We have thus very excellent demonstration of the occurrence of reflex action in the Medusæ. Further experiments show that the reflex action occurs, not between the marginal ganglia and every part of the manubrium, but only between the marginal ganglia and the point of the bell from which the manubrium is suspended--it being only the pull which is exerted upon this point when the manubrium contracts that acts as a stimulus to the marginal ganglia. But the high degree of sensitiveness shown by the marginal ganglia to the smallest amount of traction of this kind is quite as remarkable as their lack of sensitiveness to disturbances going on in the manubrium. Turning now to the physiological evidence of the distribution of nerves in Sarsia, when one of the four tentacles is very gently irritated, it alone contracts. If the irritation be slightly stronger, all the four tentacles, and likewise the manubrium, contract. If one of the four tentacles be irritated still more strongly, the bell responds with one or more locomotor contractions. If in the latter case the stimulus be not too strong, or, better still, if the specimen operated on be in a non-vigorous or in a partly anæsthesiated state, it may be observed that a short interval elapses between the response of the tentacles and that of the bell. Lastly, the manubrium is much more sensitive to a stimulus applied to a tentacle, or to one of the marginal bodies, than it is to a stimulus applied at any other part of the nectocalyx. These facts clearly point to the inference that nervous connections unite the tentacles with one another and also with the manubrium; or, perhaps more precisely, that each marginal body acts as a co-ordinating centre between nerves proceeding from it in four directions, viz. to the attached tentacle, to the margin on either side, and to the manubrium. This, it will be observed, is the distribution which Haeckel describes as occurring in Geryonia, and Schultz as occurring in Sarsia. It is, further, the distribution to which my explorations by stimulus would certainly point. But, in order to test the matter still more thoroughly, I tried the effects of section in destroying the physiological relations which I have just described. These effects, in the case of the tentacles, were sufficiently precise. A minute radial cut (only just long enough to sever the tissues of the extreme margin) introduced between each pair of adjacent marginal bodies completely destroyed the physiological connection between the tentacles. If only three marginal cuts were introduced, the sympathy between those two adjacent tentacles between which no cut was made continued unimpaired, while the sympathy between them and the other tentacles was destroyed.