Cellular agents of disease, including the bacteria, are, compared with viruses and toxins, huge and chemically extremely complex. When the whole cell, killed or otherwise made harmless, is injected as a vaccine, the antibodies formed against the mass of antigenic material hardly ever kill or inactivate living virulent cells. They may nevertheless act in various ways by helping the body's phagocytes take up the bacteria and destroy them; but generally speaking, they are less effective immunizing agents than those for viruses and toxins.

Just the same, whooping cough is caused by a bacillus which does not form a significant toxin, and neither bacillus nor its products are distributed through the bloodstream. The fact that immunity can nevertheless be produced by inoculation of killed pertussis bacilli shows that these rules are not fixed, and gives us some hope for gonorrhea and syphilis where, for the most part, the rules also do not apply.

The characteristics of gonorrhea, in fact, make it a pretty unlikely candidate for a good vaccine. The gonococcus has all the typical complexity of bacteria as a class, and more. It is not transported through the blood until late in the disease, whereas it is of course the early infectious part of gonorrhea that we need to prevent. The natural disease leaves no recognizable immunity, nor does the lack of immunity seem to be related to antigenic type variation. Possibly this lack of natural immunity is associated with the characteristic pattern of growth of gonococci inside of phagocytic cells, which, instead of destroying them as we ordinarily expect, provide nests for them and help to carry them around!