ity. One sound delays the time of reaction to another, according to his findings, 45, while a visual stimulus causes a delay of 786 in the auditory reaction. Apparently, in the researches thus far made, complication of stimuli more commonly causes inhibition than reinforcement. In the light of the results which have been considered in this paper it is of interest to inquire whether this may not be due to the fact that the temporal relation has not been considered. Possibly any two stimuli may be given in such relation that they will now inhibit, now reinforce one another.

An investigation by BOWDITCH and WARREN ('90) is of special interest in this connection, since they studied the influence on knee-jerk of various stimuli, given at different intervals with respect to the tendon blow. As appears from the following summary statement (p. 60-61) of their results there is striking agreement between their findings for this reflex and mine for the frog:

"(1) In the majority of individuals experimented upon a voluntary muscular contraction occurring simultaneously with the blow upon the knee increases the extent of the knee-jerk, but with the prolongation of the interval between the reinforcement signal and the blow this effect is reversed; the knee-jerk becomes much reduced in extent and may even entirely disappear. With a still further prolongation of the interval the kneejerk gradually returns to its normal value. The interval at which the effect changes from positive to negative varies with different individuals from 0.22" to 0.6". The interval at which the knee-jerk returns to its normal value is 1.7"-2.5". In two individuals the effect of muscular contraction on the extent of the knee-jerk was wholly positive.

"(2) The effect of a sudden auditory stimulus on the extent of the knee-jerk was, in the three subjects of experiment, almost wholly positive, though great individual differences were observed. The maximum effect was produced when the interval between the sound and the blow was 0.2"-0. 3".

(3) The effect of a sudden visual stimulus upon the extent of the knee-jerk was with two of the three subjects of ex

periment almost wholly positive, the maximum being reached when the interval between the flash and the blow was O. I"0.3". With the third individual a positive phase having its maximum when the interval was zero gave place rapidly to a negative phase reaching its maximum at 0.4"-0.8".

(4) The effect of a sudden stimulus of the conjunctiva by an air blast was in general similar to that of a visual stimulus, except that the positive phase in all three individuals reached its maximum when the interval between the blast and the blow was 0.1", and the negative phase in the individual who manifested this phenomenon had its maximum at 0.8"1.0"."

The effects of sudden stimulation of the nasal mucous membrane with a blast of air and of stimulation of the skin of the neck were very similar to those stated.

SUMMARY AND CONCLUSIONS.

1. Increase in light intensity from 1" to 2" before electric stimulation of the skin of the frog causes delay of reaction to the latter stimulus. If the electric stimulus be intense the inhibitory influence, as indicated by the time of reaction, is slight, if it be weak the inhibition is marked, and reaction may fail entirely.

2. Auditory stimuli give contradictory results. Sometimes they appear to inhibit, sometimes reinforce, the electric reaction. In case of the sound of an electric bell introduced (a) 0.1" before the electric stimulus. and (b) 1.0" before, the average for 300 reaction-times indicate a slight inhibitory influence.

3. Visual stimuli either inhibit or reinforce electric reactions according to the temporal relation of the two stimuli. (a) The visual stimulation of a moving object when given 0.1′′ before electric stimulation of the skin causes reinforcement of the reaction, i. e., shortens the reaction-time. (b) When given 0.5" or 1.0" before the electric stimulus the same visual stimulus causes inhibition, i. e., lengthens the reaction-time.

4. The experiments described prove that it is of import

ance to consider the temporal relation of stimuli in any study of the relations of complexes of stimuli to sensory or motor processes. To say that two stimuli were given "nearly simultaneously" or within a short interval of one another" does not suffice, for this unmeasured interval may make all the difference between the conditions necessary for reinforcement and those for inhibition.

5. If the meaning of the above statements in terms of the neural processes is demanded, only a speculative reply can as yet be given. Of theories of inhibition there are already enough; what we need is methods by which the neural process may be studied. Until we have more definite knowledge of what occurs in the organism in case of the mutual interference or reinforcement of stimuli it may be well for us to experiment much and speculate little.

A clear statement of what the neural changes which condition inhibition and reinforcement may be is to be found in a recent article on inhibition by MCDOUGALL ('03). So far as this paper is concerned it matters little whether inhibition be "assimilation," "drainage," "competition," or something yet unnamed.

References.

Bowditch, H. P. and Warren, J. W.

'90. The Knee-Jerk and Its Physiological Modifications. Jour. of Physiol., Vol. 9, pp. 25-64.

McDougall, W.

'03. The Nature of Inhibitory Processes Within the Nervous System. Brain, Part 102, pp. 153-191.

Merzbacher, L.

'00. Ueber die Beziehungen der Sinnesorgane zur den Reflexbewegungen des Frosches. Arch. f. die ges. Physiol., Bd. 81, pp. 222-262.

Wundt, Wilhelm.

'03. Grundzuge der physiologischen Psychologie.

Leipzig.

Yerkes, Robert Mearns.

Funfte Auflage,

'03. The Instincts, Habits and Reactions of the Frog. Harvard Psychological Studies, Vol. 1, pp. 579-638, (Psychological Review Monograph Series, Vol. 4.)

ON THE BEHAVIOR AND REACTIONS OF LIMULUS IN EARLY STAGES OF ITS DEVELOPMENT.1

By RAYMOND PEARL.

Introduction.

In morphological research two modes of procedure are usually followed in investigating the significance of some particular structure of an organism. First the form, position, relations and other characteristics of the structure in the adult

organism are studied. Then the embryological history is worked out for the purpose of ascertaining how the structure develops to the complex condition of the adult. In this way,

of course, has been gained the complete explanation of many organs and structures which were inexplicable when the adult condition alone was considered. Indeed, embryological study has come to be considered an absolutely necessary part of almost any morphological investigation which aims at completeness. The ontogenetic history of an organ is regarded as of prime importance in elucidating the adult condition.

It is evident that the same thing may be true when the problem under consideration is one in animal behavior, instead of in animal morphology. As we go up the scale from the lower to higher forms, the behavior becomes more and more complex, and less easily resolvable into simple component factors. To be sure, the increase in physiological complexity does not run exactly parallel to the increase in morphological complexity, yet one does not have to go far before the analysis of the behavior of the adult organism becomes extremely diffi

1 Contributions from the Zoological Laboratory of the University of Michigan, Ann Arbor, Mich., No. 71.

cult of accomplishment. For some time before the present piece of work was begun it was the opinion of the writer that valuable aid in the analysis of the behavior of higher organisms might be gained by following the plan of the morphologist and studying the developinent in the individual of the characteristic features of the behavior. Just as the morphologist studies the ontogeny of an organ as an aid to the understanding of the adult condition, so might the comparative psychologist study the ontogeny of a reaction. It seemed reasonable to suppose, in view of the close relationship which JENNINGS and others have shown to exist between structure and type of behavior in lower forms, that in higher forms the behavior would be simpler in character during embryonic or larval life when the structure is simpler. Of course we know in a general way that this is true; but does it hold in detail for single complex reactions and reflexes? So far as is known to the writer, very little systematic work on behavior has been done from this point of view, except on some of the mammals and birds (cf. notably the work of MILLS, LLOYD MORGAN and SMALL). In these forms the behavior has evidently a considerable psychical element in it. It was with the idea of determining whether anything of importance might be gained by studying the ontogeny of reactions primarily reflex in nature that the present piece of work was undertaken.

The form chosen for study was the king-crab Limulus polyphemus. The reasons for this choice were two fold; in the first place, I was already familiar with the behavior and reactions of the adult organism, and in the second place, Limulus is a form in which the behavior is quite complex, and yet at the same time the different reflexes are strikingly definite and machine-like in character. The adult Limulus is an almost ideal form for physiological work, on account of its tenacity to life after most extensive operations have been performed upon it, and because of the definiteness of its responses. Something of the complexity as well as the definiteness of its behavior can be gathered from the excellent account which PATTEN ('93) has given of the gustatory reflexes, for example.