It is generally accepted that complement may be split into a mid-piece and an end-piece. The mid-piece is thought to be in the globulin fraction, and the end-piece in the albumin fraction. The restoration of complement activity by putting together the albumin and globulin fractions does not prove, however, that each fraction contained a part of the complement, for the albumin fraction can be reactivated in the absence of the globulin fraction.
Complement-splitting as brought about by hydrochloric acid, carbon dioxid, and dialysis, is really an inactivation of the whole complement by certain acids or alkalis, either added in the free state to the serum, or liberated as a result of the dissociation of certain electrolytes.
That the whole complement, and not a part only, is present in the albumin fraction of the serum can be demonstrated by the removal of the inhibitory action of the acid or alkali. This can be effected by the addition, not only of alkali or acid, but also of any amphoteric substance. When hydrochloric acid, carbon dioxid, or dialysis are employed to produce the phenomenon known as complement-splitting, the complement is merely inactivated, not split.