if any, influence on the other side. Moreover, the fact of intimate local relations between nerve and muscle, as seen in the anatomy of these animals, supports the idea of neuromuscular independence instead of centralized relations. This is well exemplified in the reactions of the tentacles. If a tentacle of Metridium is stimulated by food, it turns and twists irregularly and then points toward the mouth. If the same tentacle is cut off and held filled with water so that its original relations in the animal as a whole can be kept in mind, it will be found to react to food as it formerly did, in that it will finally turn toward that side which was originally next the mouth.
The type of neuromuscular mechanism found in the sea-anemones probably also recurs in the digestive tube of vertebrates. This view is supported not only by the action of the intestine, but also by its structure (Fig. 3). Omitting for the moment the outer serous layer and the inner mucous layer of the intestine, both of which have little or nothing directly to do with its neuromuscular mechanism, there are left the outer or longitudinal muscular layer, followed internally by a nervous layer, Auerbach's plexus, which in turn is followed by the cir-