Now, we have here to deal with a class of facts which physiologists will recognize as of a perfectly novel character. Why it should be that the very tenuous tracts of tissue which I have named should have the property of responding even to a feeble stimulus by issuing an impulse of a kind which throws the contractile tissues into a spasm; why it should be that a spasm, when so originated, should manifest a power of penetration to which the normal contractions of the tissues in which it occurs bear so small a proportion; why it is that the contractile tissues should be so deficient in the power of originating a spasm, even in response to the strongest stimulation applied to themselves;--these and other questions at once suggest themselves as questions of interest. At present, however, I am wholly unable to answer them; though we may, I think, fairly assume that it is the ganglionic element in the margin, and probably also in the radial tubes, which responds to direct stimulation by discharging a peculiar impulse, which has the remarkable effect in question. For the sake of rendering the matter quite clear, let us employ a somewhat far-fetched but convenient metaphor. We may compare the general contractile tissues of this Medusa to a mass of gun-cotton, which responds to ignition (direct stimulation) by burning with a quiet flame, but to detonation (marginal stimulation) with an explosion. In the tissue, as in the cotton, every fibre appears to be endowed with the capacity of liberating energy in either of two very different ways; and whenever one part of the mass is made, by the appropriate stimulus, to liberate its energy in one of these two ways, all other parts of the mass do the same, and this no matter how far through the mass the liberating process may have to extend. Now, employing this metaphor, what we find is that, while the contractile fibres resemble the cotton fibres in the respects just mentioned, the ganglion cells resemble detonators, when themselves directly stimulated. In other words, the ganglion-cells of this Medusa are able to originate two very different kinds of impulse, according as they liberate their energy spontaneously or in answer to direct stimulation, and the muscular tissues respond with a totally different kind of contraction in the two cases. Possibly, indeed, direct stimulation of the ganglia is followed by a spasm of the muscular tissue only because a greater amount of ordinary ganglion influence is thus liberated than in the case of a merely spontaneous discharge. If this were the explanation, however, I should not expect so great a contrast as there is between the facility with which a spasm may be caused by stimulation of the margin and of the contractile tissue respectively. The slightest nip of the margin of Staurophora laciniata, for instance, is sufficient to cause a spasm, whereas even crushing the contractile tissues with a large pair of