was no noticeable difference to her in the set of the hand for the 10, 20 and 50 grams of pressure, but that the transition from any of these to 100 grams and from 100 to 200 was marked. Nevertheless L's reactions show a steady decrease in quickness from 10 to 200 grams. The shortening of the reaction time with increase in tension of the key spring is, therefore, probably physical and due to an acceleration of the motion of the reacting finger imparted by the recoil of the spring. The tension at which the acceleration would be noticeable would depend on the manner of reaction: ten grams pressure might accelerate a finger reaction but not one from the wrist or elbow. As a result of long practice, L. had become skilled in the finger reaction. This is possibly the reason why, for this reagent, the effects of each tension were noticeable. A measurement of the rapidity of the free recoil of the spring for 10 grams of tension showed that a separation of the contacts of of a millimeter-more than sufficient to break the current-took place in 0.0005 sec. which of course is considerable faster than the reacting finger can move at the beginning of its course. This is to answer the possible objection that for the weaker tensions, the reacting finger moved up faster than the key-bar. PART 2. EXPERIMENTS WITH TRIGGER REACTION Key. As is the case with much of its apparatus, experimental psychology found the ordinary telegraph key already in use and adopted it for its own purposes. The key is very convenient in manipulation and the motion it calls for is 'natural.' Serious objections to it are the antagonistic motion with the break reaction and variability of extent of the reaction motion. Another easy and natural motion is that of the finger in pulling a trigger, with the advantage of a very slight tendency, if any, towards the opposed reaction, though it may well permit an anticipatory contraction. The experiments presently to be described were carried out with a new key of the trigger type. The elimination of the antagonistic reaction would, in itself, hardly be a sufficient reason for adding another instrument to the already long list of reaction keys. This trigger key, however, measures the force and extent of the reaction movement as well as its time. As is evident from the accompanying figure, the key is simple in construction. A cylindrical, self-registering spring balance is mounted horizontally. The movable end is provided with a ring for the reacting finger and in front of this stands an adjustable post serving as a brace for the hand. Electrical contact is made through the horizontal adjustable rod R. mounted parallel to the cylinder. This rod is connected with the binding-screw S the other binding post S1? being connected with the index of the balance. In reacting, the reagent closes the hand around the post P.-adjusted to the proper distance from the 'trigger', and inserts the forefinger in the ring as far as the first joint. When the stimulus comes the reagent pulls the 'trigger', which breaks the contact between the rod R. and the scale index. The pull carries along the registering ring C. which is left in place at the forward end of the pull thus showing the force and extent of the reaction movement. If it is desired to change the initial tension of the spring, the rod R. can be pushed along the side of the scale and the spring set at any tension. In this way the influence of the initial tension on the reaction movement can be easily ascertained. Any anticipatory pull for o grams of tension is signalized by the failure of the chronoscope hands to move. For the sake of comparison of the keys, pairs of series of reactions were taken with each of 10 reagents, each key being used in one series of a pair. The number of reactions in a series was 20, and one series in each pair was taken in halves with the other series sandwiched in between the halves to compensate possible practice effects. As the object of this experiment was merely to test the relative trustworthiness of the two keys, it will be sufficient to make a general stateIment of the results. Although the reagents were inexperienced in reacting, the reaction times as well as the variations differed in no marked degree from those of experienced reagents with whom the processes of reaction had not passed over into the mechanical stage. The figures showed the shorter reaction time for 7 of the 10 reagents with the telegraph key. If we were to 'guess' at the reason for this difference we should say it might in part, be due to the upward push of the key spring, and in part to the stronger tendency towards muscular reaction involved in the reacting position for the telegraph key. With the trigger key, the hand and arm lie in a position producing less muscular strain than is the case with the Morse instrument, and no tension is required to hold in place the moving part of the instrument. Accordingly the content in consciousness of muscular strain is usually less for beginners with the trigger form of key and the reactions tend more to approach the sensory form. The differences of proportional mean variation between the two keys were less marked than differences of reaction time and in general, so far as the data go, we should be inclined to say that taking merely the question of the time factor into account, the trigger key is as trustworthy an instrument as the old key. FORCE AND EXTENT OF THE REACTION Movement So far no attention has been paid to the other factors of the reaction given by the trigger key, i. e., to the factors of force and extent of pull. The object so far has been merely to compare the time data of the instruments, and with such Versuchstiere as beginners in psychology there is every incentive to keep the conditions of investigation as simple as possible. The experiment that follows is a preliminary survey with one reagent of all the data given by the trigger key-the force and extent of the reaction movement in relation to each other and to the reaction time. The reagent for this purpose was Miss Shumate who had done much reacting both in the regular course of laboratory work and in investigations. Before coming to this work and during it, she had been using the telegraph key for reacting to light stimuli with variable signal. Under these conditions her reactions were of the sensory type. She had also taken part in the comparison experiment along with the unpracticed reagents, but had reacted eight periods instead of one. The results for seven of these periods (omitting the first where the data are imperfect) A number of series was next taken with the trigger key set at different tensions from o grams to 1,000 grams-the reagent noting each time the extent, and consequently the force, of the reaction pull. The instructions to S. were simply to mark the condition of the tension-classing it as "high, "medium," or "low" and to note extent of pull. Reactions which the reagent classed as of "low attention" are not included in results. The interval between signal and stimulus was varied slightly to prevent reactions from becoming mechanical. In response to inquiry, the reagent said that her motions were not influenced through noting their extent and force; in the process of reaction she had in mind chiefly the time factor with no thought of making the pulls uniform in extent. About 40 reactions were taken each period, distributed in four series corresponding to four different tensions of the key. Under these regular conditions, with but slight interruption for introspections, the effects of practice became very noticeable, so that at the end of a month the reagent's average reaction time had decreased 75 to 100 sigma. Table II gives the figures for all the reactions of this experiment. Taking the April result, where practice effects practically disappeared, we find with this key, too, an increase in reaction time with increasing tension of the spring and along with this a tendency towards a decreasing absolute mean variation. With the weaker tensions the reactions were fairly mechanical; the reaction motion followed the stimulus without a conscious will impulse. With the higher tensions this was less the case, but whether with 1,000 grams, for example, each reaction motion was preceded by a deliberate impulse, or whether a general state, or 'set' of preparation for stronger impulses preceded the entire series, we do not yet know. In order to get mental conditions that were as far as possible constant, the introspections were reduced to a minimum. The tendency towards a smaller m. v. with the stronger tension indicates however, that the latter condition was the case. To what is the increase in r. t. with increased tension due:to the greater time required for the greater impulse or to the greater resistance to the movement of reaction? The indications from the table are that the latter is the case. The columns headed at', ap', at2, ap2, give the averages of the three quickest and the three slowest reaction-times respectively, together with the averages of the length (and strength) of pulls corresponding to these reactions. The table shows no agreement for any given resistance between length of reaction time and strength of pull'; of the 32 series tabulated, the quickest reactions give 16 longer and 14 shorter pulls than the slowest, while for 2 series within the limits of error of reading, the pulls are equal. In one of these series (10 reactions) the times ranged from 133 to 177 sigma, while the pulls were all of 1,775 grams; in the other, the time range was from 144 to 183, 'Curiously enough Ach (Willenstaetigkeit und Denken, S. 158) assumes that the opposite is true: i. e., he regards it as "sicher" that the quicker reaction follows the stronger impulse. |