As most of the experiments required the handling of the specimens under observation once each day they were transferred to clean jars of fresh sea-water at that time.
The nervous stimulus causing pulsation originates in the sense organs, probably through a chemical reaction liberating sodium, at the nerve-centers (Mayer, 1908). At any given time only a single sense organ, which at that moment is discharging at the highest rate, controls the rate of pulsation. Consequently all the nerve-centers except one may be removed without seriously interfering with the activities of the medusa; when the last one has been removed the specimen will remain quiescent until a new sense-organ has been regenerated with sufficient completeness to take up again the normal chemical reactions which liberate sodium.
Any portion of the disk adjacent to where the nerve-centers have been removed may be insulated from the influence of the remaining centers by destroying the continuity of the subumbrella ectoderm, in which alone the muscles and nervous elements are contained, while still retaining its continuity with the remainder of the disk through the mesoglces.. This mesoglcea is of sufficient thickness to afford a support for isolated areas of the active tissues even when almost the entire ectodermal covering has been removed. Also when the medusas are kept in normal sea-water it is little subject to bacterial action and can be maintained in an apparently healthy condition even when consider ably more than half of the ectodermal layer has been removed.
Three comparisons were made in the experiments dealing with regeneration, loss of weight during starvation, total metabolism, and the influence of the sense-organs on the change in rate of pulsation in response to change in temperature. For all these experiments the oral arms and stomachs were removed and the disk alone used, to avoid the contamination of the water by mucus and to facilitate the opera tions. In each experiment disks of the same size were used, although specimens of the same diameter showed considerable variation in weight due to differences in the thickness of the rnesoglcea. For
studying the general metabolism large specimens were used because it was easier to perform the necessary operations. In the first type of experiment, normal disks, or half-disks, were compared with others that had been rendered inactive by the removal of all sense-organs (figs. 1, 2, 3). The sense-organs were cut out from the border of the disk by a sharp cork-borer just large enough to completely remove this structure at one stroke. From between the sense-organs of the "active" specimens a piece of tissue of equal size was removed so that the amount of injury was the same for both specimens.
In all the experiments on regeneration recorded in the tables given in this paper the halves of each disk were insulated by the removal of two diametrically opposite strips of subumbrella ectoderm extending from the periphery of the disk on each side to a cavity in the center of the disk where the amount of regeneration was measured (fig. 3). In later regeneration experiments and all other experiments where this type of operation was used the medusa disk was separated into halves and the regular operation was performed upon each half. For measur ing the amount of regeneration a disk of tissue 22 mm. in diameter was removed from a similar part of each half-disk. This complete separa tion of the half-disks did away with the folding of the disks at the point where the insulating strips were removed, which often interfered with accurate measurement of the regenerated tissue. (These series will be referred to as the "active" and " inactive " series.) In the second type of experiments normal disks or half-disks were compared with others from which all sense-organs had been removed, but in which the subumbrella muscles had been activated by electrical stimulation. In entire half-disks, from which the central portion has been removed, a circuit wave of contraction may be initiated by induc tion shocks and maintained for as many as 11 days unless interrupted by some strong stimulus (Mayer, op. cit.; Harvey, 1911).