acquainted with the method, when it is stated that the upper line is a record of the movements of the electro-magnet writer, and the lower line that of the movements of the other writer. It will be observed that the point _a_ in the upper line marks the point at which the induction shock was thrown in; so that by first producing the perpendicular till it meets the lower line at _b_, and then measuring the distance between the point _b_ and the point _c_, at which the curve in the lower line first begins, the latent period (_b c_) is determined--the time occupied by the rotation of the cylinder from _b_ to _c_ being known. [12] It may here be stated that in all the experiments on stimulation subsequently to be detailed, there is no difference to be observed between the behaviour of an entire swimming organ deprived of its ganglia, and that of a portion of any size which may be separated from it. _Summation of Stimuli._--In this way I have been able to ascertain the period of latent stimulation in Aurelia aurita with accuracy. It must be stated at the outset, however, that the period is subject to great variations under certain varying conditions, so that we can only arrive at a just estimation of it by understanding the nature of the modifying causes. To take the simplest cause first, suppose that the paralyzed Aurelia has been left quiet for several minutes in sea-water at forty-five degrees, and that it is then stimulated by means of a single induction shock; the responsive contraction will be comparatively feeble with a very long period of latency, viz. five-eighths of a second. If another shock of the same intensity be thrown in as soon as the tissue has relaxed, a somewhat stronger contraction, with a somewhat shorter latent period, will be given. If the process is again repeated, the response will be still more powerful, with a still shorter period of latency; and so on, perhaps, for eight or ten stages, when the maximum force of contraction of which the tissue is capable will have been attained, while the period of latency will have been reduced to its minimum. This period is three-eighths of a second, or, in some cases, slightly less. Now, we have here a remarkable series of phenomena, and as it is a series which never fails to occur under the conditions named, I append tracings to give a better idea of the very marked and striking character of the results. The first tracing (Fig. 2) is a record of the successive increments of the responses to successive induction shocks of the same intensity, thrown in at three seconds' intervals--the cylinder being stationary during each response, and rotated a short distance with the hand during each interval of repose. [Illustration: Fig. 2.]