1. None of the salts tested produce a marked inflammation in vivo in concentrations under 10 per cent. Potassium salts and the different citrates produced atypical inflammatory reactions in mice, but not in frogs. There was no true inflammation, however, characterized by blood vessel changes, migration of polymorphonuclear leucocytes and erythrocytes, and fluid exudation.
2. Synergistic action occurs when equal parts of strontium and magnesium salts are employed. There is a change in the appearance of the mesentery without a true inflammation, and this change does not occur with either salt alone.
3. Amino-acids and amines as a class do not produce inflammation, but histamine produces a marked inflammatory reaction in frogs and mice.
4. Tyramine does not cause an inflammatory reaction but has other marked effects; agglutination thrombi occur within the smaller blood vessels, both veins and arteries; in frogs there is a rapid clumping of the white blood cells followed by a true coagulation with strands of fibrin and entanglement of erythrocytes. This is very widespread and often kills the animal within an hour after injection. In mice it is the erythrocytes that clump and coagulation occurs very much later, usually at the end of 24 hours; still later there is complete absorption of the coagulated masses and the mesenteric circulation returns to normal. None of the mice died during the stage of clumping, and the clots never extended up the larger vessels as they did in the frogs. These effects are similar to the phenomena observed in the in vitro work, in which clumping of the cells appeared constantly.
5. Cantharidinum, histamine, and turpentine produced the most rapid and marked inflammation of any substances tried. These substances are all strongly positively chemotactic in vitro. The differences occurring when these substances are used in different species is a quantitative rather than a qualitative one, the body temperature being of some importance. Papain acted only in warmblooded animals; this is consistent with its chemotactic action in vitro. The degree of positive chemotaxis varied markedly with the blood employed and in the in vivo work the inflammation varied with the species of animal used.
6. Certain substances produced inflammation only some time after injection; this is true of scarlet R and croton oil in weak dilutions. These are not strongly positively chemotactic.
7. Parazol produces an inflammation associated with necrosis of the tissues. This is similar to the results obtained in vitro, parazol being positively chemotactic in low concentrations and negative in high concentrations.
8. The exact chemical nature of many of the substances which produce marked inflammation is unknown. This is true of cantharidin, and the active constituents of turpentine and croton oil.
9. All substances which produce marked and rapid inflammation on injection are positively chemotactic, but not all strongly positively chemotactic substances produce inflammation; i.e., calcium compounds, sodium phosphate, etc.
10. Only substances which are positively chemotactic and also soluble in oil seem capable of producing inflammation in animals.