3. _Caffein._--The effects of caffein on Sarsia may be best studied by immersing the animals in a saturated sea-water solution of the substance. In such solutions the Medusæ float to the surface, in consequence of their lower specific gravity. I therefore used shallow vessels, in order that the margins of the nectocalyces might rest in the level of the water that was thoroughly saturated. The immediate effect of suddenly immersing Sarsia in such a solution is very greatly to increase the rate of the pulsations, and, at the same time, to diminish their potency. The appearance presented by the swimming motions is therefore that of a fluttering nature; and such motions are not nearly so effectual for progression as are the normal pulsations in unpoisoned water. This stage, however, only lasts for a few seconds, after which the spontaneous motions begin gradually to fade away. Soon they altogether cease, though occasionally one among a number of Sarsia confined in the same saturated solution will continue, even for several minutes after the first immersion, to give one or two very feeble contractions at long intervals. Eventually, however, all spontaneity ceases on the part of all the specimens, and now the latter will continue for a very long time to be sensitive to stimulation. At first _several_ feeble locomotor contractions will be given in response to each stimulus; and as on the one hand these contractions never originate spontaneously, while, on the other hand, _paralyzed_ Sarsia never respond to a single stimulus with more than a single contraction, these multiple responses must, I think, be ascribed to a state of exalted reflex irritability. After a long exposure to the poison, however, only a single response is given to each stimulus; and still later all irritability ceases. On now transferring the Sarsia to unpoisoned water, recovery is effected even though the previous exposure has been of immensely long duration, _e.g._ an hour. An interesting point with regard to caffein-poisoning of Sarsia is, that as soon as spontaneity ceases the tentacles and manubrium lose their tonus and become relaxed to their utmost extent. This is not the case with anæsthesiation by chloroform, even when pushed to the extent of suspending irritability. If, however, Sarsia which have been anæsthesiated to this extent in chloroform be suddenly transferred to a solution of caffein, the tentacles and manubrium may soon be seen to relax, and eventually these organs lose their tonus as completely as if the anæsthesia had from the first been produced by the caffein. Moreover in this experiment the irritability, which had been destroyed by the chloroform, returns in the solution of caffein--provided the latter be not quite saturated--though spontaneity of course remains suspended throughout.