all the controls counted." 
Of 635 estimations made in this way 5 gave a reading of less than 0'9, 64 fell 
between 0 9 and 0'95 ; 488 were between 0"95 and 1"05, 68 between I'Oo and I'l, 
and 10 were greater than Tl. 
We take this method to be superior to that employed by Dr Bulloch, since the 
"normal" mean value for any given day is deduced from more than one count 
made with more than one person's serum. We must, however, remark that the 
results to be described in this memoir seem to indicate a very considerable range 
of variation, whether the serum used be taken from a normal or diseased person, 
and that our work refers to much larger samples than those considered by either 
Dr Bulloch or by Dr Fleming. 
Dr Fleming also considers various sources of error which, being dependent upon 
technical manipulations, hardly fall within our province, but it is right for us 
to summarise them as evidence of the complexity inherent in the opsonic problem 
from whatever standpoint it may be regarded. 
(a) Agglutination of the washed red corpuscles increases the amount of 
phagocytosis, and the error thus introduced may be large*. 
(6) If red corpuscles are taken up with the serum, the amount of phagocytosis 
is reduced. 
(c) Blood capsules left widely open for several hours give untrustworthy 
readings. 
We now turn to the actual counting of the cells, a vital matter, in respect of 
which Dr Fleming's views are so peculiar that we quote them in extenso : 
"Having now discussed some points in connection with the different factors in connection 
with the estimation of the opsonic index, we turn to another most important part of the technique, 
namely the counting of the number of bacteria which have been ingested by the leucocytes. 
Needless to say, this must be done with great care and conscientiousness, otherwise errors will 
creep in. In this connection it should be pointed out that it is a great mistake to have any 
arbitrary number of leucocytes which one counts, neither counting more nor less whatever the 
conditions may be. A number of leucocytes, which will give one a good average of a slide with 
one emulsion, will not furnish a true estimate of a slide with another emulsion which may be 
uneven or clumpy. If all emulsions were perfect, it would be very easy to say that a certain 
number of leucocytes per slide would furnish a result differing from the true i-eading by not 
more than ten per cent., but, unfortunately, in practice emulsions are rarely perfect, and, as life 
is short, one wants to get a true estimate of the slide without counting very large numbers of 
leucocytes such as have been advocated in one or two quarters. It is here that experience will 
come in and enable a person, who counts intelligently, to obtain a good average with a smaller 
See Fleming, op. cit. p. (il7. 
48—2 
