_Nitrous oxide._--This gas at first accelerates the motions of Sarsia, but eventually retards them. I omitted, however, to push the experiment to the stage of complete anæsthesia, which would doubtless have supervened had the pressure of the gas been sufficiently great. _Deficient aëration._--It may now be stated that the Medusæ are exceedingly sensitive to such slight carbonization of the water in which they are contained as results from their being confined in a limited body of it for a few hours. The rhythm becomes slowed and the contractions feeble, while the pauses between the swimming bouts become more frequent and prolonged. If the water is not changed, all these symptoms become more marked, and, in addition, the rhythm becomes very irregular. Eventually the swimming motions entirely cease; but almost immediately after the animals are restored to normal sea-water, they recover themselves completely, the rate and regularity of their rhythm being then quite natural. The suddenness with which this return to the normal state of things is effected cannot but strike the observer as very remarkable, and I may mention that it takes place with equal suddenness at whatever stage in the above-described process of asphyxiation the transference to normal sea-water is accomplished. CHAPTER VIII. ARTIFICIAL RHYTHM. If the umbrella of Aurelia aurita has been paralyzed by the removal of its lithocysts, and if it is then subjected to faradaic stimulation of minimal intensity, the response which it gives is not tetanic, but rhythmic. The rate of this artificial rhythm varies in different specimens, but the limits of variation are always within those which are observed by the natural rhythm of different specimens. The artificial rhythm is not in every case strictly regular; but by carefully adjusting the strength of the current, and by shifting the electrodes from one part of the tissue to another until the most appropriate part is ascertained, the artificial rhythm admits in most cases of being rendered tolerably regular, and in many cases as strictly regular as is the natural rhythm of the animal. To show this, I append a tracing of the artificial rhythm (Fig. 25), which may be taken as a fair sample of the most perfect regularity that can be obtained by minimal faradaic stimulation.[24] [24] This and all the subsequent tracings I obtained by the method already described. [Illustration: Fig. 25.] [Illustration: Fig. 26.] This artificial rhythm may be obtained with a portion of irritable tissue of any size, and whether a large or small piece of the tissue employed be included between the electrodes.