In

I. Error in Judgment of Direction of Motion on the Skin. making these observations the part of the body on which the experiments were to be made-namely, the back, more commonly the leg, or much the most frequently the fore-arm—was placed on the support in a fixed position under the car, the eyes of the percipient being closed. The operator set the apparatus in motion and also a metronome, and, after the avertissement "ready," dropped the weight gently and noiselessly into the cup, which thereby pressed the button upon the arm. As soon as the percipient had determined whether the motion was up or down the limb, or (more generically) to or from the head, he said plus for up and minus for down. The time and judgment were recorded by the operator, and subsequent trials in the same way and over the same dermal tract, sometimes to the number of twenty or thirty in a sitting, were made till signs of fatigue began to appear. The following Table gives the gross results of many observations.

The headings of the fifth and sixth columns mean respectively number of times when motion up the limb was judged to be motion down it, and vice versâ. In this Table no account of rates of motion, of weights or of the surfaces of skin tested is taken. For each of these conditions, so far as they were explored in our research, as well as for all in the aggregate, the following law appears. We are more likely when in doubt to judge motion on the surface of the limbs to be up rather than down their axis. On the breast, shoulderblades and back between them, the tendency was to judge movement to be towards the head although these parts were less fully tested. Man's experiences with sweat and rain, especially without or before clothing, must have made him more familiar with downward than with upward movement on the surface of his body, and the latter, as being more apt to be caused by living things-insects, parasites, &c.-or by aggressive outward movements with the limbs, would be more likely to attract his attention. Movement also against the direction of the hairs, "which strokes the wrong way," would for anatomical reasons seem at first view

to be a stronger stimulus than motion coinciding with their direction. Mainly for this reason probably, minus or "from" movement often failed to be felt with the lighter weights which in the opposite direction caused a distinct sensation. Whether the general law above stated holds for all parts of the surface of the limbs cannot be inferred on the basis of our observations, which were made mainly on the upper and inner fore-arm and on the middle of the upper thigh, but it seems not unlikely that it may for most of it.

A few general remarks may be appended to this section. The percipient is quite prone, unconsciously and with the best intentions, to judge direction from accessories rather than from the simple elements of motor impressions. If there have been several consecutive judgments in one direction, he expects the other direction and often judges on general grounds without laboriously fixating the sensation. Even an inadvertent noise of the hand in adjusting or a squeaking of the apparatus is liable to enter as a factor of judgment. Again, when four or five consecutive movements are given in one direction, the time of the first and last judgment or judgments is apt to be longer than of the intermediate ones; also, after such a series is given, the first movement in the opposite direction is often wrongly judged. Thus a more frequent alternation of direction was found to constitute a better condition for correct judgment. The sense of motion is strongest during the first few seconds and slowly and irregularly diminishes with time. The fall of the button on the skin must not be too forcible or the direction of movement can be told by the swing inward toward the axis of the button, which, as it depresses the skin, stretches it slightly before it begins to slip over it. This sense of stretching, which seems from repeated notes to that effect in our protocol, to be a possible factor in making the skin over muscles more susceptible to motor impressions than skin immediately over bones, is the sensation which comes immediately after that of contact. It can be somewhat reduced by rubbing vaseline over the surface of the skin tested, but it does not necessarily interfere with the exclusive fixation of other elements of impression of motion. Then comes an indeterminate sense of motion, of which more is said later, sometimes preceding any judgment or even impression of direction. The first impression of direction is quite likely in all cases to be that the movement is upward even when it is downward, a reversal of this impression and sometimes an alternation of impressions leading at length to a correct judgment. This alternation has repeatedly led us to anatomical conjectures. Two nerve-fibres, a and b, could, e.g., near their tips bend back and hook into each other in such a way that there might be a particular spot on the skin where a straight line would first stimulate the body of a, then the tip of b, then the tip of a, and finally the body of b.

In almost every possible motion in this direction a is

561

stimulated before b, and we have not learned to differentiate sensations finely enough, or dermal experience has not sufficiently educated this one spot, to rectify the general rule.

All the observations in this section are concerned with only two opposite directions, and those parallel with the axis of the body or limb. It may be added however that, with an apparatus to be described later, observations are being conducted involving discriminative judgments of any horizontal direction as a function of time, rate of motion and dermal area, by a drop-cup and button like the above which can be set by the operator to move Enough out from a central starting-point along any radius. results have already been reached to show not only the great complexity and indefiniteness of the sensations on which a judgment of the direction of motion over the skin is based, but the great inaccuracy of such judgments if not supplemented by muscular innervation; from which it seems not unlikely certain inferences to retinal action may be drawn.

The II. Time-relations of Judgments of Motion on the Skin. following Table is based on the same observations as Table I., the only difference being that a number of series available for direction and not for time are excluded, and a few new ones added. It is therefore also a gross Table in which several conditions of rate, weight and place are undiscriminated, with a predominance as before of values for the inside of the fore-arm. These variables, if we except rate, which is but little varied, cause, as will be seen later, no very high average error of time.

The most uniform conditions of the above Table were with J. V. D., upon whom all observations were taken in groups of either eight or ten per sitting, with an equal number each way, on the volar fore-arm, and with very slight variation of other conditions. The least uniform conditions were those with W. N. From this Table it appears (1) that a judgment of motion down the limb

takes more time than that of an upward movement; (2) that the time for a wrong judgment is much greater than-sometimes more than twice as great as the time needed for a correct judgment; (3) that of the two errors it takes longer to judge minus to be plus movement than the reverse. Why it takes so long to mistake downward for upward movement, when that error is far more common than the opposite one and when also we are so likely to get an early though faint plus impression from all motions, it is not easy to say. That an erroneous judgment is given after so prolonged an impression, may be said to show the strength of the tendency. If, as would appear, minus movements are a feebler stimulus than plus ones, something might be due to less sharpness of attention; but then why should these verdicts be so much longer than correct minus judgments? The only explanation we can suggest is, that in these longest of all judgments the sense of time past since the movement began shrinks in consciousness faster than the sensory after-image arising from the moving point fades, so that there comes a moment when we interpret time past as vividness of seemingly shorter impression, and the more vivid the sense of motion, other things being equal, the more like a plus motion it seems. Beyond a certain length of time, varying with many conditions, this tendency, if such it be, would be corrected by a sense of distance and direction between the remembered spot where the point first touched the skin and its present position. Again, after all allowances are made for distracted attention, we believe that the short time of correct as compared with wrong judgments shows what has been often remarked in the course of this experimentation, viz., that a too laboured fixation of attention confuses a more rapid and instinctive divination of the direction of motion, which is apt to be correct, though with the feeblest assurance of correctness, before the attention feels itself fully focussed and ready for its more selfconscious and artificial activity.

In another Table, which it is not needful to reproduce in full, the above results were worked out with each group of observations for one day taken as a unit, and these units averaged. This Table, when calculated for an equal number of similar observations, gives as the percentage by which the time of judgment of a minus motion exceeds that of a plus

Here, where not single observations but different diurnal verdicts and states are aggregated, even W. N. and G. S. H. fall under the general rule that minus judgments require the longer

time. The results were once more tabulated by grouping together all experiments on the back, upper and lower fore-arm, ball of thumb, shin and thigh respectively, to see if the general law that minus judgments took longer than plus had local exceptions or was peculiarly great for any of these parts; but, although with three of our subjects many observations were taken on all these parts, the excess of minus judgments was so uniform that we can infer no such difference, and we believe it to be a general law valid in about the same degree for all these parts.

III. Effect of varying the Rate of Movement or the Distance which must be traversed before the Judgment is made. We now began to fix our variables. The following Table is accordingly made on one person H. H. D., with a constant weight of 15 grammes, with a circular metallic point 2 millimetres in diameter, and on a definite part of the volar surface of the right arm so arranged that the point traverses the same tract in moving up or down and in successive sittings.

The most obvious result from this Table is that while the discriminative sensibility for compass-points on this part of the arm, measured longitudinally, could rarely be brought below 25 mms., even in single observations, motion is recognised and its direction discriminated at an average distance of between 6 and