I will now sum up this rather lengthy discussion. The two theories of ganglionic action may be stated antithetically thus: in both theories the accumulation of energy by ganglia is supposed to be a continuous process; but while the resistance theory supposes the rhythm to be exclusively due to an intermittent and periodic discharge of this accumulated energy on the part of the ganglionic tissues, the exhaustion theory supposes that the rhythm is largely due to a periodic process of exhaustion and recovery on the part of the responding tissues. Now, I submit that my experiments have proved the former of these two theories inadequate to explain all the phenomena of rhythm as it occurs in the Medusæ; for these experiments have shown that even after the removal of the only ganglia which serve as centres of natural stimulation, the excitable tissues still continue to manifest a very perfect rhythm under the influence of any mode of artificial stimulation (except heat), which is of a constant character and of an intensity sufficiently low not to produce tetanus. And as I have proved that the rhythm thus artificially produced is almost certainly due to the alternate process of exhaustion and recovery which I have explained, there can scarcely be any doubt that in the natural rhythm this process plays an important part, particularly as we find that temperature and gases exert the same influences on the one rhythm as they do on the other. Again, as an additional reason for recognizing the part which the contractile tissues probably play in the production of rhythm, I have pointed to the fact that in the great majority of cases in which rhythmic action occurs the presence of ganglia cannot be suspected. For it is among the lower forms of life, where ganglia are certainly absent, and where the functions of stimulation and contraction appear to be blended and diffused, that rhythmic action is of the most frequent occurrence; and it in obvious with how much greater difficulty the resistance theory is here beset than is the one I now propose. Granted a diffused power of stimulation with a diffused power of response, and I see no essential difference between the rhythmic motions of the simplest organism and those of a deganglionated Medusa in acidulated water. Lastly, the facts relating to the tonus of the manubrium in Sarsia furnish very striking, and I think almost conclusive, proof of the theory which I have advanced. CHAPTER IX. POISONS.