We may wish to know, however, what the action of the drug has been on the vagus-roots, and it is evident that if the ends in the heart are paralysed, no action on the vagus-centre could alter the pulsations of the heart any more than nervous stimuli proceeding from the cord could move the legs of an animal poisoned by curare. Nor can we separate the vagus-centre from the heart by ligature of the vessels so readily as one isolates the frog's leg. It can be done no doubt by tying the carotid and vertebral arteries and keeping up an artificial stream of blood through the head. Instead of this, however, the simpler method is generally adopted of injecting the drug to be tested into the carotid artery, so that it will reach the vagus-centre before it gets to the heart, instead of injecting it as usual into the subcutaneous tissue or veins, whence it will be carried to the heart before it can reach the vagus-centre.

By experimenting in this way it is shown that atropine stimulates the vagus-roots so that when injected into the carotid it causes slowing of the heart's action. When it has passed through the cerebral vessels, and returns with the blood to the heart it paralyses the ends of the vagus in the heart, and therefore the pulse again becomes very rapid, notwithstanding the continued stimulation of the vagus-roots.

We cannot always conclude with certainty that a drug has excited the vagus-roots merely because it has caused the pulse to become slower and has had no action after the vagi have been divided, for it is possible that the terminations of the vagus in the heart may be rendered more sensitive than usual by a drug, so that they may respond to a slighter stimulus than usual or with greater energy to a normal stimulus. Such an action appears to be exerted by physostigmine, which in a certain stage of poisoning renders the vagus more excitable, so that when irritated in the neck by a faradaic current a slighter stimulus suffices to stop the heart after the administration of the drug than before.