The advantage of this method is that it affords information regarding the action of the poison upon other parts of the nervous system, viz. the nerve-centres and sensory nerves, as well as upon the motor nerves. It also gives the order in which the poison affects the various nervous structures, and shows whether the quantity of poison conveyed to the nerves by the circulation is sufficient to paralyse them or not. For some substances, directly applied to the ends of the motor nerves, may paralyse them, although they do not have this effect when injected into the blood: the reason being that the quantity applied to the nerves directly may be much greater than that which reaches them through the circulation.

The muscles and ends of the motor nerves being protected in the ligatured leg from the action of the poison while it still remains in connection with the nerve-centres by means of the sciatic nerve, this method serves as an index to show what is going on in the nerve-centres. Thus in a frog poisoned by curare it is found that the ligatured leg moves on irritation of the sensory nerves, while all the poisoned parts remain perfectly still. This shows that the afferent nerves are still capable of conveying impressions to the spinal cord, and the cord itself of reflex action, although the poisoned limbs give no indication of the changes which are occurring in the nerve-centres. By-and-by irritation of a sensory nerve or root ceases to produce any movement even in the ligatured limb. This effect is shown to be due to paralysis of the nerve-centres by observing the effect of irritation of the nerves in the ligatured limb, for the muscles still respond readily to irritation of the nerve by a moderate stimulus. We may conclude with tolerable certainty that the motions have ceased in the limbs because the nerve-centres have become paralysed.