These chambers are piled into hexagonal columns, which lie lengthwise in the organ (they are set dorso-ventrally in Torpedo). Each chamber contains a jelly-like substance which embeds an electric disc. The disc divides the chamber into a smaller anterior and a larger posterior compartment. Each chamber is supplied with several nerves which ramify into innumerable twigs on the front surface of the disc. The development of the disc must be considered for a moment if its structure is to be understood. It starts life looking as if it would grow into a voluntary muscle-fibre. A nerve joins it, forming an end-organ in the usual way. Then the end-organ increases its spread unduly, while the rest of the fibre fails to grow. The structure becomes toadstool-shaped, with the nerve arborizing on the seat of the stool. The front aspect of the disc, therefore, corresponds to a nerve-ending in a muscle. Its middle layer indicates clearly that the fibre makes an abortive attempt to develop cross-striation. It is laminated, the laminæ strangely contorted; in section they appear, not as plain lines, but as rows of dots, evidently a suggestion of longitudinal striation. The posterior layer of the disc consists of granular protoplasm drawn out as a number of short backwardly directed tongues, and one long process, the stem of the stool. No structure could be more suggestive of the function of the organ; but no one has as yet succeeded in catching the suggestion and pressing it into a definite explanation of the way in which it works. Certain physiologists, laying great stress on the fact that the functional connections between an electric organ and its nerves are not easily interrupted by the administration of curari, atropin, and other drugs, which block the passage of impulses from nerves to muscles, look upon the nerve-layer of the disc as the generator of electricity, and the rest as an accumulator or resonator, which stores, or exaggerates, the electric charge. Others consider that the portion of the disc which is altered muscle-fibre—the middle, or middle and posterior layers—generates the electromotive force, the nerve simply calling it into activity. All agree that a brief interval (about 0·003 second) elapses between the arrival of the nerve-impulse and the discharge of an electric shock. This “latent period” may be used as an argument in favour of either view. It would be in harmony with the general account which we have already given of protoplasm as a liberator of energy to suppose that a nerve-impulse, having reached a disc, immediately infects the protoplasm of the disc, inducing molecular commotion, and that the ions move in such directions as to disturb the electric equilibrium of the disc, its front surface becoming in relation to the back as zinc to copper in a battery. The current generated in the fish is in the direction from head to tail. It is certain that the change does not occur until an impulse reaches the organ. The organ is not