The original of tliis book is in tlie Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924024828620 A Plethysmographic Study of Attention BY^ HERMAN CAMPBELL STEVENS THBSIS PRESBNTED TO THE UNIVERSITY FACULTY OF COKNBI.L UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY 1905 QP 405.S8ri905"""' '■"•""' ""''finlifffifliiilPraiPiiifMy *" attention ... 3 1924 024 828 620 A Plethysmographic Study of Attention BY HERMAN CAMPBELL STEVENS Reprinted from The American Journal of PsYCHOUXSy October, 1905, vol. XVI, pp. 409-483. H A PLETHYSMOGRAPHIC STUDY OF ATTENTION.^ H. C. Stevens, Ph. D. TABLE OP CONTENTS. A. Introduction. page 1. Reasons for the failure of the method of expres- sion as applied to the study of the feelings. 410 2. Apparatus and conditions of experimentation. 411 B. Experiments on the state oe active attention. I Attention to visual stimuli. 1. Masson Disc. Exps. 1-14 414 2. After-image. Exps. 15-26 419 3. Radii of circle. Exps. 27-36. .... 423 4-. Radii of circle with intrinsic interest. Exps. 37-43 427 5. After image with intrinsic interest. Exps. 44-50. 430 6. Questions in philosophy. Exps. 51-56. . . 432 II Attention to auditory stimuli. 1. Difference-tone. Exps. 57-62 435 2. Watch moved out to limen and back again. Exps. 63-69. 437 3. Watch moved in from limen and out again. Exps. 70-74 439 4. Cessation of a tone. Exps. 75-82. . . . 441 5. Beats. Exps. 83-90. 444 III Attention to tactual stimuli. I. Cold cylinder. Exps. 91-98. .... 446 ^ From the Psychological Laboratory of Cornell University. 2 STBVENS : [410 Page 2. von Frey pressure hair. Exps. 99-102. . . 449 3. Cylinders. Exps. 103-105 45° IV Attention to ideas. Problems in multiplication. Exps. 106-112. . 451 V Attitude experiments. Exps. 113-125. . • -454 VI Miscellaneous experiments. Exps. 126-136. . . 460 C. Generai, Survey of Results and Discussion. 1. Changes in rates of pulse and respiration. . . 464 J.. The significance of the change in volume. . . 471 3. Summary of conclusions. . . . _ • • 47^ 4. Classification and discussion of the equivocal cases 47° 5. Remarks on the experiments with intrinsic in- terest and on the attitude experiments. . . 480 6. Concluding remarks. 480 D. Appendix. 1. Table VI. 2. References. 3. Illustrative Curves. A. Introduction. 1. Reasons for the failure of the method of expression as applied to the study of the feelings. If the success of a scientific method is to be tested by the consensus of result which follows its application, the method of expression as applied to the study of the feelings must be pronounced a failure.^ The failure is the more surprising, since the method seemed well adapted to its purpose. The organic concomitants of emotive phenomena are so patent, that they in themselves seemed to assure the effectiveness of the method; and, furthermore, in the case of the simpler affective processes, evolutionary theory seemed to indicate a logical ground. Although the conception of the method is supported by common observation and by theory, a long series of investiga- tions has ended in contradictory or negative results. This failure of the method of expression cannot be attributed to the instruments at its disposal, because they are as perfect as the present state of physiology permits. Nor is it likely that the experimenters have been at fault. The causes must be sought elsewhere. Two possible causes suggest themselves: (i) the complication of afiective processes by other mental processes; (ii) the purely physiological processes set up by the stimulus itself. If the method is to be made serviceable for the study of affective processes, the results of these two factors must be known and must be taken account of. ^See Appendix, for proof of this statement. 41 1] STUDY OP ATTENTION. 3 Already something has been done towards the study of these two factors. Mosso, Gley, F^r6, and Delabarre have all studied the effects of intellectual activity as something different from the effects of the feelings or emotions. So, also, have Binet and Angell. McDougall has made a special study of the phy- siological characteristics of states of attention. lychmann de- votes, in his later work, a chapter each to the feelings and to attention and their relations. Recently, Zoneff and Meumann have approached more closely the relation of attention to affec- tion. But with the exception of the last two references the aim of this work has been to apply the method of expression to another mental phenomenon, and not to see how far the ex- pression of affective processes is being masked by the expres- sion of other mental processes. The second factor has been greatly neglected. Mentz, in 1895, made a long study of the effects of auditory stimuli on pulse and respiration, with the result that auditory excitations cause a slowing of pulse and respiration, independently of any mental process. So also, in his second work, did I,ehmann determine the purely physiologi- cal effects of certain stimuli. Very lately, Kelchner has raised the question how much of the effect of the method of expres- sion is due to the stimulus and how much to the mental pro- cess. The aim of the present study is to add something to the knowledge of these two factors, with a view to forcing either the rehabilitation of the method of expression or its final aban- donment. It should be understood that we did not attempt a psychological analysis of attention. Introspection was inci- dental rather than essential to our method. Our interest was chiefly psychophysical. We wished to know what physiologi- cal concomitants those states have, which without too much refinement of definition may be fairly characterized as states of active attention. The experimental material was obtained from a detailed study of attention under a great variety of con- ditions, by means of the expressive method. 2. Apparatus and conditions of experimentation. The in- struments used in this work were I^ehmann's plethysmograph and the Verdin pneumograph. The volume curve furnished the change in rate of pulse as well as the changes in volume. The measurement of the pulse rates from the plethysmogram seems to be warranted by the practice of Lehmann, and of An- gell and McLennan. Both recorded the radial pulse on the one arm while the volume of the other arm was being re- corded. But as no essential difference came out between the measurements of the two curves, the sphygmograph was not often used. The height of water in the manometer of the plethysmograph which was found to yield the optimal height of pulse was about i8 cm. The temperature of the water was 4 STEVENS : [412 kept as near 37" Centigrade as was possible. The plethysmo- graph was always placed on the right arm. The left hand held the key of an electromagnetic signal. The pneumograph was applied to the lower thorax in the case of male observers, and to the upper thorax in the case of female observers. Both the pneumograph and the plethysmograph were connected with their respective Marey tambours by a thick walled rubber tube I meter long, with a lumen 4 mm. in diameter. The length of the writing lever for the pneumograph tambour was 230 mm. The distance from the fulcrum to the point at which the force was applied was 6 mm. The magnification of the curve was, therefore, 38 times. The corresponding dimensions for the lever of the plethysmograph tambour were 235 mm. and 8 mm. The magnification, therefore, was 29 times. The records were taken on a clock-work kymograph made by C. H. Sto«lting. The drum was 28 cm. long and 15 cm. in diameter. With the middle rate of revolution, the period of the drum was, approximately, 140 seconds. The mean va- riation of the kymograph was determined by measuring the number of seconds that fell within a 30 mm. space. The number of seconds varied about 9 as a norm. When the drum was first wound up and stopped after each revolution, the value fell slightly below 9. After 20 minutes of continuous running the value was slightly above 9. Under the former condition the mean variation was 0.12; under the latter, 0.08. The experiments were all made under the former condition. One electromagnetic signal was made to serve three purposes. It served to mark time in seconds from a Zimmerman clock; it served also as a signal for both experimenter and observer. In the two latter cases, a closure of the circuit prevented the making of seconds, so that a dash appeared in the time line. Since only i or 2 seconds were ordinarily obliterated, they could easily be interpolated when the reaction was measured. This arrangement saves space on the drum, and for the condi- tions of our experiments was without any essential error. Each record, therefore, consisted of three curves (counting from above downward), pneumogram, plethysmogram, and the combined time and signal line. The pneumograph tam- bour was inverted in order to close up the gap between the pneumogram and the plethysmogram. This fact must be taken into account in reading the pneumogram, since an up stroke of the .stylus means an inspiration, while a down stroke means expiration. With these instruments and conditions four physiological characters were determined in each curve; rate and depth of respiration, rate of pulse, and volume of the arm. Height of pulse was not attended to, as we did not consider the apparatus 413) STUDY OF ATTENTION. reliable for that determination. The rates of respiration and pulse were measured by counting the number of respirations and pulses that fell within a lo second interval. This interval was determined by erecting, at every nth second, a vertical on the time line. The space, therefore, between two successive verticals represents lo seconds. The changes in volume and depth of respiration were determined by inspection. A de- crease in the height of the breathing curve is always spoken of as 'inhibited breathing.' Each record has 3 periods: the normal or iudiflFerent period, the reaction period, and the pe- riod of recovery. It was attempted to make each period 40 seconds in length; but this was not feasible in some experi- ments. For short reaction periods, the rates were measured in 5 second intervals. In such cases, the results were multiplied by 2, so as to make the averages comparable with the averages of the normal and recovery periods, which were measured in 10 second intervals. The following observers shared in the experiments: Drs. Bentley and Baird, Miss Jenkins and Miss Andrus, Messrs. Sa- bine and Galloway. Mr. Galloway began to observe with Exp. 57. Miss Andrus did not observe after Exp. 36. The normal curves of Dr. Baird', Miss Andrus, Miss Jenkins and Mr. Sabine were recorded for a period of several months prior to the experiments with reactions. The average rates of pulse and respiration, with the mean variation of each, are given in a table. The numbers which enter into these averages were obtained by measuring the number of pulses and respirations for each 10 seconds of each record. The average rate for each record is found by taking the average of the 10 second values. The final average for each observer is made by taking the aver- age of all the records. The mean variation is the average deviation of the average of each record from the average of all the records. The number of records of both pulse and respira- tion is also given. Observer R P MVoi R MVoi P No. of R No. of P Sabine 3 -06 12.8 0.138 1. 21 42 39 Miss Andrus 2.58 14.6 0.199 1-31 34 33 Miss Jenkins 3-24 "■5 0.237 0.24 26 20 Baird 1.83 10. 0.123 0.43 9 9 This table not only brings out marked individual differences in the normal rates of P and R, but the mean variations STEVENS : [414 show the degree of susceptibility to spontaneous variation ot different observers. Experiments were made once or twice a week at the same time of day for each observer. These hours were 12-1, 3-4, and 4-5 p. m. The observers were comforta- bly seated, holding a signal key in the left hand, A curtain shut off the kymograph and experimenter from the view of the observer. Two or three records were made at each hour. Some abbreviations and terms used in the description of the experiments require explantion. Each experiment is divided, as we have said, into three parts: normal, reaction, and re- covery. An unequivocal change in the reaction value of pulse and respiration means a change which differs absolutely from the rates of both normal and recovery. For example: an un- equivocal increase in the rate of pulse means an increase over both the normal and recovery rates. An equivocal change in the reaction rate is one that does not differ absolutely from both normal and recovery rates. For example, the reaction rate may be faster than normal, but slower than recovery; or slower than normal, and faster than recovery. It is a question of some importance to know how to treat these equivocal re- actions, i? stands for respiration; /* for pulse; F for volume. The length of time of each period is given in seconds above the measurements for each period. The average rates of R and P are given below the measurements for each period. For ex- ample, in Exp. I, the normal period was 85 seconds long; the reaction 35 seconds; the recovery 30 seconds. The figures standing opposite R and P are the number of respirations or pulses which occur in ten seconds. A blank in a series ot figures means that the curve, for some reason, was not legible. 3-5i 3-7> 3-8, are the average rates of R for the three periods of the experiment. 13.7, 13.3, 12.8, are corresponding values for P. The changes in depth of breathing and volume are given in words. It will be seen that the reaction of both P and R in this experiment is equivocal. O stands for observer; E for experimenter. B. Experiments on the State op Attention. /. Attention to Visual Stimuli. I. Stimulus : A Masson Disc. The disc was rotated on a hand color mixer. O attended to the fluctuations of one of the rings. Exp. I. SabinB. i"-85" R 3.00 4.00 3.50 4.75 3.25 3.75 3.00 3.00 P 12-5 IS'O 14-0 14.0 14.5 13.0 13.0 3-5 13-7 I"— 30" 4-25 13-5 3.25 4.00 13.0 12.0 3-8 12.8 ibited [ ; there is one deep 415] STUDY OF ATTENTION. i"-35" 4.00 3.50 3.75 3.75 14.0 13.5 13.0 13.0 3-5 13-3 During reaction : R is somewhat inspiration. V falls slowly to the end of reaction ; then rises during recovery. Exp. 2. Sabine. i" — 49' 1' — 32" R 3.50 3.50 3-50 3.25 3-75 3-5° 4.00 3.75 3.50 3.50 P 12.0 12.0 12.5 12.5 12.5 12.0 13.0 12.5 12.0 12-5 3-S 3-6 12.3 12.4 1* — 30* 3.25 2.75 3.00 3.25 3.25 12.0 12.5 12.0 12.0 12.0 3-1 12.1 During reaction: 2 ^ are inhibited (just after signal); for 20* depth is normal; then 4 hypernormal R followed by 3 inhibited R. V falls slowly throughout reaction; reaches minimum at end of reaction; rises during recovery. Exp. 3. Sabine. i'— 39'' i"— 42" K. 3.50 3.50 3-00 3.50 3-75 3-75 3-75 4.oo 3.75 P 11.5 13.0 12.0 13.5 14.0 14.0 13.0 12.5 35 3.8 12.1 13.4 I '—60* 3.25 3-25 3-50 3-25 3-50 3-25 12.0 12,5 11.5 12.0 11.5 11.5 11.6 3-3 11.8 During reaction: i ^ is deeper than normal; inhibition sets in and reaches maximum at the middle of reaction. It gradually lessens to the end of reaction. V falls slightly after first signal, rises, then falls steadily to near the end of reaction. It rises sharply and steadily during recovery. Exp. 4. Miss Jenkins. i"— 39" I"— 55' R 3.30 3.00 3.45 3.27 3.60 4.00 3.95 3.95 3.80 3.40 P II. 5 II. 5 12.5 12.3 12.5 13.0 12.5 12.0 II. 5 12.0 3.27 3-78 II. 9 12.2 I"— 38" 4-05 3'4o 3.05 3-37 12.0 11.5 II. 5 II. 8 3.46 II. 7 8 STBVENS : [416^ During reaction: R is inhibited; inhibition seems to pass off toward the end of the reaction. f'does not fall immediately; falls suddenly and remains low for 15'; rises to end of reaction. Exp. s. Miss Jenkins. I"— 62" R 3.40 P 12.0 3.20 3.50 3.75 12.5 12.5 3-51 12.2 3-50 3-75 12.0 II. 2 I'— 36" 3.7s 4.00 4-00 12.0 12.5 13.0 3-87 12.5 3-75 4-25 12.5 13.5 I'- 35" 3.70 4.00 3.60 11.5 13.0 12.0 3-88 12.5 During reaction: R is slightly inhibited. V is not very constant. After the beginning, it falls slightly; rises to a first maximum; falls slowly to the middle of the reaction; then rises to a second maximum at the end of the reaction. Exp. 6. Miss Jenkins. i''-35 ' r"-35' R 3-75 3-75 3-75 3.80 4-05 3.95 4-00 4-0O P 13.0 12.0 12.5 12.0 .12.5 12.0 12.0 12.0 376 4.00 12.3 I'-SS" 12.1 3-75 4-IO 4-00 3-25 3-9° 3-50 13.0 13.0 12.5 II. 5 12.5 3-75 12.2 10.7 During reaction: R is slightly inhibited. V rises 20" after the beginning of the reaction and continues high until 10" after reaction. « Exp. 7. Miss Andrus. I'- 53' R 3.10 3.30 2.25 2.75 3.45 2.50 P 12.0 12.0 12.0 13.0 13.3 2.90 12.0 i"- 35° I"- 38' 3-5° 3-35 3-9° 3-oo 3-75 3-oo 3-oo 3.00 13.2 13.0 13.5 14.0 13.0 12.75 12.0 11.25 3-43 3-i8 13-4 12.2 During reaction: R is inhibited at first. There are then 2 hyper- normal R; 2 normal R; and 2 inhibited. V does not change. The plethysmograph functioned as a sphyg- mograph. 417] STUDY OF ATTENTION. Bxp. 8. Miss Andrus. I"— 40" i»— sfi-s" R P 2.80 2.50 2.70 3.10 2.9s 2.95 3.00 3.08 10.5 10. II. II. II. g II. 5 II. 12.3 2.83 2.99 10.6 II. 5 I'- 59-5' R 1.75 1.80 2.25 2.00 2.7s 2.87 P II. II. II. II. II. II. I 2.23 II. During reaction: the height of R is increased . V is diminished . Exp. 9. Baird. iv_40" I "—39" R 2.25 1.90 1.85 1.75 2.25 2.25 2-35 2.50 1.92 2.43 i"— so* 2.75 2.25 2.50 2.05 2.20 2-35 During reaction: ^ is greatly inhibited. K is diminished, (/"is not legible.) Exp. 10. Sabinb. I'— 48' i'-32.5" R 3.00 2.75 2.90 3.45 3.43 3.60 3.65 3.55 3.33 P 13-5 13-5 13-0 13-7 14-0 I4-0 I4-0 15-0 3-14 3-53 13.4 14.2 i"-5i-5" 3-50 3-25 3-20 3.00 3.00 3.00 13-7 13-5 I4-0 135 12.5 13.0 3-31 13-3 During reaction: R is inhibited after the first and the second sig- nals. V falls to a low level in the middle of the reaction and rises to the end of reaction. Exp. II. Sabine. i"-34" i"-32.S" R 3-25 3-15 3-OI 3-65 3-35 3-55 3-4° 3-6o P 14.0 13.2 13.5 14.4 13.2 12.7 13.5 16.0 3-27 3-47 13-7 13-8 i"-77-5' 3-75 3-20 3.50 3.15 3.30 3.00 3.07 13-5 13-7 I4-0 I4-0 13-0 13-0 130 328 13-4 During reaction: R is inhibited more during the first half of reac- tion than the last. V is diminished during the reaction. lO STBVBNS : [418 Exp. 12. Baird. 1"— 42" R 2.6 2.6 2.4 2.6 2.5 P 10.5 10.5 10.2 10.5 10. o 2-54 10.3 I"— 48" 1"— 40" 2.6 2.2 2.6 2.4 2.5 2.6 2.4 2.4 2.5 10. i! 10.7 10.5 10.5 10. o 10. o lo.o 10. o 10. o 2.47 2.47 10.4 10. o During reaction: R is inhibited. K falls during reaction. It begins to rise 10" after reaction. Exp. 13. Baird. I "—40" I"— 40" R 2.4 2.6 2.4 2.5 2.6 2.5 2.5 2.5 P 10.2 1Q.2 10.2 10. o 10.5 10. o lo.o 10. o 2.47 2.54 10.8 10. 1 I"— 60" 2.5 2.4 2.5 2.2 2.4 10. o 10. o 10. o 10.2 10.2 2.42 10. 1 During reaction: R is inhibited. V is diminished. The respiratory oscillation which was very marked during normal and recovery was suppressed duting reaction. Exp. 14. Baird. I "—40" i"— 46" R 2.5 2.2 2.6 2.6 2.6 2.7 2.5 2.4 2.3 P 10.2 9.7 IQ.O lO.O 10. 9.2 9.5 10.2 10.4 2.48 2.51 10. o 9.8 i"-47" 2.7 2.4 2.7 2.4 2.5 10.5 10. o 10. o 10. o 10. o 2.56 10. 1 During reaction: R is inhibited, more in the middle than at the beginning. K falls somewhat after first signal; but rises before the end of the reaction. The respiratory oscillation is suppressed. Summary of Results. Attention to Masson disc is characterized by i. Unequivocallyincreasedrateof i?(iocasesoutof 14). In 3 cases R is greater than normal and less than recovery; in i case less than normal and the same as recovery. 419] STUDY OF ATTENTION. II ii. Unequivocally increased rate of P (8 cases out of 14). In I case less than recovery and greater than normal. In 2 cases less than normal and greater than recovery. In 2 cases un- equivocally slowed. In i case no experiment. iii. Inhibited breathing. (In part in every experiment. ) iv. A fall in volume of the forearm. (In all cases except Exp. 6, in which V rises. ) 2. Stimulus: Black Paper and After-image. Attention was directed to an oblong piece of velvet-black paper, 4 x 3.5 cm., exposed on a gray background. O fixated the paper for 15-20 seconds; then watched the development of the after-image on the screen. Both fixation and after- image phases are included in the attention period. The measurements of the fixation and after-image phases are sepa- rated within the total reaction, by a line of vertical dots. The average rates of P and R for each phase are to be found at the two ends of the reaction period. The average rates oiP and R for the reaction considered as a whole are given between the averages for the fixation and after-image phases. For exam- ple, in Exp. 15, the average rate of .^ and P for the fixation phase of the reaction is 4.00 and 11.8 respectively; for the after-image phase, the rate of .ff and P is 3.63 and 11.8 re- spectively; for the reaction as a whole, the values are 3.87 and 1 1.8. In any general discussion of the results of these experi- ments only the last values are considered. Exp. 15. Miss Jenkins. iff— 69' i'— 20" i'— 11° R 3.30 3.60 3.40 3.00 3.90 3.80 3.60 4.25 3.75 : 3.63 P 12.0 12.6 12.0 12.0 12.0 II. 7 12.2 12.0 II. 7 : II. 8 3-52 4-0O 3.87 3.63 i2'o II. 8 II. 8 II. 8 I"— 37" 3.95 3.60 3.60 4.00 12.0 12.0 12.0 II. 4 3-78 II. 8 During reaction: Ji is inhibited during fixation; but not during after image. K falls after the first signal at the beginning of the reaction, and rises again to normal level. It falls at the end of the reaction. Exp. 16. Miss Jenkins. i"— 65" R 3.75 3.50 3.90 4.00 4.00 P 12.0 11.75 I2.0 12.0 12.5 3-83 12.0 12 STEVENS : i"— 20" i"— 20" i"- ■29" 3.90 4.00 : 3.2 4.2 12.0 12.75 : I2-0 12.0 3-75 12.0 3-90 12.0 3.66 II. 6 3-95 3-86 3.70 12.0 12. I 12.0 3-77 II. 8 [420 During reaction: R is inhibited during fixation; there is a short period of deeper breathing between fixation period and after-image; the breathing is inhibited during after-image. V falls at the begin- ning and the end of the reaction with normal level between. During the normal period, there was a decided fall in V about 15 seconds before reaction; V returned to the normal level before the reaction. O reported that she looked at the screen and expected E to move it. Exp. 17. Miss Jbnkins. 1"- -58" R 3 -50 3 -So 3-00 3- 60 3- 80 3- 75 P II .0 12.0 II .5 12.0 I] :.o II .0 3-57 11.4 1"— 22" i"— 21" I"— 34" 3-6o 3.75 3.40 : 3.80 3.60 3.90 3.75 3.75 3.7s II. 5 II. 5 12.0 : 11.5 10.8 II. 5 II. s II. 5 11.5 3-58 3-63 3-70 3-78 II. 6 II. 4 II. I II. 4 During reaction: height of R is not much changed. V falls after each signal (at the beginning and the end of the reaction^, with normal level between. During the normal, there was a noise made by a falling body. This disturbance may account for the increase in rate of ^in the middle of the normal. XP. 18. Baird. I"— 40" I"- 33" l"-2l" R 2.50 2.30 2.20 2.00 2.50 2.30 2.25 : 2.55 2.45 2.40 P 10.0 10.5 10.7s 10.25 10.2 10.5 10.75 : 10.25 10.5 II. 2.25 2.35 2.40 2.46 10.3 10.4 10.5 10.5 1"— 40" 2.00 1.50 2.00 1.80 10.25 10.25 lo-o 10. 1.82 10. 1 Daring reaction: R is inhibited strongly during fixation; some- what less during after-image. V remains nearly level throughout; there are marked respiratory oscillations during normal and recovery; these are suppressed almost entirely during fixation. They are not so much suppressed during after-image. Exp. 19. Baird. I"— 39" R 2.55 2.40 2.50 2.80 P 10.5 II. o 10.75 ii-o 2.56 IO;8 42 1] STUDY OP ATTBNTION. 1 3 I"— 37" l"— 27" 2.25 2.40 2.30 2.60 : 2.50 2.30 2.30 10.6 10.5 10.0 10.5 : 9.0 II. o 10.5 2.38 2.31 2.27 10.4 10.5 10. I 10. I During reaction: R is inhibited strongly during fixation; some- •what less during after-image. There are no changes in V, as the ple- thysmograph functioned as a sphygmograph. Exp. 20. Baird. I"— 30" 2.00 2.00 1.90 1.90 10. 11. 10.5 10.00 1-93 i"— 40" I"— 42" I"— 19" R 2.50 2.40 2.10 2.60 2.25 2.25 2.90 P II. II. II. 5 11.75 "-25 10. 10.75 2.43 2.50 II. 3 10.6 2.50 2.60 : 2.30 2.50 10.5 10.75 : 10.25 lo.o 2.47 2.40 10.5 10. I i"-34" 2.0 1.9 1.6 1.5 10.5 10.25 9.5 lo-o 1-75 I"— 40 // 3.30 3-40 3-45 3- 25 14.5 14.25 15.0 15- 25 3-36 • 3 .20 14.5 14 .1 During reaction: R is inhibited strongly during fixation; less strongly (but still inhibited) during after-image. K rises during fixa- tion; falls before the end of fixation; rises during after-image. Exp. 21. Sabinb. i"— 40" R 3.4 2.75 3.4 3.5 P 14.5 15.75 15.5 14.75 3-25 15. 1 i"_3o" i"— 30" 3.45 2.60 3.10 2.9 3.00 3.25 13.25 14.0 14.0 15.0 15.5 14.75 3.05 3.05 13.7 15-0 During reaction: height of R is irregular; there is no change dur- ing fixation; but there is considerable inhibition during first half of after-image, although less during latter part. V falls during first part of fixation, rising to normal during latter part of fixation. It falls during the after-image and remains low throughout the rest of the re- action. It rises during recovery. O reports the most vivid after- image he ever saw. He was surprised and pleased. Exp. 22. Sabine. i"— 34" I"— 22" 3.50 2.60 : 3.40 3.40 3.10 15.0 16.5 : 14.0 14.2 13.7 3.00 3.10 3.20 15-2 14.5 13-9 -40" 3.00 3.00 14.5 14.5 2.80 14.8 During reaction: R is inhibited as compared with the height of R in recovery. On the whole, there is a fall in V. I"- -40' / R I. ,60 2. 45 2.10 3.00 3.10 P 17 .0 15 2 15 •S .28 ■3 14.5 14.5 3-3° 14.7 14.2 I"—, 2.25 15.5 i 14 STBVENS : L422 Exp. 23. STBVENS : m 1 Sabine. i"-S3" R 2.50 3.00 3.00 3.00 P 15.0 14.25 13.2 14.2 2.80 13-8 3.10 2.5o 14-5 12-5 i"— 30'- I"— 22" I"— 30" 2.90 3.10 : 3.00 3.00 3.75 15-25 14-75 : 13-5 13-75 13-0 3-25 3-20 3-30 14-75 15-25 14-0 2.80 3.02 3.25 14.8 14. I 13.4 3-25 14-6 2.65 During reaction: R is markedly inhibited during after-image; but not during fixation. K falls after the initial and the final signals of the reaction. Exp. •. 24. Miss Andrus. i"-42-5" R 3-00 3-30 3-20 3.20 P 13.8 15.1 15. I 14.2 3-17 14-5 i"-29.5" I"- 50" 2-94 14-2 3-°5 3-33 : 3-45 3-45 325 3-25 15.0 15.5 : 15.0 14-7 15-0 14-6 3.20 No 15 . I recovery 3.10 3.21 3.32 14.9 14.8 14.8 During reaction: R is not changed during fixation, but it is deeper and more regular during after-image. V falls at the beginning and end of fixation; it is normal between. There is a gradual fall in V during the after-image. Exp. 25. Miss Andrus. 1"— 46" R 3.10 2.90 2.7s 3.00 3.20 P 15.0 15.0 14.4 14.8 14.6 2-99 14-7 1"— 17" i"— 29.5" 1"— 33" 3.05 2.85 : 2.95 3.05 3.10 3.33 2.75 3.10 14.4 13.8 : 13.7 14.3 14.4 13.3 13.5 13.8 2-95 3-°3 3-06 14. I I4-I 13.5 During reaction: R is inhibited during fixation and the first part of after-image. V is subnormal, with marked depressions after the beginning and the end of the reaction. Exp. 26. Miss Andrus. R P i"_ 44" 3.40 2.80 2.90 3.3s 14.0 13.6 14.0 15.0 I"— 14" I"— 25" 3-23 : 3-25 3-20 3-0O 13-5 : 13-8 14-6 13.2 3-" I4-I 3-23 3-19 3-15 13-5 13-7 13-9 423] STUDY OF ATTENTION. 15 I"- 45" 3.00 2.90 3.05 3.4s 2.10 13.7 12.8 13.0 13.6 12.0 2.90 13.0 During reaction: on the whole, there is a fall in y, with sudden dips at the beginning and the end of the reaction. Ji is inhibited during fixation, although it is normal for after-image. Summary of Results. Attention to stimulus and after-image is characterized by i. Unequivocally increased rate of .ff (8 out of 12 cases). In 2 cases R is less than normal, but greater than recovery. In I case R is same as normal, but less than recovery. In I case R is greater than normal, and less than recovery. For the unequivocal increases in rate of R, the fixation rate is greater in four cases and the after-image greater in four cases. In the two cases where reaction is less than normal but greater than recovery, R during fixation is faster than during after-image. In the two cases where reaction is the same or gfreater than normal but less than recovery, the rate of R for after-image is greater. In six cases each, is the respiration rate faster, during fixation and after-image. ii. Unequivocal increase in the rate of /" (3 out of 12 cases). In 5 cases P is greater than recovery, but less than normal. In I case P is greater than normal, but less than recovery. In 2 cases P is unequivocally less than either normal or re- covery. In I case P\% greater than normal; but there is no recovery in that experiment. For the unequivocal increase in rate of P, fixation rate is faster in two cases. For the cases where P is greater than re- covery but less than normal, the fixation rate is faster in three cases, the same in one case, slower than the after-image rate in one case. For the one case of reaction P less than recovery but more than normal, fixation rate is greater. For the two cases of unequivocal slowing, fixation is faster. For the case where there is no recovery, fixation is faster. In all, fixation P is faster than after-image /" in 9 cases, slower in two cases, same in one case. iii. Inhibited breathing in nearly every case; more in fixa- tion than in after-image. iv. I^ower level of F of arm. In one case (exp. 20) there is a rise in V during fixation. In most experiments, there is a sudden fall in V at the beginning and end of reaction with normal level between. 3. Stimulus : Closely drawn radii of circle. O was asked i6 STEVENS : [424 to count the number of radii, about 10 cm. in diameter. Exp. 27. Sabine. There were 69 drawn in a circle i"— 40" 1"— 40" R 3-3° 3-45 3-25 2.75 3-25 3-25 3-30 3-20 P 13-5 14.75 13.0 13.0 12.5 12.5 3.18 3.27 14. I 12.7 i"-37" 3.40 3.40 2.60 3.23 13-5 13-75 13-5 14-4 3.15 13.7 During reaction: R is more regular and perhaps slightly inhibited. V falls. The respiratory oscillation is suppressed. Exp. 28. Sabine. I"— 40" I"— 42" R 3.20 3.30 P 13.0 12.5 3-00 3.0 3.55 3-46 3.10 3.27 12.5 12.25 13.75 13.25 13.0 12.7 3.01 12.5 3-34 13.0 1"— 40" 3.00 3.05 3.05 3.30 ' 14.5 13-75 13-0 12.5 3-10 13-3 During reaction: R is markedly inhibited, more at first than after- wards. K falls noticeably. The respiratory oscillation is suppressed. Exp. 29. Sabine. I "—40" I "—42" E. 3.10 3.30 2.95 2.50 P II. 5 12.0 12.5 13.0 3.60 3.30 3.75 3.50 12.75 12.5 12.3 12.7 , 2.96 12.2 3.53 12.6 I"- 38" 3.00 2.30 2.95 13-75 13-5 12.75 3.50 12.8 2-93 13.2 During reaction: R is inhibited. Kis oscillation is suppressed. diminished. The respiratory Exp. 30. Baird. 1"— 40" I"— 28" R 2.60 2.65 2.35 2.60 P 12.0 13.0 11.75 12.5 2.40 2.50 2.50 2.50 12.0 11.75 ".75 12. 1 2.55 12.3 2.47 11.9 425] STUDY OF ATTENTION. 1 7 I"— 50" 2.75 2.4s 2.40 2.50 2.25 12.25 12.5 12.5 12.5 13.0 2.47 12.5 During reaction: Ji is inhibited. The plethysmograph functioned as a sphygmograph. Exp. 31. Baird. i"— 40" I"— 40" R 2.40 2.10 2.45 2.40 2.45 2.25 2.75 2.45 P 12.5 12.5 12.5 12.7s 12.5 12.0 12.0 II. 5 2-33 2.47 12.5 12.0 I"— 50" 2.60 2.45 2.45 2.25 2.65 12.5 12.0 12.0 12.25 12.5 2.48 12.5 During reaction: R is inhibited; F falls at first, but rises gradually to the end of the reaction. Exp. 32. Miss Jenkins. i"— 50" i"— 40" R Z.80 3.05 3.15 3.50 3.45 4.00 3.90 3.90 4.25 P 13-5 12.75 13-0 13-0 12.75 13-25 13-25 13-50 3.19 4.00 13.0 13. I I"— 40" 4.00 3.80 3.75 3.90 13-5 13-75 13-0 13-0 3-86 13-3 During reaction: R is markedly inhibited, f is unchanged, except for a slight fall after the first signal. Exp. 33. Miss Jenkins. i'/_ 40" I"— 40" R 3-7° 3-85 3-8o 3.00 3.80 3.95 3.75 3.75 P 13.0 12.0 13.0 13.75 12.75 13-0 13-5 I3-I 3-58 3-86 12.9 13.0 I"- 55" 3-70 3-75 3-50 3-55 3-60 13-75 13-5 13-5 13-5 13-0 3.60 13-4 During reaction: R is inhibited. V is very slightly diminished. The respiratory oscillation is markedly suppressed. I8 STEVENS : L42D Exp. 34. Miss Jenkins. I "-40" / I"— 40" R 3-75 3 -40 3-00 P 13-5 I3'5 13-5 3-38 13-7 3 -40 13 -o I"- 50" 3.60 3.90 3.50 13.0 12.75 12.0 3 •68 12.56 3-75 12.5 3-85 12.5 3-75 3-55 12.8 12.3 3.60 12.52 3-55 3-30 12.5 12.5 During reaction: R is inhibited very slightly. The oscillation is slightly suppressed. V is not changed. respiratory Exp. 35. Miss Andrtts. I "-39" I'— 40" R 3.70 3.60 3.55 P 13.5 13. I 13.8 3-61 13-4 3 -60 13-2 3-75 3-5° 3-75 14.0 14.2 13.5 3-69 13-7 3-75 13-8 l"— 40" 3-7° 3-15 3-20 3.70 12.0 12.8 12.7 12.5 3-44 12.5 During reaction: height of R is increased. Cf. Exps. 8 and 24 for deeper breathing during attention in the same O. Cf. also Exps. 25 and 26 for inhibited breathing during attention. There is a marked fall in V. Exp. 36. Miss Andrus. I"— 40" I"- -40" R 3 40 3-35 3-0O 3-05 3-30 3-35 3^35 3 40 P 13.6 12.9 12.6 3-20 13-0 13.0 I''- -49" 13.8 14-3 3 14 14.6 .62 .2 14.0 3-3° 2.25 2 • 50 2 •95 2 .90 13-5 13.1 12 2 .0 12 •74 .7 12 •9 12.8 During reaction: Height of R is increased. V is more markedly decreased than in the last experiment. Summary of Results. Attention involved in counting radii of circle is characterized by i. Unequivocally increased rate of R (8 out of 10 cases). In 2 cases (same O), /? is once slower than normal, but same as the recovery; and once it is greater than the normal, but. very slightly less than the recovery. 427] STUDY OF ATTENTION. I9 ii. Unequivocal slowing of /* (3 out of 10 cases, 2 by same O). Unequivocal increase (2 out of 10 cases). In 4 cases reaction P is faster than normal but slower than recovery. In i case, it is faster than recovery but slower than normal. iii. Inhibited breathing (8 out of 10 cases) ; deepened breath- ing in same O in two instances. iv. Suppression of respiratory oscillation where it occurs. V. Fall in ^"(7 out of 10). 2 cases unchanged; i case V not recorded. At this point, it was suggested that the experiments should be taken as a part of the day's work and be worked through mechanically. Two -variations were devised: (i) an experi- ment in the performance of which O should have an intrinsic interest, e. g. , looking for an illusion in the radii of a circle; and (ii) an experiment which should involve the peculiar in- terests of O; e. g-., a question in philosophy for a student of philosophy. The first variation was carried out with the radii of a circle with a view to detecting an illusion in the lines; and also with the after-image stimulus with a view to marking the periodicity of the after-image. 4. Stimulus: Suggested Illusion in the Radii of a Circle. O was asked to discover the illusion. XP •31 Baird. l"- -48 (/ I "— 48' R P 2. II. 25 ,0 2 II .00 2. .0 II, 10 .0 2 II .10 .0 2.20 10.9 2 10 ■30 •5 2 ID f 2-15 II. 2 II .60 .0 2 10 ■50 •3 2.13 10.9 2.40 10.6 l"- -35 II I- II ■50 .0 2.25 10.4 2, 10 ,40 •5 2.40 19-5 2. 10. 6^ During reaction: i? is markedly inhibited. The respiratory oscil- lation is suppressed. There is no K change, since the plethysmograph functioned as a sphygmograph. Exp. 38. Baird. 1"— 30" l''-26" R 2.85 2.50 2.7s 2.50 2.50 2.16 P II. 5 II.O II. 5 II. 4 10.7 II.O 2.70 2.38 "•3 II.O 20 STEVENS : [428 I"— 68" 2-75 2-75 2.30 2.50 2.30 2.10 2.30 II. 7 il-S 11.4 10.9 II. 4 II. 2 II. 2 2.42 II. 3 During reaction: M is markedly inhibited. The respiratory oscilla- tion is suppressed. The plethysmograph functioned as a sphygmo- graph. Exp. 39. Baird. i"_ 37" i"_ 22" R 2.30 2.10 2.50 2.57 2.60 2.50 2.41 P II. o II. 6 II. 5 10.7 12.0 II. 5 II. 6 2.36 2.50 11.2 II. 7 I"— 65" 2.85 2.45 2.50 2.45 2.20 2.30 2.50 12.0 II. o II. 3 II. o II. o 10.8 10.6 2.46 II. I During reaction: R is greatly inhibited. The plethysmograph functioned as a sphygmograph. Exp. 40. Sabin«. i"— 40" I"— 32" R 3.10 2.80 3.40 2.60 3.50 3.20 3.41 P 14.6 15.0 15.23 14.75 13.6 13-7 3.10 3-37 14.9 13.9 I"— so" 3.60 3.5s 2.80 3.60 3.10 14.5 14.5 15.25 15.0 15.25 3-35 14.9 During reaction; Ji is slightly inhibited toward the end. There is no change in V except a very slight fall after the initial and final signals. O reported that he thought of the illusion during the nor- mal. Notice the inhibited breathing and suppressed respiratory oscillations during the last 20" of the normal. The rates of R and P are slowed and quickened respectively. This change, however, is opposed to that of the reaction. Exp. 41. Sabine. I"— 40" I"— 42" R 2.90 3.00 3.05 3.05 3.50 3.20 3.05 3-37 P 14.0 14.5 14.0 14.25 13.75 14.1 3.00 3.28 14.2 14.0 I"- 55" 3.50 2.10 3.05 2.25 2.35 4.00 14.2 15.2 15.0 15.0 15.5 14.0 2.87 14.8 429] STUDY OF ATTENTION. 21 During reaction: R is markedly inhibited (after three initial normal respirations). The respiratory oscillation is suppressed. V remains constant. Bxp. 42. Miss Jenkins. I"— 34" 1"— 44-5" R 3.00 2.45 3.05 2.75 3.50 4.00 3.95 3.70 4.00 P 12.5 12.0 12.0 12.5 12.5 13.5 13.0 12.5 12.5 2.81 3.83 12. i! 12.8 I"- 49-5" 3.70 3.60 3.90 3.80 3.41 12.0 12.5 12.0 12.0 12.3 3-68 12. 1 During reaction: R is somewhat inhibited. V is constant except for a slight fall after the signals at the beginning and end of the re- action. O reports that during recovery she tried to imagine what the illusion was; she was, therefore, attentive. This fact may account for the fast rate of R in recovery. Exp. 43. Miss Jenkins. [A control record was taken between exps. 42 and 43 to discover whether O was reacting normally. The slow normal rate of R in exp. 42 was unusual. The results show that normal R in exp. 42 is too slow — 2.81 against 3.39 in the control record. The rate of pulse is practically the same — 12.2 against 12.1 in the control.] I ,"— 40" 1— 39" R 3-70 3-9° 3 20 3.30 3-65 3-85 3-9° 3-61 P 12.5 12.7 12 !.o II. 5 11.25 12-25 12.5 II .9 3-47 3-75 12.1 I"— s8" 11.97 3-90 3.50 3.60 3.30 3-75 3-50 11.6 12.0 12.0 12.3 3-70 11-94 13-5 11.8 During reaction: R is slightly inhibited at first and afterwards be- comes normal. There are no V changes. Summary of Results. Attention to radii of circle with intrinsic interest is charac- terized by: i. Unequivocally increased rate oi R {6 out of 7 cases). In I case R is unequivocally slowed. ii. Unequivocally decreased rate of P (3 out of 7 cases). Unequivocally increased rate of P (2 out of 7 cases). In 2 instances, the rate of P of reaction is less than normal, and in I case the same as that of recovery : in the other case, more than rate of recovery. iii. Inhibited breathing in every case; but usually not marked. 22 STEVENS : [430 iv. Fall in volume only after signals. 5. The fixation and after-image experiments were repro- duced with intrinsic interest. O was required to mark the periodicity of the after-image. Exp. 44. Sabins. I"— 40" R 2.95 3.03 3.05 3.40 P 12.5 12.5 12.5 12.0 3-II 12.3 I"— 33" I"— 47" I"— 10" 3-35 325 3-33 : 3-25 3-io 2.95 3.00 2.85 3.05 12.0 12.25 12.3 : 12.75 12.0 13.0 13.0 12.5 12.5 3-31 3-17 303 3-05 12.1 12.4 12.6 12.5 During reaction: R is not inhibited during fixation; but is mark- edly inhibited during after-image. There is a fall in V after each signal. Exp. 45. Sabins. i"— 40" i"— 20" i"— 21" R 3-05 3-30 2.95 2.9s 3.45 3.22 : 3.40 2.95 P 12.0 12.0 12.0 11.75 12.0 12.0 : 12.0 11.75 3-o8 3.33 3.25 3.17 11-93 12.0 11-93 11.87 i'-52" 3.45 3.00 3.25 3.30 2.85 12.7s 14.0 13.0 12.5 11.8 3-17 12.8 During reaction: in fixation the height of JR is not much changed; during after-image R is markedly inhibited. V is slightly less dur- ing reaction. Exp. 46. 46. Sabinb i"- -38" 1"— 20" I' '-19" R P 3-75 12-75 3-45 12-0 3-62 12.6 3-55 12.25 3-75 13-4 I"- -48 3 -50 12.0 3-4= 12.4 3 -40 : 13-75 : 1 3-37 12.1 3 11 -25 3 •75 12. 3-29 11.9 ■38 .1 3-05 3 -30 3 -SO 12.0 12.25 12.0 3.20 3.00 12.0 II. 5 3 11 .21 •9 During reaction: in fixation R is normal; but inhibited during after-image— more at the beginning than at the end. V falls sharply after the first and last signals; it is lower during after-image than dur- ing fixation phase. There is a sharp rise in V in the middle of fixa- tion. 43l] STUDY OF ATTENTION. 23 Exp. 47. Miss Jenkins. 1"— 41" 1"— 21" i"— 23" R 2.90 3.25 3.30 3.63 3.70 4.00 : 4.00 4.00 3.7s P 11-75 12.25 12.25 13-1 13-25 12.25 : 12.25 12.6 12.2 3-27 3-85 3-8i 3-91 12.3 12.75 12.54 12.34 i"-4o" 4.50 4.00 3.90 3.60 12.25 12.25 12.3 12.0 4.00 12.2 During reaction: R is normal for fixation; but is inhibited for after- image. There are no V changes in this and the three following ex- periments because the plethysmograph functioned as asphygmograph. Respiratory oscillations are suppressed throughout reaction, but more during after-image than during fixation. Exp. 48. Miss Jenkins. i''— 41" I"— 22" I"— 22" R 3.60 3.25 3.85 4.36 4.00 3.81 : 3.80 3.90 P 12.8 12.25 12.75 12.9 12.0 12.0 : 12.0 12.0 3-76 3-9° 3-87 3-85 12.6 12.0 12.0 12.0 1"— 46" 4-00 3.95 3.95 3.30 4.00 12. 1 12.75 13.0 12.75 12.5 3-84 12.6 During reaction: R is somewhat inhibited during whole of reac- tion; but more during after-image than fixation. The respiratory os- cillation is slightly suppressed. Exp. 49. Baird. I"- 37" R P 2.55 2.50 2.50 12.5 12.5 12.75 2.58 12.56 2. 15 ,78 •■■5 1"- 30" I '—29" I"— 30" 2.25 2.50 2.50 12.55 12.56 12.55 : 2.60 2.50 2.22 : 12.55 12.0 11.4 I. 8s 2.25 2.55 12.0 II. 4 11. 2 2.41 2. 12.57 12. 42 2.44 21 11.85 2.21 "■53 During reaction: the result of this curve is inconclusive, as the rates show a gradual falling-off throughout the experiment. There is no apparent change in depth of breathing. Exp. so. Baird. i"-37" R 2.20 2.60 2.40 2.43 P 12.5 12.7s 12.6 12. s 2.40 12. S 1"- -20" I 1/ ■18" 2. ■55 3^' 00 : 2. 75 2. 50 13 12 .6 : 12 :.o II. 25 2 ■V 2 .70 2 .60 12 .8 12 .2 II .6 24 STEVENS : [432 i"-55" 2.7s 2.00 2.00 2.25 2.40 II. 8 II. 8 11.6 II. 8 II. 6 2.28 II. 7 During reaction: ./? is inhibited throughout; but much more during after-image than fixation. Summary of Results. Attention to fixation and after-image, with intrinsic interest, is characterized by: i. Unequivocally increased rate of R (4 out of 7 cases) . In 1 case R is slower than normal but greater than recover}'. In 2 cases greater than normal but less than recovery. In 5 out of 7 cases, the fixation rate is greater than the after-image rate; in 2 out of 7, slower (only slightly). ii. Unequivocally increased rate of Z' (i out of 7 cases). Unequivocally decreased rate of /• in i out of 7 cases. In i case P is unchanged. In 2 out of 7 cases P is faster than nor- mal, but less than recovery. In 2 out of 7 cases P is slower than normal, but faster than recovery. The pulse changes in rate with respect to normal and recovery are inconclusive. But with reference to the fixation and after-image rates of P, the fixation rate is faster than the after-image rate in 5 out of 7 cases, one case unchanged, one case slower. iii. Inhibited breathing. .There is more inhibition during after-image than during fixation (6 out of 7 cases), i case is unchanged. iv. Very slight fall in V. These results, compared with the results without intrinsic interest, yield the following points of agreement and disagree- ment. They agree in i. Increased rate of i? and the division of increase (6 cases each) between fixation and after-image phases. ii. Inconclusive reaction of P; but rate of P in fixation is greater than the after-image. iii. Inhibited R. iv. Slight fall in arm volume. They difier in that i. Without intrinsic interest, R is inhibited more in fixation; while with intrinsic interest, the after-image phase is the more inhibited. ii. With intrinsic interest, R is increased more in fixation than during after-image. 6. Stimulus: Questions in Philosophy. The following six experiments were made with a view to 433] STUDY OF ATTENTION. 25 comparing the phj-siological reactions of attention with natural interest with the attentional reactions in which other interests are involved. (See above, Exp. 37.) Exp. 51. Sabine. Question: Arrange in order of birth a list of six philosopliers (Leib- nitz, Spinoza, Locke, Hume, Kant, Descartes). O was required to signal when the question was satisfactorily answered. R P I"— 40" 3-75 3-25 3-35 13.0 13.25 13. I 3-25 12.96 2.65 12.5 I"— 41" 3.10 3.60 3.20 12.0 12.25 12.5 3.20 12.37 2.90 12.7 I"— 48" 3.05 3.05 2.75 3.5 3.25 13-0 13-75 13-4 3.12 13.0 During reaction: R is inhibited more during the middle of reaction than at the beginning or end. The plethysmograph functioned as a sphygmograph during this and the next three experiments. O reports that he had expected an easy question; he was surprised by the names, and at first somewhat confused. Attention became more intense toward the end of the reaction. During the recovery he thought of the series again. Exp. 52. Sabine. Question: The relation of mind to body in Descartes' system. l"— 40" I"- 44" R 3.40 3.20 2.70 P 11.75 "-25 12-25 2.60 3.05 3.25 3.05 3.40 3.00 12.5 12.75 12.6 12.0 12. r II. 7 2-97 11.9 3-15 12.23 I"— 48" 3-60 12.6 3.40 3.25 3.25 3.37 12.3 12. I 12.1 3-37 12.27 During reaction: R is inhibited. O reports curiosity as to what the question would be. He had difficulty in thinking himself into Des- cartes' system. He attended to the question in an indistinct sort of way during the recovery. Exp. 53. Miss Jenkins. Question: Same as in last experiment. i"— 46" R 3-25 3-05 3-40 P 13.75 14.75 13-5 3-45 13.6 3-55 4-00 I — 39" 4-50 4-25 3-75 4-38 13-5 12.9 13-2 13-3 12-9 4.22 I3-I 13-1 26 STEVENS : [434 I"— 27" 4.05 4.00 3.50 13.0 13.4 12.8 3-85 13.0 During reaction: R is inhibited. Exp. 54. Miss Jbnkins. Question: Malthus' doctrine of population and food supply. I"— 37" I"— 37" R 4.00 3.90 3.25 2.14 3.90 4.00 3.75 3-57 P 13.5 13.5 14.0 13. 1 13.0 13.1 12.8 3.32 3.80 13.6 13-0 1"— 50" 3'75 2.95 3.00 3.05 3.10 13.2 13.6 12.5 12.5 12.6 3-17 12.8 During reaction: R is inhibited, but more during the first half than in the last. The inhibition continues for 14" in the recovery. Exp. 55. Baird. Question: Arrange names of philosophers, as in Exp. 51. i"_55" i"— 20" R 2.80 2.60 2.20 2.75 2.70 2.80 2.50 2.90 P 11.9 11,4 11.4 II. 7 11. 5 11.8 II. 7 11.7 2.64 2.70 11.6 11.7 I"— 58" 2.95 2.70 2.30 2.55 2.50 2.70 II.6 II.5 II.4 II.O II. I 2.61 11-3 During reaction: R is decidedly inhibited. V sinks greatly. O reports that the question was easy. Exp. 56. Baird. Question: Relation of mind to body in Descartes' philosophy, i"— 36" I".— 38" R 3.00 2.50 2.45 2.60 3.00 2.50 2.75 2.50 P 11.9 11.5 II. 9 12.0 11.9 II. 7 II. 2 II. 2 2.63 2.68 11.8 11.5 i"-55" 2.90 2.45 2.40 2.30 2.50 3.00 II.3 II. I 10.5 10.8 10.6 II.O 2.59 10.9 During reaction: R is inhibited and V falls very decidedly. O re- 435] STUDY OF ATTENTION. • 27 ports that there was a slight unpleasantness during recovery, since he could not get his mind off the subject. Summary of Results. Attention with natural interest is characterized by: i. Unequivocally increased rate oi R (4 out of 6 cases). In I case rate of reaction R is less than normal and greater than recovery. In i case rate of R is greater than normal and less than recovery. Both reactions are from the same O. ii. Unequivocally faster /" (i out of 6 cases). Unequivocally slower P (i out of 6 cases). In I case /'is faster than normal, and less than recovery. In 3 cases P is greater than recovery, and less than normal. iii. Inhibited breathing in every case. iv. Fall in V in each of the two cases in which the plethys- mograph functions properly. //. Attention to Auditory Stimuli. I. Stimulus: DifEerence tone. Attention was directed to a difference tone produced by Quincke's tubes, g^-b^. The tone was very prominent. O signalled when the difference tone was heard. Exp. 57. Gai,i,oway. i"— 53" I"— 17" R 3.20 3.30 3.00 3.30 3.25 3.33 2.95 3.28 P 14.3 14.2 14.0 14.0 14.0 15.0 3.26 3.11 14- 1 14-5 i"— 60" 3-70 3-75 3-55 3-75 3-75 34o 14.5 14.7 13.5 14.0 13.5 14.0 3-65 14.0 During reaction: R is not much inhibited. F first rises slightly, then falls decidedly. O reports a feeling of strain and a tendency to hold the breath. Bxp. 58. Sabine. i"— 25" I"— 7-5 R 3-7° 3-25 3-8o 4-00 P 15.8 16.2 15.6 15.6 3.58 4.00 iS-8 15-6 i"— 70.5'' 3.80 3.70 3.30 4-00 3.90 3.7s 3.30 15.8 16.6 14.8 15.0 15.9 16.1 15.2 3.68 15-6 28 STEVENS : [436 During reaction: Ji is greatly inhibited, but the respiratory oscil- lation is not much suppressed. K falls, reaching a minimum after the end of the reaction; and then returns rather slowly to normal. Exp. 59. Miss Jenkins. I "-36' I'-e" I'— 29' R 3-SS 3-8o 4-75 4-9i P 15-3 15-5 I5-0 4-83 14.8 4.50 4-55 4.20 14.8 15.2 14-3 4-25 15-2 4.83 14.8 4.41 14.7 During reaction: ^ is slightly inhibited. Ftends to fall; it reaches a minimum after the reaction. Bxp. 60. Baird. I"— 40* i"— 10.5* R 2.30 2.40 2.20 2.75 P 15.4 14-7 15-2 15.4 2.85 2.38 14.4 14.6 2.41 I5-I 2.85 14.4 I'— 41.5' i"— 9« I"— 20' 2.55 2.96 2.68 2.00 2.20 2.10 14.7 14-6 14-9 I5-0 13.9 13-8 2.64 14.7 2.00 I5-0 2.15 13.8 During reaction: J? is inhibited, tains a minimum after the reaction. K falls in. both reactions, and at- Exp. 61. Baird. 1'— 30" 1"— 6" i'-37" i«—j.5' I '-8.5' R 2.85 2.35 2.45 2.25 2.00 P 10.9 ii.o 11.4 ii.o ii.o : 2.45 2.50 to.9 10.5 2.64 2.53 2.52 10.5 II. 3 10.5 2.55 2.25 II. I II.O 2.39 10.7 2.53 2.52 11.3 10.5 During reaction: i? very slightly inhibited. There is no apprecia- ble change in V. O reports low degree of attention. Exp. 62. Gauoway. i"-39'' I"— 14" I "—20" R 3-75 3-25 3-75 3-83 P 15.6 13.7 14.8 14.0 3-64 14-5 3.60 4.00 4.40 4.05 14.2 14.7 14.4 14.5 3.80 4.22 14.45 14.45 During reaction: Ji is slightly inhibited at the end of the reaction. There is no definite change in V. O reports that he was confused be- cause one of the pipes at first failed to sound. He was not comforta- ble because the room was too cold. Summary of Results. Attention to difference tone is characterized by: i. Uinequivocal increase in rate of R (4 out of 8 cases). Unequivocal decrease in rate of R. 3 out of 8 cases. 437]i STUDY OF ATTENTION. 29 In I case the rate of R is faster than recovery, but less than that of normal. ii. Unequivocal increase in rate of P (3 out of 8 cases). Unequivocal decrease in rate of /"(i out of 8 cases). In 2 cases the rate oi Pis faster than recovery, but less than nor- mal. In 2 cases the rate of Pis the same as recovery, but less than normal. iii. Inhibited respiration in every case, although in some cases only slightly. iv. Fall in Fin 5 out of 8 cases. In 2 cases F is unchanged; in one case it rises, falls, and rises again. It may be questioned whether it was advisable to attempt two reactions during the same experiment, as was done in Exps. 61 and 62. The second reaction in both experiments has changes opposite to those of the first reaction. However, there is precedent for the procedure in Zoneff and Meumann's work. 2. Stimulus: Watch-tick receding and coming in. O sig- nalled when the watch ceased to be audible and when it again became audible. Exp. 63. Gai,i.oway. - i"-45" I"— 12" R 3.60 2.90 3.10 3.10 P 14.1 14.2 14.9 3.00 14.4 2.45 15.0 3-04 14.1 2.45 I5-0 i"-73" 3-75 3-6o 3-50 3-65 3-75 15.5 14.5 14.9 14.7 14-2 3.40 14.2 2.75 3-00 13.6 14.0 3-42 14.4 During reaction: R is greatly inhibited; indeed almost suppressed. V falls greatly at the beginning of the reaction, and remains low for 25 after the reaction. Exp. 64. Sabinb. i"— 40" 1"— 15" R 4.00 3.45 3.25 3.90 4.10 3.80 4.00 P 15. 1 14.7 15.2 15.9 15.0 16.2 15.6 3-65 3-96 15-2 15-6 i"— 72" 4.30 4.00 3.90 3.90 4-00 3.30 3.75 16.6 15.9 15-2 I5-0 15-5 14-6 3-88 154 During reaction: R is greatly inhibited; in fact, almost completely suppressed. V falls moderately and soon rises to normal. 30 STEVENS : [438 Exp. 65. STEVENS Baird. I' "-40" I "—41" t. 2.90 2.65 2. 20 2 .85 2.65 2.75 2.55 2 •50 ' 12.6 12.6 12.0 12 .1 II.9 12.5 12.3 12 •5 2.65 2.61 12.4 l"- -48" 12.3 2.50 2 .65 2 •55 2 .60 2.43 12.5 12.0 12 .0 II •9 11.6 2 ■54 During reaction: ^ is greatly inhibited. K falls decidedly. Exp. 66. Baird. I"— 40" i"-36.5" R 2.70 2.20 2.7s 2.75 3.00 2.75 2.50 2.23 P 12.7 13.2 12.4 II. 8 II. 8 11.9 II. 2 2.60 2.62 12.7 11.6 i"-55-5" 2.71 2.35 2.20 2.10 2.25 2.50 12.0 12. 1 11.4 11.4 11.8 11.8 2.35 II. 7 During reaction: R is markedly inhibited, although more at the beginning than at the end. The respiratory oscillation is suppressed. F falls slowly, and reaches a minimum after the end of the reaction. Exp. 67. Gaii" ■L'-'L^Yi" R 2.60 2.65 2.65 2.20 2.83 2.90 P II.O 11.4 10.9 II. I 12.0 13.6 2.50 2.86 II. I 12.8 3.00 2.60 2.85 2.90 2.70 2.50 13.0 II. 9 11.6 11.9 10.4 9.5 2.76 11.4 46 STEVENS : [454 During reaction: R is inhibited. V rises. There is a slight dip in the curve after the second signal. V falls during recovery. O reports that he visualized as before ; he multiplied 97 by 60, then subtracted 97. No muscular tension. Summary of Results. Attention to ideas (multiplication) is characterized by: i. Unequivocally increased rate of R (3 out of 7 cases). Unequivocally decreased rate oi R (2 out of 7 cases). In I case R is the same as normal, but faster than recovery. In I case R is faster than the normal, but less than recovery. ii. Unequivocally increased rate of P (4 out of 7 cases) . In 3 instances R is less than normal, but faster than recovery. iii. Inhibited R in every case. iv. Slight changes in V/ in two instances there is a rise in V (Exps. 112 and 113) ; in 4 cases a fall. In I case V is constant. V. Attitude Experiments. Stimulus: Konig fork A^. The following experiments were devised with a view to showing whether it is possible to obtain from the same characteristically different reactions to the same stimulus. It seemed possible to take three different attitudes to the same stimulus. These attitudes were described to O, who was afterwards required to assume in three succes- sive experiments: (i) an attitude of indifference to the stimu- lus, both as regards affection and attention; (ii) an attitude of active attention; O was asked to make a judgment as to the place of the tone in the tonal scale; (iii) an attitude of affec- tion exclusively; O was asked to give himself up to the pleas- antness or unpleasantness of the experience. Exp. 113. Baird. Ixdipfbrbnt Attitude. I "-43" i" — 27" R 2.4s 2.50 2.50 2.50 2.50 2.00 2.20 2.00 2.10 P 10,9 10.4 10.6 11. 1 II. 2 lo.o ii.o II. II. 7 2.48 2.20 10.7 II.O i"-55" 2.50 2.30 2.40 2.35 2.40 II. I 10.9 10.6 II. 2 II.O 2.48 10.9 During reaction: R is not inhibited. V rises during reaction. O reports that he did not know what to do. He began to introspect, to see whether he was behaving indifferently. He did not attend either to the pitch of the tone or to the affection. 455] STUDY OF ATTENTION. 47 Exp. 114. Bairb. Activb Attitude. i"— 47" i"— 12.5" R 2.40 2.30 2.60 2.30 2.50 2.00 2.00 P II. I 10.8 10.7 10.6 10.7 10.6 10.4 2.42 2.00 10.8 10.5 i"-65" 2.09 2.20 2.20 2.25 2.40 2.10 2.00 10.3 10.2 10.4 10.4 10.5 9.8 9.8 2.17 10.2 During reaction: R is slightly inhibited. V falls markedly. There is a sudden dip in the curve after the second signal. O reports that he was all at sea. He imaged the c of the great octave as a stand- ard; he knew that the stimulus fork was above that. Exp. 115. Baird. Affbctivet Attitude. I "-53" I "-16" R 2.20 2.60 2.35 P 10.7 lo.i 10.2 2.42 10.4 2.65 2.30 II. 10.3 I "-54" 2.00 2.00 2.50 10.4 9.8 IQ.3 2.15 lO.I 2.20 2.20 1. 16 2.20 2.25 9.9 10.3 lo.o 10.2 2.19 10. 1 10.2 During reaction: R is plainly inhibited. V falls slowly, then rises sharply. O reports that he had trouble in finding any affective tone. He compared the affective tone of the stimulus with that of the im- aged c of the great octave. He thought that the stimulus fork would be pleasanter if lower. About one-third from the end of the reaction the pleasantness began to appeal to him. Exp. 116. BENTLEY. Indifferent Attitude. I "—49" I "—27" R 3.20 3.50 3.1S 3.05 2.85 3.00 2.90 2.75 P 13-5 13-4 13-9 13-5 134 13-5 I3.6 13-5 3.15 2-88 13.5 13-5 I "-54" 3-25 3-25 3-20 3.20 2.73 13-5 13-8 13-9 136 13-3 3.12 13-6 During reaction : R is not inhibited. Kis slightly less. O reports: Passive at first; became a little excited by the thought of responsibil- ity ; this was unpleasant. After three or four seconds settled down. When tone tailed off felt some strain, but it was not at all intense. Attention was not constant. 48 STEVENS : [456 Exp. 117. Bbntley. Active Attitude. I "—45" 1"— ai)^" R 3-IO 3-55 3-25 3-50 3-4o 3-3° 3-27 P 14.5 14.0 14.2 14.3 14.4 14.3 14.6 3-36 3-28 14.2 14.4 3.20 3.10 3.50 3.55 3.40 3.12 14.5 14.7 14.0 14.0 13.6 13.0 3-31 13-9 During reaction : R is perhaps slightly inhibited. V does not change. O reports that he tried to get hold of some means of placing the tone; but was kept on edge. He decided that it was C" above mid- dle C. Got to reflecting on previous tone; the association was persist- ent; could not relax in recovery ; kept thinking of the diflSculty in placing the fork ; got to ruminating, and so got off from the experi- ment. Exp. 118. Bentley. Apeective Attitude. I "-47" I'-iS" R 3.40 3.50 3.80 3.20 3.28 3.40 3.00 P 14.7 14.2 14.5 15.0 13.9 14.0 3.43 3.20 14.6 13-9 I''— 61" 3-5° 3-15 3-25 3-00 370 3-50 14.3 14.4 14. 1 14.0 I3.3 13.5 3-45 i3-9 During reaction : R is slightly inhibited. V is not markedly changed. (9 reports a little inclination to compel enjoyment; said to himself 'that's a mellow, pleasing tone.' Pleasantness was weak, without much glow: dull, vague. Attention was directed to the smooth, sweet side of the tone. Exp. 119. Sabine. Indifferent Attitude. I"— 31 I"— 5" R 3.25 3.10 3.25 2.95 P 13.9 13.1 13.0 13.3 3.30 12.8 3-16 13-3 J:r I"— 8i" 3.60 3.50 3.30 2.70 3.40 12.7 13.3 13.1 14.0 13.6 3.00 3.30 3.II 12.7 133 13-2 3-24 13.2 During reaction : R is inhibited and continues inhibited for ten seconds after the reaction. F begins to fall during the reaction, at- taining a minimum after the end of the reaction. O reports curiosity just as the tone began to sound ; wanted to know whether it was high or low. Thought it was a pleasant tone; realized that he had sig- nalled too soon ; felt foolish. 457] STUDY OP ATTENTION. 49 I"- -51" R 3-75 3-oc > 3-00 P 13-3 12.7 13-2 3.26 13-0 Exp. 120. Sabinb. Indifpkrbnt Attitude. O was asked to make the reaction longer than in the last experiment. i"-5" 3-00 3.55 3.30 13.0 13. 1 12.0 3-30 12.0 I "—66" 3.40 3.10 3.00 3.00 2.75 3.20 3.30 12.8 12.7 12.8 12.7 II. 6 12.0 12.3 3-II 12.4 During reaction: R is noticeably inhibited. V falls sharply, at- taining a minimum after the close of the reaction. O reports no effort to attend; tone rather pleasant; attention was divided between the idea of signalling too soon and the tone itself. Not able to keep the pleasantness out. More pleasant than before. Exp. 121. Sabine. Active Attitude. I "—61" i"— 9'' R 3.40 3.45 3.20 3.00 3.30 3.20 2.55 P 12.0 12.7 12.5 12.7 13.0 13.2 13. 1 3-26 2.55 12.6 13. I i"-55" 12.0 12. 1 12.6 12.3 12.0 12.2 3.00 12.2 During reaction: one respiration is less than normal, one is more. R is inhibited before the reaction. V falls. O reports: Reaction very unsatisfactory because of confusion. Attention was distracted by click of signal; tried to run through the scale for ground of compari- son; asked himself, " How does this experiment differ from the last one?" There was a feeling of activity; consciousness of work to do; consciousness of movements in the throat. Exp. 122. Sabine. Affective Attitude. i"-53-5" i"-io.5" R 3.20 3.10 3.40 3.25 3.00 3.00 2.90 P 12.7 II. 6 12.3 12.6 11.8 12.0 11.4 3.12 2.90 12.2 11.4 i"-65" 3.30 3.20 3.05 3.00 3.35 3.05 2.10 12.6 12. 1 10.5 ii.o 11.7 II. 5 12.8 3.00 II. 7 During reaction: R is inhibited a little. V falls sharply in the second half of the reaction. There is a fortuitous fall in F 30" after the reaction. It coincides with the very slow pulse (10.5). O reports tone not nearly so pleasant as he thought it was; as he listened, the 50 STBVENS : [458 pleasantness became flat and dull; disappointed; unpleasant at the end. Exp. 123. Gai^ // I "-30%" R 2.25 2.55 2.45 P 9.9 10.6 10.7 2.10 2.40 10.2 11.7 2.12 2.40 2.20 10. 1 10.6 10.3 2-35 10.6 I "-47' / 2.24 10.3 1.80 lO.O 1.70 2.00 9.7 9.8 2.05 9.8 2.30 lO.O 1.97 9.8 During reaction: R is considerably inhibited, more in the middle of the reaction than elsewhere. F falls sharply, reaching a low level. O reports a feeling of strain which he could not localize: it seemed as if the clicks were never going to come. Exp. 132. Sabine. Stimulus should have been a cold cylinder. E, seeing the volume fall and suspecting expectation, said 'no experiment.' i'— 50.5' R 3-50 3- P 15-5 15- 00 3-So 3-75 3-33 1 13.6 13.4 13.5 3-40 14.0 i"-7o.5" 3 12 .8c •9 1 3-65 I3-0 3-35 13-3 3-8o 3-55 14.0 13.5 3-52 13.2 3-50 3-00 13.0 13.0 O reports: nothing prominent in consciousness; rather surprised when E said 'no experiment;' thought something had gone wrong. There is second spontaneous fall in F 30 seconds before the end of the curve. E was probably wrong in attributing the first fall in F to expecta- tion, inasmuch as O's introspection did not bear him out. The three following experiments were devised with a view to secur- ing a high degree of attention without direct stimulation of the sense organs. It seemed possible to attain a moderately high degree of at- tention by letting ideas, as in a revery, take their own course through consciousness. The experiment began with normal indifference on the part of O. As soon as he became interested in his ideas, he sig- nalled; when his interest was interrupted, he signalled again; so that the period of the greatest clearness of ideas was marked off from the normal and recovery periods. 463] STUDY OF ATTENTION. 55 Exp. 133. BenTlby. I "-39" R P 2.80 14.0 3-20 3.05 13.6 13.6 3.12 13-7 I"— 86" 3-44 13-5 2-95 3-5 2.90 13-5 3.10 13.6 3-05 13-3 3-05 3-00 13-4 13-4 3-05 12.9 3-3° 12.6 3' 12 ,00 .6 3-04 13.2 During reaction : J? is, if anything, a little higher; but the differ- ence is very slight. V rises very gradually throughout the curve. O reports that during normal he was thinking ' Now I must get myself into this state;' there was more tension, more excitement than in re- action. During reaction attention went up and down; there was a high degree of attention because the ideas were clear ; but attention was not constant. Also there was a conscientious pulling together now and then, along with a feeling of responsibility. There were excite- ment and unpleasantness with the sense of responsibility. Good bodily tone; quite relaxed. Exp. 34. Benti,ey. I "—25" i"— 35' I "-3" R 2.80 3.10 3.40 2.50 2.75 2.75 P 13.4 13-5 12.8 13.4 13.4 12.6 3.10 2.67 13.2 13.2 i"-47'' 2.70 4.00 13.4 12.6 4.00 12.6 I"— lo" 3.00 3.25 2.80 2.80 2.85 13-2 13-5 13-4 13.5 13-5 3-00 13-5 3-" 13-4 3.00 13-5 During the first reaction, F is a little higher; there is a slight drop in the curve at the beginning of the second reaction. O reports that during normal he is getting adapted to the situation. During the first reaction, attention was flighty and there was some excitement. Un- pleasantness because attention was spasmodic and fitful. A little bit disappointed with the state of things. During the second reaction, he picked his way through ideas. Attention if anything was unpleasant Exp. 135. Benti,ey. I "—16" R 2.90 3.33 P 13.5 12.6 3-II 13.0 I "—no" R 3.10 3.10 3.60 3.00 3.3s 3.40 3-20 3-IO 3-05 3-25 3-IO P 13.6 13.5 13.6 13.0 13.0 13.0 13.8 13.5 13.5 13.7 13.0 3.20 13-3 During reaction there is no noticeable change in the depth of J?. V tends to rise gradually. O reports: normal free from affection; set- 56 STEVENS : [464 tling down. Reaction. Attention to auditory ideas and visual perceptions. O made a false signal, whicli made him uncomfort- able, because he thought it might confuse E. There was a strain of unpleasantness all through. An occasional idea with fleeting unpleas- antness. Moderate degree of attention; not sharp. Summary of Results of Last Three Experim,ents. i. There is no consistent change in rates of /'and R. ii. Slightly increased V and slightly deeper breathing occur in revery. iii. Probably sensory stimulation causes fall in V. iv. Apparently centrally excited feelings have very slight efifect on P, R and V. Exp. 136. Sabine. E traced with a pencil the letters H C Son O's cheek. O attempted to decipher them. I "—42'' I "—24" R 4.00 3.10 3.10 3.33 3.50 3.25 3.10 P 12.8 II. 5 12.0 12.5 II. 2 10.6 II. 2 3-38 3-28 12.2 ii.o I"— 58" 3.90 3.50 3.30 2.80 2.60 2.70 11.4 II. 6 II. I 11.4 12.4 12.2 3-13 11.6 During reaction: H is considerably inhibited. V falls markedly, reaching a minimum after reaction. C. General Survey op Results and Discussion. /. Changes of Rates of P and R. I. In order to put the main results of the study in form for discussion, the experimental data may now be tabulated. Since there were many equivocal results, either some rule must be adopted for their interpretation, or they must be excluded from the Table. But since these results may be instructive in them- selves, it was not deemed proper to exclude them. There are six sorts of equivocal cases: 1. Reaction rates of P and R faster than normal and slower than recovery. 2. Reaction rates of P and R slower than normal and faster than recovery. 3. Reaction rates of P and R same as normal and faster than recovery. 4. Reaction rates of P and R same as normal and slower than recovery. 5. Reaction rates of P and R same as recovery and faster than normal. 465] STUDY OF ATTENTION. 57 6. Reaction rates of /'and R same as recovery and slower than normal. The following disposition was made of these cases. 1,3, and 5 were counted with the unequivocal increases in the rate of P and R; cases 2, 4, 6, were counted with the unequivocal de- creases in rate. Since there were four physiological symptoms determined, there will be four main rubrics in the Table: rate of pulse; rate of respiration; depth of respiration ; and volume. Under each of these rubrics there will be three sub-heads, since each physiological symptom may vary in any one of three ways, viz. , P and R, with respect to rate, may increase, decrease, or remain unchanged; R, with respect to height, may increase, decrease, or remain unchanged; V may rise, fall, or remain unchanged. The experiments are grouped according to the sense department to which the stimulus appealed. Thus, there is a visual, auditory, and tactual group. The number of ex- periments is also given (Table I, p. 466). The Table shows characteristic differences of reaction for the different kinds of experiments. For the visual group the rate of R is very greatly increased. The rate of P, on the other hand, does not change consistently. In approximately half the cases, the rate of P is faster, and in the other half, the rate is slower. The auditory group shows another correlation. The rate of P in these experiments is as decidedly decreased as the rate oiR was increased in the visual group. The rate of R, on the other hand, in the auditory group is almost as equivocal as was the rate of P in the visual group. The tactual group varies in still another way. Both R and P are very decidedly slowed. The changes for multiplication resemble those for visual stimuli. This may be due to the fact that the O's as a rule visualized the figures. The other physiological symptoms seem to show no differentiation according to stimulus. The over- whelming result for all experiments is inhibited breathing and diminution of the V oi the arm.. In order that the character- istic changes in rate of P and R may not be attributed to the manner of distribution of the equivocal results, the unequivocal results are given alone, in a similar Table (Table II, p. 467). The Table of unequivocal results brings out the same fact as the Table of both equivocal and unequivocal results. In the visual experiments, the rate of R is always faster, while the rate of P, with the exception of the experiment with the Mas- son disc, is about evenly divided between faster and slower. In the second group, the rate of P is very decidedly slower, while the rate of R is in some cases faster and in others slower. The slowing of both P and R, in the third group, is even more pro- nounced than in the Table of mixed equivocal and unequivocal results. 58 STEVENS ; [466 Bo 1 p s ■c P. Tj- W t^ !>. \0 00 I-- 10 00 00 •* to r^ 1 ro 'i- n- N CT M M H -H H H M H C< ro M *N, c>. CO W m vD •* a> 00 00 to H •* 1 a H ■a u H J. 1- 5| 00 t^ lo 00 00 ■* to t^ a. H N -a 1 V Is W to M W M M ^ w ^ t^ m VO to CI « K5 ON 0\ VO VO 10 ^ 10 i-t ro 10 SO N CO « 10 « Tj- M M ■* ■c P •s C 5 MassoD disc After-image Radii of Circle Badii with Intrinsic Interest After-image with Intr. Interest Natural Interest Difference tone Watch in-out-in Watch out-in-out End of Tone Counting Beats Cold Cylinder Pressure Hair Comparing Size of Cylinders a 1 •a •3 467] STUDY OF ATTENTION. Table II. 59 Pulse 1 Respiration No. of Experiment Faster Slower Vn- changed Faster Slower Un- changed Experi- ments Masson disc 8 2 lo 14 After-image Radii of Circle Radii (Intrinsic Int.) After-image (Intrinsic Int.) Natural Interest 3 2 2 I I 2 3 3 I I 8 8 6 4 4 I 12 lO 7 7 6 Difference tone Watch out-in-out Watch in-out-in End of tone Counting beats 3 2 O I O I 4 4 5 4 4 5 I 2 3 3 I 3 7 2 8 7 5 10 8 Col'd Cylinders Pressure Hair Size of Cylinder I o o 6 3 3 I o I 6 2 I 8 4 3 Multiplication 4 o 3 2 7 The experimental results of this study force, then, the fol- lowing conclusion upon us. Either we must say that active attention is different in kind according as the object of atten- tion is a visual, auditory, or tactual sensation, in just the same way as we speak of visual and auditory types of memory; or we must say that the changes in the rates of P and Ji are not due to attention at all, but are due to the psychophysical process of the sensation; or, as a third possibility, to the co- operative influence of both attention and the psychophysical process of sensation. The first possibility may be correct, although at present there is no theory of attention which dis- cusses such a view.' As for the third possibility, the method of expression has assumed that the purely reflex effect of the stimulus, if any at all, is very slight, compared with ■the expression of the mental process. There is, however, positive evidence in favor of the second possibility. Mentz" says, "With tones and clangs, as well as with noises, in consequence of increasing intensity, there appears an increasing lengthening of pulse which, nevertheless, with high intensities is again decreased on account of unavoidable unpleasantness, and indeed even passes over to a shortening. The cause of the original 1 C/., however, Art. Attention in Baldwin's Diet., I, 86; also Titchener, Am. Jour, of Psych., XVI, 214. ' Phil. Stud., XI, 95. 6o STEVENS : [468 lengthening is probably a direct effect of the psychophysical process of the sensation, since this process physiologically spreads out more widely in the organism with greater inten- sities, and thus draws into sympathetic action not only the involuntary and voluntary muscles, but also the vessels and, with greater intensities the heart." Mentz gives results from Mosso's work which show the same effect of auditory stimuli. Ivchmann^ says with regard to the physiological effect of cold: "It follows that cold, in every case, produces a lengthening of pulse, except where it is painful; then it, like every other unpleasant stimulus, has as a consequence shortening of pulse. ' ' Warmth has the opposite physiological effect. Finally, Kelch- ner^ finds that "with pleasantness the pulse is accelerated, if the pleasantness is set up by taste stimuli; the pulse is slowed if tones and colors serve as stimuli." In the case of un- pleasantness, there was no differentiation of results accord- ing to stimulus. Changes in respiration were also independ- ent of stimulus. While it is true that there is not a uniformity among these purely physiological effects, as one might expect if the reaction were reflex, it is quite possible that different intensities as well as different qualities of stimulus give diff- erent physiological reactions. The Ritter-Valli phenomenon is a case in point. With a certain very low intensity of stim- ulus, the extensor muscles of the claw of the crayfish are stimulated; with higher intensities, the flexors.' Something of this sort may hold for smooth muscle. Again, it is known that exposure to low temperature depresses the vaso-constrictor nerves more than the vaso-dilator nerves. Also, if a mixed nerve is stimulated with rapid induction shocks, the effect is vaso-constriction; if with slow induction shocks, vaso-dilata- tion. Assuming the results to be purelj' reflex, there seems to be sufficient evidence, on physiological grounds alone, to account for divergences. It will probably be urged against this view that it takes account of only a few negative cases, while it neglects a well confirmed correlation, that between changes in rate of P and R and active attention. We propose, therefore, to examine the experimental evidence for this correlation. The results may be thrown into a Table (Table III) which will include the name of the investigator and the date of his work; the stimu- lus to attention and the resultant changes in rates of P and R, together with the number of experiments and the number of O's, so far as these data are obtainable. The plus and minus ' Die korp. Aeusser. psych. Zustande, 114. ^ Arch. f. g. Psych., vol. V, 39. ' V. Schafer's Physiol., II, 480, for a description of this and similar phenomena. Fold out '^9 469] STUDY OF ATTENTION. 6 1 signs mean acceleration and retardation of P and R. ± means acceleration followed by retardation. TABI.E III. With regard to the rate of P, the Table shows uniformity, with three exceptions: (i) the slowing of P with tactual stim- uli, in McDougall's experiments; (ii) the slowing of P in all of ZonefF and Meumann's experiments; (iii) the denial of correlation by Angell and Thompson. There is also the fact, first pointed out by McDougall, that frequently the rate of P first increases, then decreases even below the norm. Binet and Courtier confirm this result, as does Gley in his later work. Unfortunately, most investigators have given only the main results of their work. The total number of experiments, and especially the number and kind of discrepant results, are almost always lacking. The rate of R is uniformly increased, with three exceptions: (i) Mosso found no correlation; (ii) McDougall found slowing of .^ with tactual stimuli; (iii) An- gell and Thompson deny correlation. And Lehmann found increase in rate of R only under certain conditions. On the other hand, Delabarre, Mentz, McDougall, Binet and Courtier, and Zoneff and Meumann find, on the whole, an increased rate of .^, although had Delabarre taken more observations from other O's than 'D,' his results would, apparently, have been difierent. A fact which may possibly help to explain the universal quick- ening of ^ is the almost universal use of visual stimuli, such as reading, counting dots, arithmetical operations, which were probably performed visually, and reading-over nonsense sylla- bles. Of course, against this suggestion, it can be shown that Mentz used auditory stimuli mainly and nevertheless found quickened rate of ^, although there were many cases of retarded R, and that McDougall did not use visual stimuli at all. But on the other hand, Zoneff and Meumann make 19 experiments with optical stimuli against 12 with acoustic and tactual together. A reasonably long series of experiments with auditory and tactual stimuli has never been made. What- ever may be, the cause, the present status of the matter is this: for pulse, (i) two contradictory correlations, (ii) a denial of correlation; for respiration, (i) a uniform correlation, (ii) a partial contradictory correlation, and (iii) two denials of correlation. We think, therefore, that the appeal to the facts does not sustain the objection that we have based our thesis on a few irregularities in the face of a well established correlation. The well established correlation does not exist. And furthermore, what correlation there is depends on a narrow range of ex- 62 STEVENS : [470 periments, in which discrepancies have been largely neglected, carried out on a small number of O's. Again, it may be objected that the discrepancies in our re- sults can be explained by differences in the state of attention itself. It may be said that attention to minimal pressures and cold spots is not of the same sort as attention to a Masson disc. In the one case attention is powerfully attracted; in the other it has to be sustained by effort. In short, the tactual and part of the auditory experiments are expressions of passive attention, while the visual experiments are expressions of active atten- tion. This theory of the results would bring about a happy compromise. According to it, the results would confirm Mentz' thesis that active attention accelerates P and J?, while passive attention retards them. And so the question of the validity of the method would not be raised at all. This view has a certain amount of plausibility, but it leaves out of account several considerations, (i) The known physiological fact that stimulation of a centripetal nerve or of a sense organ does pro- duce reflex acceleration or slowing of the heart. There is no uniform result (siich is the impression one gets from Tiger- stedt's review' of the literature) from this sort of stimulation; but the important point for this argument is, that the heart does respond reflexly. (ii) The purely physiological effects of certain stimuli, demonstrated by Mentz, Lehmann and Kelch- ner. (iii) The difficulty of discovering a reason for calling at- tention to tactual and auditory stimuli passive, and attention to visual stimuli active. Attention to a liminal pressure or auditory stimulus is, surely, just as active as attention to a limi- nal visual stimulus. A third objection may be made: that we use the method of expression to obtain results, while we deny that the method is expressive. This objection is merely verbal. The method, apparently, is not expressive in that it consistently character- izes a total state which we call active attention; but it is ex- pressive in the sense that it responds to certain physiological processes which may vary while the total state of attention, as regards degree and kind, remains the same. That is, the method is expressive for certain inessential features of the at- tentive experience; but it is not expressive for the state as a whole, which is itself the object of investigation. The argument of this paper, up to this point, may now be summarized. The method of expression has failed in the do- main of feelings. Two possible reasons were suggested: (i) complications with other mental processes, and (ii) the psy- chophysical processes of the sensation. The results of this 1 Phys. des Kreislaufes, pp. 282-289, 1893. 47 1 ] STUDY OF ATTENTION. 63 study point definitely to the second factor, unless one is willing to speak of visual, auditory and tactual attentions, as we speak of visual or auditory memories. Three objections were brought against this view, (i) It is based on a few discrepancies from a well established correlation. Reply: the correlation is not well established. (2) The results are the expressions of active and passive attention. Reply : liminal stimuli all require active attention, yet the expression is different. (3) We deny that the method of expression is expressive, and yet make use of its results. Reply: this objection is merely verbal. //. The Significance of the Changes in Volume. So far, we have considered only two symptoms of the expressive method, viz., changes in rate of P and R. It may now be asked whether the changes in V and depth of brealhing are expressive or merely reflex. In these experiments, as Table I. shows, F'and depth of R changed consistently with active attention. Wherever there was a state of active attention, there were always a fall in V and inhibited breathing. These two symptoms, therefore, would seem to be the characteristics of active attention. But there is, at least, positive evidence against the F change. Just as the changes in rates of Pand R brought suspicion on themselves by a too pliant response to experimental conditions, so the F change brings suspicion on itself by a too indiscriminate response to experimental condi- tions. The all but invariable reaction in our own experiments was a fall in V. The all but invariable reaction in the whole literature of the expressive method, whether applied to the feelings or to attention, is a fall in V. We conclude, there- fore, that every sensory stimulus tends to produce a fall in vol- ume of the arm. The evidence for this thesis is as follows, (i) In exps. 133, 134, 135 of this paper, in which there was a state of active attention without sensory stimulation, Fshowed no tendency to fall, (ii) The greatest and most sudden changes in Fcome at the very beginning of the reaction, during or im- mediately after the period of stimulation, although it is im- probable that the highest degree of attention has been attained at that point, (iii) The fact, already mentioned, that fall in V independent of mental conditions has been the almost uni- versal reaction in all experiments with the plethysmograph. A detailed review of the results of the method of expression as . regards V will, we think, bear out this statement. The re- sults may be tabulated. The name of the author, the date of his work, together with the kind of stimulus, affective or at- tentive, and the change in volume expressed by a plus or minus sign to indicate a rise or fall respectively are given. The Table shows a uniform decrease in V for all mental con- 64 STEVENS : [472 Tabi,E IV. Investigation Mosso 1879, 1881 F6r6 1887 Lehmann 1892 Hallion and Comte Shields Binet and Henri 1895 Binet and Courtier 1896 Angell and McLennan 1896 1894 1895 McDougall 1896 Angell and Thompson 1899 Lehmann Gent Gley Bonser 1899 1903 1903 1903 Stimulus Every psychic activity Pleasant Unpleasant Pleasant Unpleasant Tactual, auditory pain Olfactory sensations and mental work Cold, tactual sensations Pleasant and unpleasant emotions Intellectual work Pleasant and unpleasant emotions Active attention Agreeable Disagreeable Watch tick, mental arith., recall of past events Emotions, intellectual operations, sensations Voluntary attention Involuntary attention (fright) Spannung Unpleasantness Pleasantness Strain Relaxation Pleasantness Depression Mental arithmetic Mental arithmetic + + - 75% + 25% + ? % - 90% + + + 473] STUDY OF ATTENTION. 65 ditions, with the exception ofFdr^'s, Lehmann's and Gent's work. Angell and McLennan found a rise in V with agree- able stimuli in an unspecified percentage of their experiments. But since they do not publish curves, or state their results pre- cisely, one is justified in not taking their results into account. We propose, now, to examine, so far as this is possible from their published results, the evidence for the rises in volume asserted by 'B€r€, Lehmann and Gent. 'B€r€, Sensation et Mouvement, finds that sensory stimuli when tested with a dynamometer fall into a dynamo- genie series. In chapter XVII, he finds that taste, smell, auditory and visual stimuli, when tested with the plethysmograph, also arrange themselves according to the degree of their effects. F6r6 does not say, as he does in the case of the dynamometry experiments, that the change in V is correlated with the affective tone of the stimulus. On the contrary, at least in the case of diffusible excitants as alcohol and ether, he suggests that the type of reaction is due to intensity of stimulus. "Uue excitation faible produit la dilatation; une pen plus forte, la dilatation suivie de retraction proportionelle; une excitation tr&s forte, une retraction immediate" (p. 108). On p. 113 Ffir^ says, "Toutes les Amotions ddpressives d^terminent une diminution de vol- ume dans les membres; toutes les Amotions excitantes ou agrdables dSterminent un effet inverse." We have failed to discover any curves in either edition of Sensation et Mouvement which exemplify this conclusion. I,ehmanu, Die Hauptgesetze etc., publishes nineteen curves, of which thirteen are reactions to simple pleasant or unpleasant stimuli. The curves with the initial of the observer and stimulus are as follows: Sugar solution. Eau de Cologne. Saccharin. io% quinine sulphate. Lukewarm water with a few drops of tartaric acid. Hand in 50° water. Unpleas. Pin stuck in nates. Carbon bisulphid. Told it was pleasant. Arm in water 43°. Decidedly pleasant. Tobacco cigarette. Pleasant. " " Unpleasant. Wall paper. .Esthetic pleasure. Tuning fork sympathetically actuated. Intellectual pleasure. Fright. Fright. Depression. Fear. Fright. Anger. These experiments, where possible, were carried out on two other observers. In all there must have been fifty-seven curves, of which thirty-nine were simple feeling curves. But we are here concerned only with the curves of pleasantness. These curves are I A, I B, I C, II E III A, III C, IV A. If these experiments were repeated on two other observers, there must have been twenty-one curves to generalize from. These curves must now be examined with regard to the amount and character of the rise in V and its coincidence with the period of stimulation. I A. M IB. C IC. M ID. D II A. D II B. D lie. M II D. M HE. Ill A. D Ill B. inc. M IV A. D IV B. M IV c. E IV D. D V A. E VB. E V C. M 66 STEVENS : l474 I A. The K before stimulation is very low. Just after stimulation there is a sudden though very slight rise in V. The V remains almost level for thirty seconds, although there is a slight gradual in- crease, which reaches a maximum twenty-eight seconds after the stimulus was applied. The absolute and relative change in V is so slight and comes so long after the stimulus that it seems very doubt- ful whether any positive conclusion can be drawn from it. I B. There is undoubtedly a rise in K coincident with the period of stimulation. Although it should be pointed out that the V was low when the stimulus was applied. I C. This curve is very much like I A. The normal is very low and the reaction is very long delayed (due to the dissolving of the saccharin, Lehmann says) and very slight when it comes. About twenty-two seconds after the substance is tasted and forty seconds after it is administered, there is a slight rise in V. II E affords a good illustration of the freedom of interpretation that is possible in work of this kind. Lehmann says that for ten seconds after stimulation the K diminished, then increased until it exceeded the normal level. So it does; but not until thirty seconds after stim- ulation. Without some good introspective reason, it seems inadmis- sible to call a subsequent rise in V, after an initial fall, the expression of an affective process. III A. Lehmann calls this curve an ausgepragte Lustcurve. The fact is this. The V for nearly thirty seconds after stimulation is less than that at the time of stimulation. A decided rise in V does not occur until forty-five seconds after stimulation. Furthermore, the rise in V does not occur so soon nor is it so marked when it occurs, as is the rise in Kin III B, the unpleasant tobacco curve. In III B a very decided rise in K occurs a few seconds after the stimulus is ap- plied. There is a low V for twenty seconds, with a very decided increase before the period of stimulation is over. This curve is very much like II B in its general type. The stimulus for II E was entschieden lusterregend. Lehmann finds a physiological reason for the fall in V in II E, although, apparently, he is willing to attribute the fall in volume in III B to the unpleasantness of the tobacco. If II E is called a curve of pleasantness, we see no reason why III B should not be so called. III C is a curve of sesthetic pleasantness. The K falls for about ten seconds, then slowly attains and gradually surpasses the level before stimulation, and sinks slightly below the normal level before the stimulus is removed. Here, again, the curve has the same general aspect as II E, III A and III B, namely, a low level for several sec- onds after stimulation, with a gradual return to the normal level or above it. But there is a positive rise in this curve, coincident with the period of stimulation. IV A. This curve is an example of intellectual pleasantness. The stimulus was the perception of a tuning fork actuated sympathetically by another fork. The K begins low, and had already begun to rise before the stimulus was applied. It continues to rise throughout the period of stimulation, reaching a maximum after the stimulus has ceased. TJie result of this examination is that out of seven curves of pleas- antness in only three is there an unequivocal rise in V coincident with the period of stimulation. And in each of these curves there is some questionable circumstance. In I B and IV A the Kwas low and had already begun to rise when the stimulus was applied. In III C, the rise in V follows a marked fall in V. Lehmann gives fifteen examples of Lustzustande in the KSrp. A'uss. 475] STUDY OF ATTENTION. 67 psych. Zustdnde. These experiments are described in the text, pp. 13I-136- The list of curves, with the initials of the observer and the stimulus, is as follows. I. XLIII D. P. L. Smell of saffron. 2. XLIII B. P. L. Nitrobenzol. 3- XLIV A. A. L. Menthol. 4- XLIV B. A. L. Chloral. 5- XLIV C. P. L. Patchouli. 6. XLIV D. S. N. Attar of roses. 7- XLV A. A. L. Saffron. 8. XLVB. Dr. B. Menthol. 9- XLVC. A. L. A bit of sugar. 10. XLVD. P. L. Powdered chocolate. II. XLVI A. H. K. Chocolate cake. 12. XIvVI B. P. L. Cheerful tune. 13- XLVI C + D. A. L. Colored photograph. 14. XLVI E. A. L. Spontaneous pleasant memory. IS- XL VII A. A. L. Satisfaction at solution of easy problem, No. I. About nine seconds after the beginning of stimulation, the curve reaches its lowest level. It then rises sharply, reaching the normal level eighteen seconds after stimulation and finally exceeding the normal niveau. No. 2. There is a slight rise in V jusl after stimulation, followed by a fall in V, which remains low. Lehmann does not call this a rise. Nos. 3, 4, 5. Show short, sharp rises in V either in part coincident with the stimulus or entirely so (as in No. 5). No. 6. There is a slight rise in V coincident with the stimulus. But one can hardly refrain from the thought that had the stimulus occurred a few seconds earlier, under a respiratory oscillation, the re- sult would have been just as good. No. 7. The Arises sharply out of the normal level coincident with the first half of the period of stimulation. It falls during the last half of the stimulation, reaching a lower level than the normal. No. 8. A very decided fall in V, which Lehmann suggests may be due to the very long deep respiration and strain of attention. No. 9. The stimulus is followed by a long, low fall in K which never exceeds the normal level. No. 10. The stimulus is applied at the lowest point in the curve; the V gradually rises, reaching the normal level about eight seconds after the chocolate was tasted, and surpassing the norm. No. II. The stimulus is followed by a fall in V, which at no time afterwards exceeds the normal level. No. 12. The F falls during half the time the melody was played; it rises during the other half. But it does not exceed the normal level at any time during the period of stimulation. Afterwards it rises con- siderably. No 13. Closely resembles No. 12. The V is low during the whole period of stimulation; it rises, and remains continuously high, after the exposure of the picture. No. 14. The V begins to rise with the first signal and rises steadily to an acme just above the second signal. No. 15. The curve rises slightly after the first signal, falls, then begins to rise a little before the second signal. It does not reach a high level at any time. The review shows that positive rises in V coincident with stimula- tion occur in curves 3, 4, 5> 6, 7i 1°. and 14. That is, in seven out of fifteen cases. An examination of all the curves reveals three kinds of 68 STEVENS : [476 rises in V. (i) An initial rise coincident with the period of stimula- tion. As in curves XLIV A, B, C, D, and XLV A. (ii) A rise when the normal V is very low, as in curves XLV D and E. (lii) A rise which follows a more or less prolonged fall in V. It seems possible to bring all of these curves, with the exception of XLV D and E, under one head. The general character of the curves is the same. There is a positive rise in For, at least, a tendency to rise just after or along with the application of the stimulus. This initial rise is followed by a fall in V, after which the V regains the normal level or exceeds it. Instances of this general character are curves XIvIII D and E, XLIV A, B, C, D, XL A, B, C (not D), XLVI A, B, C (not E) and XLVII A. All the curves, with two exceptions, conform to this general type. Of course, the three parts of the curve, the initial rise (or tendency to rise), the fall and the subsequent rise may vary considerably in different curves. For example, there is no tendency for the Kto rise just after stimulation, in curves XLVI A, B and C. But the subsequent fall and rise is, apparently, a constant character of every curve mentioned. Now these two rises, the initial, and the final rise after the fall, permit great latitude of interpretation. If the initial rise is very slight, the expres- sion of the pleasantness is looked for in the final rise, as was the case in XLIII D and XLVI B, C. On the other hand, if the initial rise is considerable, even if the final rise is large, the first rise is called the expression of pleasantness. Witness curves XLIV A, B, C, D, and XLV A. The fall in Fand subsequent rise occur in these curves as in all the others ; but the characteristic expression of the feeling is said to be the initial rise, which in these cases is large. This license in interpretation is responsible for much of the apparent success of the method. So long as the expression of the feeling may be looked for anywhere from 2 seconds to i minute, after the time of stimulation, without an exact introspective check on the interpretation of the curve, one must expect that arbitrariness of interpretation with which not only Lehmann but others after him may be justly charged. But, waiving that objection, the fact remains that there are seven positive rises in F coincident with the stimulus. This number is not sufficient to establish the correlation, even if those curves were unobjectionable. But it may be urged against them that in the best examples, XLIV A, B, C, and XLV A, the sudden rises occur where there is considerable disturbance of the breathing, due, probably, to the use of olfactory stimuli. And in one case, XLV D, the stimulus was applied when the V was very low. Gent' describes four reactions of mixed pleasantness and strain, al- though only one curve is published. The curves with the initial of the O and the kind of stimulus are as follows: 1. Ch. Smell of fresh lemon. 2. Ch. Menthol. 3. Ha. Violet. 4. Wi. Violet. (Fig. 13.) No. I. We quote Gent's description of these curves, with the excep- tion of the last. "Indie Reizphase tritt die Volumcurve ansteigend ein mit wachsenden Pulshohen und wird in ihrem Fortgange kaum von dem Geruchreiz alteriet. Sobald derselbe sein Ende erreicht hat, sinkt er plotzlich, so dass es den Eindruck macht, als sei der bisherige Hochstand des Niveaus nicht der normale gewesen" (p. 757). No. 2. "Die Volumcurve tritt ansteigend ohne Spontanschwank- ungen und Respirations-oscillationen in die Reizphase ein" (p. 758). iphilosophische Stud., XVIII, 757. 477] STUDY OF ATTENTION. 69 No. 3. "Das Beispiel bietet den allgemeinen Ztigen nach dasselbe Verhalten dar, wie die bereits besprochenen Beispiele, nur dass hier das Armvolumen nicht ansteigend, sondern mit Neigung zur Senk- ung in die Reizphase eintritt" (p. 758). No. 4. The V rises very slightly for three pulses, falls for six pulses and rises during the remainder of stimulation, reaching a max- imum after stimulation. This curve has the same general character as those of Lehmann already discussed. There are an initial rise, fall, and final rise in V. It is an important matter for interpretation to knowwhich of these phases is the essential reaction. Of these four curves there seems to be only one. No. 2, which shows an unequivocal rise in {^coincident with the stimulus. The first curve is 'scarcely altered' by the smell stimulus. No. 3 yields a fall in V. No. 4 is half-rise and half-fall. Yet, on the basis of these experiments, Gent feels justified in concluding that: "Das Gefiihl der Lust erzeugt imnier ein Anschwellen des Armvolums mit Zunahme der Pulshohen" CP- 759)- Gent has also some Fcurves to show the effect of pleasant emotions. But he concludes that "das Armvolumen zeigt weder ein constantes Steigen noch Sinken" (p. 781). Gent finds that the feeling of relaxation {Losung) is characterized by a rise in V. He describes five experiments with that feeling, the stimulus to which, in every case, is a problem in multiplication. The characteristic reaction is a fall in V, while the problem is being solved {Spannung'), followed by a rise in Y, which may begin before the problem is solved, and which continues to rise above the normal level after the solution is reached. It is possible that this rise in V, after stimulation, is an active expression of the feeling of relaxation, as Gent holds. But it seems more likely that it is' a purely passive restoration of physiological equilibrium which has been disturbed by the stimulus.^ Only one curve (Fig. 8) is published, so that it is not possible to give a detailed criticism. But it seems unlikely that the satisfaction of performing a simple arithmetical operation should express itself by such a powerful and continuous rise in Kas is shown in Pig. 8, while it is next to impossible, by any sort of pleasant sen- sory stimulation, to obtain even a very moderate rise in V. We have now fini.shed the review of those instances of rise in V on the basis of which a correlation with pleasantness has been established. Fer6 has published no curves with which we are acquainted, showing the effect of pleasant and unpleas- ant stimuli on V, although he is the author of the correlation. In I^ehmanu's first work there are three rises in V coincident with the stimulus; in his second work there are seven rises coincident with the stimulus. In Gent's curves, there is one rise in V coincident with the stimulus. Therefore, there are eleven rises in V coincident with a pleasant stimulus in the literature'' of the expressive method. This paucity of positive rises in V has not escaped other observers. Lehman n says, "It is a well known circumstance that the bodily ex- ^ Cf. Zoneff and Meumann, Phil. Stud., XVIII, 63, for the similar view that after retardation of /"by pleasantness, P tends to increase from purely physiological causes. , 2 F6r6 cites the results of L'bermiuier, Etudes plethysmographiques en psycho-physiologie. These, Bordeaux, 1897, in support of his theory. I have not been able to consult this paper. 70 STEVENS : [478 pressions of pleasurable states can be demonstrated only with great difficulty. While I found earlier (in the Hauptgesetze) that all sim- ple pleasurable feelings are accompanied by increase in F, in addition to increased height of pulse, most later observers in this field have been able to demonstrate no essential difference between pleasant and unpleasant stimuli. As we shall see in the sequel, such a difference does nevertheless exist; and if authors like Kiesow, Shields and Binef are not able to find this, probably a whole series of interesting cir- cumstances is at fault" (Korp. Aeus. psych. Zust., 128). Lehmann objects to Binet and Courtier's work that they investigated emotions and that children were observers. He suggests that the difference may have been overlooked by Shields. Lehmann gives four reasons for the paucity of rises in V with pleasantness, (i) It is difficult to set up strong feelings of pleasantness, (ii) Since the feelings are weak the curves must be represented very plainly, especially with reference to length and height of pulse, in order that small differences may be visible, fiii) The observer must be kept in a state of affective equilibrium, (iv) It is impossible to avoid active attention, which has the opposite expression of pleasantness. In reply to these reasons it may be said (i) that it is true that the objective expression of pleasantness is difficult to secure. Such, at least, has been the expe- rience of the Cornell Laboratory.' But to attribute this difficulty to the weakness of the pleasantness during the experiment without valid reasons for the attribution and in face of the fact that in ordi- nary life pleasantness is easily excited is almost a question begging argument, (ii) This does not apply to V but merely to changes in length and height of pulse, (iii) This is undoubtedly true, (iv) This reason involves a discussion of the relation of attention to feeling, which is not within the province of this paper, and in the second place, although it may be a difficulty, Lehmann himself has proved that it is not insuperable, by the fact of his own curves. ///. Summary of Conclusions. The fourth physiological symptom with which this paper is concerned, namely, depth of breathing, must remain the sole characteristic, the validity of which is not impugned either by our own results or those of other observers. As we have al- ready said, with a few exceptions it is the universal character- istic of active attention. The position of this paper, therefore, with regard to the four physiological symptoms may be sum- marized in three theses. (i) Changes in rates of P and R are brought about by the psychophysical process of sensation. (ii) Every sensory stimulus (probably in proportion to its intensity) tends to produce a fall in V. (iii) Inhibited ^ is a characteristic of active attention. IV. Classification and Discussion of Equivocal Cases. A classification and analysis of the equivocal cases brings out some interesting relations as to the sensitiveness to change ' Cf. Titchener, Exp. Psych., Instructor's Manual — Qualitative, 181, for a curve of pleasantness. 479] STUDY OF ATTENTION. 71 of the circulatory and respiratory functions. We assume that if either function reacts pliantly to the conditions of the ex- periment, there will be few equivocal cases. On the other hand, if the function is inert, there will be many equivocal cases. _ The following Table shows the total number of experi- ments in each sense department, together with the number of equivocal reactions of P and R. Tablb V. Equivocal Cases R p of Bxpts Masson disc After-image Radii of circle Radii with intrinsic interst After-image with intrinsic interest Questions in philosophy 5 4 2 .1 3 2 17 5 7 5 2 5 4 28 14 12 10 7 7 6 56 Difference tone Watch from out in Watch from in out End of tone Counting beats I 4 I I I 1 I 4 3 4 7 t4 8 7 5 ID 8 38 Cold spots Minimal pressure Comparing size of cylinders 1 I 2 I I o 4 2 .8 4 3 15 Multiplication The Table shows that there are fewfer equivocal cases of R in the visual and auditory experiments than there are equivo- cal cases of P. On the other hand, there are more equivocal cases of R in the tactual group than of P. This result, if our assumption as to the meaning of an equivocal case is correct, means that the respiratory function is more sensitive to audi- tory and visual stimuli than is the cardiac function; but that the heart responds more readily to tactual stimuli than does respiration. The Table also shows that the total number of 72 STEVENS : [480 equivocal cases with relation to the whole number of experi- ments is less in the tactual and auditory experiments than in the visual. The proportion 1345:57 for visual; 21:38 for audi- tory; and 6:15 for tactual. This fact would seem to indicate that both P and R respond less readily to visual stimuli than to tactual and auditory stimuli. V. Remarks on the Experiments with Intrinsic Interest and on the Attitude Experiments. It remains to say a word about the experiments with intrin- sic and natural interest, and the 'attitude' experiments. The former experiments were suggested as a means of varying the central conditions of attention while the peripheral conditions remained the same. It was thought that if peripheral factors were the cause of the changes in rate of P and R, the case would be strengthened if it was shown that a wide variation of central conditions made no essential difference in the results. Of course, such a conclusion is valid only on the assumption that the central conditions would express themselves differently from the peripheral conditions. Apparently, the variation in conditions made no marked difference in result. The latter experiments were instituted with a view to showing whether it is possible for the same observer to give consistently different reactions to the same stimulus. If it were possible, it was thought that it would show a possible ground for the discrep- ancies with which the literature of the expressive method abounds. The changes in V and depth of R show no differen- tiation according to attitude. The reaction for the most part is fall in V and inhibited R. The rates of P and R for the in- different attitude do not change consistently within the group, although the pulse shows an evident slowing. The active and the affective attitudes, however, are differentiated quite .sharply by the changes in rate of P, which in the active attitude is faster (except one reaction), and in the affective attitude is slower. VI. Concluding Remarks. The writer, after consideration and in the light of all the facts, has felt under the necessity of saying that the method of expression as applied to the study of the feelings and of attention has failed. When one compares Lebmann's extrav- agant suggestion^ with the actual state of this branch of Psy- 1 "Sind alle diese Erscheinungen erst untersucht und ihre charac- terischen Aeusserungen festgestellt, so wird man am Plethysmographen ein wirkliches Psychoskop besitzen, einen Apparat, mittels dessen man mit nicht geringer Sicherheit den Gemiitszustand einer Person zu diagnostizieren vermag." Korp. Aeuss., etc., text, p. 216. 48 1 ] STUDY OF ATTENTION. 73 chology, one cannot but wonder at his temerity. If the position which the writer has taken with regard to the changes in rate of pulse and of respiration and in volume is correct, the plethysmograph will never serve as a psychoscope for the diag- nosis of affective processes. And if by any means the method is rehabilitated, it will require more rigorous rules of use than have yet obtained, to secure results of any degree of certainty. D. Appendix. Table VI was made in support of the statement that a long series of investigations -with the method of expression had ended in contra- dictory or negative results. The negative results are those of Shields, Binet and Henri, Binet and Courtier, Angell and Thompson and Bon- ser. file contradictory result is that of Kiilpe, as regards the rate of pulse in unpleasantness, and that of Gent, as regards rate of pulse for pleasantness. The height of pulse in pleasantness and unpleas- antness is consistent for those who have observed it. But many have failed to observe it at all. Zoneff and Meumann alone find an opposition in rate of respiration between pleasantness and un- pleasantness. Binet and Henri and Binet and Courtier found no diff- erentiation. Lehmann found an increased height of respiration (he characteristic of pleasantness in his earlier work. He does not confirm it in his second work. Zoneff and Meumann find the opposite change in height of respiration for pleasantness. For changes in Klvchtnann agrees with himself and with Gent. F6r6, apparently, has not pub- lished any affective curves. Angell and Mclennan's results are not stated precisely enough for treatment. The other observers have either found vasoconstriction for all stimuli or they deny any kind of correlation. Rbfbrbnces. We append here a list of the works referred to in this study. Mosso, Pi.., Die Diagnostik d. Pulses. 1879. Ueber den Kreislauf d. menschl. Gehirns. 1881. VON Thanoffer, Der Einfluss d. Gehirns- Thdtigkeit auf den Puis. Arch. f.,d. g. Phys., XIX, 254. Gley, E., Etude expirimentale sur Vital du pouls corotidien pendant le travail intellectuel. Arch, de Phys., normale et pathol., 1881. [I have not been able to consult this paper.] Fere, Ch., Sensation et Mouvement. 1889, 1900. Delabarre, E. B., V influence de V attention sur les mouvements res- piratoires. Rev. philosophique, 33, 639. 1892. Lehmann, a., Die Hauptgesetze des menschlichen Gefuhlslebens. 1892. KtJl