View report page

46
the excess would be left in the tubes after laking had been completed. Such specimens
of sera require greater dilution to give one dose per volume (for example,
63, 66, or 69 times), or in other words their true titre is one of these values higher in
the range.
If for any such serum within the range a lower titre-value was used as the
working-titre instead of its true and higher value, this would result in using an
excess of complement; but the table shows that the excess at the maximum value
would not exceed .15 unit for each dose. After some experience it has been found
possible to perform this test efficiently with any such lower value, providing the
excess over one dose is limited to .1 dose.
It is important always to choose the working titre-value so that the complement
is used in excess, and never less than the number of doses required. Since dilutions
corresponding to the higher values (not being the true titre, which is lower) would
all contain less than one dose per volume (as shown in the table), a higher value
must never be chosen instead of the lower. Since the anti-complementary property
is variable with different sera, laking might not be complete with such dilutions, and
by using less than 3 M.H.D. in the test proper the results would not be reliable.
3. The particular titrations described.—On the occasion referred to in the
introduction, the guinea-pig serum had been titrated for complement as usual, but
the preliminary titre given by the front row tubes was so much higher than the
confirmatory titre (as obtained by considering the readings of the other three rows)
that the latter readings were considered unreliable. It was consequently decided to use
the unmodified preliminary titre, with the result that the readings for the whole batch
were unsatisfactory. On the following day, using another specimen of guinea-pig
serum and a fresh suspension of sensitised red cells, the titre given by the front row
tubes was read as 1:78 (see diagram), but when the readings of the other three rows
of tubes were considered this titre was modified to 1:60. This modification gave
reliable results when the test was repeated later with the same batch of sera.
The readings of this titration are tabulated below in table II, using table I and
1ts method of construction to give the doses of complement corresponding to each
reading.

Assuming that the unit volume of red blood cells requires the same amount of
complement for laking as is used in the front row tubes, it will be seen that the
excess necessary for the tubes of the other rows (as shown in the end column of
table II) must be due to the anti-complementary property of the other ingredients
added, thus: 1 vol. control serum uses 1.6 dose more complement, and 1 vol.
diluted antigen uses 2.7 doses more complement; but, when these volumes of serum
and antigen are mixed, only an extra 2.3 doses of complement are required, instead
of 4.3 doses which might have been expected. This seems to show the presence of
some substance in the serum which to some extent neutralises, or corrects, the
anti-complementary body or bodies in the antigen. This is a constant observation
when performing these titrations. The unusual feature in this particular case is the
fact that so much complement was inactivated as to require the preliminary titre
to be modified from 1:78 to 1:60. This point will be considered in detail later.