Bichloride of mercury readily loses the greater part of its antiseptic power in presence of pus and the substances of which the tissues are made. Besides, it is very irritating, even in dilute solution.

1 Dakin, Presse Medicale, 1915.

Nitrate of silver has a greater value than bichloride of mercury. But it becomes irritating when used in sufficiently strong solution. Many substances which enter into the composition of the tissues inhibit its action in a marked manner. The sensitiveness to light of silver compounds is also an objection to their use.

Iodine, so valuable for sterilisation of the skin, has yielded results much less satisfactory when applied to the disinfection of deep wounds, because it coagulates proteins and irritates the tissues. The penetrative power of iodine is feeble. Treated by this substance, wounds continue to suppurate, and heal more slowly than the rest.

Hypochlorite of soda has a high germicidal power and many other useful qualities. But the hypochlorite of soda found in commerce has an extremely variable composition. Besides, it contains free alkali, and often free chlorine. Consequently, it is irritating when applied to a wound.

The deleterious action of antiseptic solutions upon living tissues should be studied as carefully as their bactericidal action. It is, in fact, absolutely necessary that the substance should be tolerated by the tissues during a prolonged period. The disfavour with which the antiseptic method is regarded by the majority of surgeons is partly due to the use of destructive substances, such as carbolic acid or corrosive sublimate, which have done harm without sterilising the wounds.

In a series of experiments which he made with Mme. Carrel in M. Tuffier's laboratory at the Hopital Beaujon, Dakin studied the action of a great number of substances on connective tissue. The experiments were made on guinea-pigs. Small fragments of sponge of similar weight were placed under the skin of the abdominal wall by means of a short incision which was immediately closed by a suture. On one side, by means of a hypodermic syringe, 1 c.c. of the substance to be studied was injected. In the sponge placed on the other side, which served as a control, 1 c.c. of physiological saline solution was injected. At the end of forty-eight hours examination of the region showed a thickening, more or less considerable, of the tissues surrounding the fragment of sponge which had received the solution. By the change in volume of the sponge, the action of the substance employed upon connective tissue could be estimated. In this manner carbolic acid, iodine, bichloride of mercury produced marked tumefaction. The animals injected with bichloride of mercury usually died rapidly. Those injected with carbolic acid showed extensive necrosis of the abdominal wall.

It was only after having made the comparative examination of a large number of substances, from the point of view of their bactericidal action and of their irritating action upon normal tissues, that Dakin decided upon neutralised hypochlorite of soda and chloramines.