tains the medusa in a delicately balanced fluid for it contains poisons and antidotes as does our blood, which exactly counteract one the other. For example, the jellyfish dies in less than two hours if placed in a solution having the amounts and proportions of the common salt and the potassium of the sea-water, but if we simply add the calcium it pulsates very rapidly for more than twenty-four hours. Finally, however, the calcium produces so strong a muscular tetanus that the pulsating tissue is torn literally to shreds; but all of these injurious effects disappear when we add the magnesium, which causes the pulsation to become much slower and more regular, and wholly prevents the calcium from producting tetanus. Another curious fact is that were it not for the presence of the calcium the magnesium would so stupefy the nervous and muscular tissue that no pulsation could arise. This is the more remarkable because magnesium and calcium are both inhibitors of pulsation, yet when both are present they tend in a measure to offset each other, magnesium mainly inhibits the muscles, while calcium stupefies the nerves.
But to return to our subject, let us carry out some experiments to discover the nature of the stimulus which produces each pulsation of the jellyfish. If we cut a ring from the medusa's disk such as is shown in Fig. 3 and leaves a long narrow strip AB attached to it, and then 
Fig. 3. A Pulsating Ring with an "Index-strip." start a contraction-wave traveling e around the ring; every time the wave passes the point A a side-tracked portion of the wave will pass along the strip from A to B. When each sidetracked wave comes to the end B it dies out, for it can not return over the recently stimulated tissue along which it has just passed. Thus we see that the index strip AB simply serves to catch a portion of each wave which passes its base.
Now suppose we place the ring in a pure solution of magnesium chloride, and allow the index strip AB to remain in natural sea-water. Then the contraction-wave gradually dies out in the pulsating ring, for the magnesium paralyzes the muscles; and at the end of about a quarter of an hour all movement will have ceased in the ring, but long after this we find that the strip AB still continues to transmit contractions at regular intervals of time. We see then that whenever the something which produced the contraction in the ring comes around to the point A it is still capable of setting up a contraction in the strip AB, although it can not now cause the muscles of the ring itself to pulsate.
The explanation is that the stimulus which produces pulsation is
