topples over and rolls down till it lies in the usual position in the bottom of the "vicarious chorion" with the haemal side up. I have seen this series of actions repeated time after time by the same embryo. It usually takes several minutes of strenuous labor for the embryo to get righted in the "vicarious chorion," and then within a few seconds of the time it attains. that position, it suddenly loses its balance and falls back to the point from which it started. Then the same series of events begins again. There is nothing in the behavior which would in the least suggest any process of "learning by experience," or of perfecting a reaction by practice.

The stimulus which induces these almost continuous leg movements in the embryo is probably of the same sort as that which causes the righting reaction in the adult Limulus. If adult Limuli are placed in water so that the haemal side is in contact with the bottom, they immediately give a characteristic reaction which brings them into the normal position. From a long series of observations and experiments on the adult animals it appears that this apparent "equilibrium sense" is primarily due to a strong positive thigmotaxis of the neural surface of the body, together with a negative thigmotaxis of the haemal surface. A similar condition of affairs has been shown to be the cause of the righting reaction in other organisms (cf. PEARL, :03 for Planaria). It seems to me probable that the leg movements of the embryos when they are in an inverted position are thigmotactic responses. The only essential difference between embryo and adult in respect to this thigmotaxis would then be that the definite, purposeful reaction with which the adult meets and solves the difficulty has not yet developed in the embryo. Instead the embryonic thigmotactic reaction is simply a generalized response to a general stimulus. The reaction becomes specialized and better adapted to the accomplishment of its end as development proceeds.

The movements of the thoracic appendages which have been described are the only ones which I have observed before the time of hatching. None of the complex, coördinated re

flexes of the legs, such as the gustatory and swimming movements, appear at this early stage.

Reactions to Stimuli.-It is rather surprising to find that before the time of hatching the embryos are extremely sensitive to mechanical stimuli applied to the external surface of the "vicarious chorion." If the surface of the "vicarious chorion" be touched very gently with a needle the embryo stops all movement at once, draws the legs back as far as possible into cephalothorax and strongly flexes the abdomen, so as to make practically a right angle between it and cephalothorax. These positions are maintained as long as the stimulus is continued. In a short time after the stimulation ceases the abdomen is extended, and the respiratory and leg movements begin again.

Fig. 1.-Diagrams showing appearance of Limulus embryo immediately after hatching. 4. Haemal side. B. Neural side. (After KINGSLEY). This is the same reaction as that given in response to mechanical stimuli after hatching. It is somewhat remarkable that the organism should respond in the same way to the pressure of a needle point when in one case this pressure has first to be transmitted by the surrounding fluid to the embryo, and in the other case the needle point touches the surface of the body directly.

Hatching.-The behavior at the time of leaving the "vicarious chorion" I have not been able to observe. In the material which I had all the hatching occured during the night.

The Behavior after Hatching.

Appearance of the Organism.—In general form the embryo at this stage closely resembles the adult Limulus except for the absence of the elongated telson. The appearance of the embryo in dorsal and ventral aspects is shown in Fig. 1, A and B. This stage is KINGSLEY'S (loc. cit.) Stage K.

Movement of Abdominal Appendages; Respiratory Movements.—The respiratory movements continue after hatching in the same characteristic manner as has been described above for the preceding stage in development. The only difference in them is found in the rate, which becomes somewhat more rapid. According to my observations the rate of the beat after hatching stands in about the ratio of 5 to 4 to the rate while the embryo is within the "vicarious chorion." During some experiments in which the embryos were taken from the water and placed on moist sand, with the haemal side uppermost, it was observed in several cases that the respiratory movements continued in the normal manner, while the embryo was out of water. This was a rather unexpected finding, for the reason that adult Limuli never perform continued, normal rhythmical movements except in the water. The only explanation for the case of these young embryos which has suggested itself to me is that possibly at this stage of development the gills are less sensitive to changes in the surrounding medium than they are to the adult. This, however, does not seem very probable, in view of the fact that the general tactile sensitivity of the embryo at this stage is greatly in excess of that of the adult.

Swimming.-Immediately after the embryo leaves the "vicarious chorion" characteristic swimming movements begin. So far as the abdominal appendages are concerned these movements are precisely the same in embryo and adult. They consist of strong extensions and flexions of the gills with reference to the abdomen. They are rhythmical, and are essentially

1

I Throughout the paper where statements are made concerning the performance of swimming movements by the "gills" it will be understood that the gill covers and the operculum are the organs to which reference is made. The term "gills" is used merely to avoid circumlocution.

like the respiratory movements except that the amplitude and force of the beats are much greater in the swimming than in the respiratory movements. On account of the fact that the gills are extended so as to form nearly a right angle with the body at the beginning of each beat a considerable portion of the effective force of the stroke is directed nearly straight backward.

There is a very marked and fundamental difference between the swimming movements of the embryo and the adult with respect to the thoracic appendages. In the adult Limulus all the walking legs beat strongly back and forth in time with the gills during the swimming. As the gills are raised the legs are extended and thrown far forward in the cephalothorax, and as the gills strike backwards the legs accompany them in this movement. In the embryo of the stage under discussion, however, the legs take no part whatever in the swimming movement. Whenever the gills begin swimming motions the legs are extended as much as possible and thrown forward until they touch the antero-lateral margins of the cephalothorax. Then they are held rigidly in this position as long as swimming continues. The appearance in swimming is as if the legs became temporarily paralyzed and held in a "forced" position while the gills were performing swimming movements.

A very interesting course of development is to be observed in connection with the ability of the embryo to direct its movements when swimming. The usual position of the embryo, if lying by itself in the dish, is the same as its position in the "vicarious chorion," namely, lying on the bottom in an inverted position. As has already been stated, swimming movements begin at once after the embryo is hatched. For some time after these movements begin, however, the embryo is not able to rise from the bottom. The gills beat most energetically, but the only result is to cause the embryo to slide along the smooth bottom of the dish. It has not acquired the faculty of so bending the abdomen with reference to the cephalothorax as to make the force of the swimming raise it from the bottom. That this failure to rise is not due to lack of power in the gill

strokes is certain, from evidence to be presented later. To such an extent is the organism unable to direct its movement in these early stages that, instead of tending to swim up from the bottom, it actually tends at times to swim downward. In many cases I have seen the beating of the gills become so forcible as to raise the abdomen and send the embryo skimming along the bottom on the antero-haemal surface of the cephalothorax. If, under these circumstances, the anterior margin of the cephalothorax happens to meet an obstacle in its path the embryo will in many cases turn completely over on its anterior end as a pivot, and come down with the haemal surface uppermost. The "summersault" in such cases is caused solely by the continued violent swimming movement of the gills while the anterior end is held.

How then does the embryo get up from the bottom at this stage, so as to swim freely through the water? This is done in one of two ways, during the earliest stages after hatching. The first of these methods is purely accidental so far as the organism is concerned. If an embryo which is sliding along the bottom as the result of the violent swimming movement of the gills happens to strike squarely a very small obstruction in its path, the anterior end of the body will in some cases slide up onto the obstruction. This, of course, gives the body as a whole an upward tilt and if the swimming movements continue the embryo will rise clear of the bottom and swim freely through the water. Sand grains and pieces of cast egg membranes usually serve as the means for starting the animals upward in this way. Rising in this manner only occurs infrequently, since it is not often that all the necessary conditions will be fulfilled together. After the embryo gets started in this way it is able to sustain itself in the water for as long as a minute even in very early stages of its free existence. Such cases show that the ordinary inability of the embryo to rise from the bottom is not due to lack of force in the swimming movements.

The second and more usual method by which very young embryos rise from the bottom is by first turning over from the usual position with the haemal side down, and then starting to