taste for bitter seems to be most frequently absent, while that for sour is most commonly present.

"To study this taste of the brightest feeble-minded children, solutions of the above substances were made, and these were added a little at a time to a quantity of water until the subject perceived the proper taste, the tongue being cleaned and dried at proper intervals. From the quantity of taste solution and water taken the strength of the solution tasted could be calculated. Twenty children were tested. For fourteen of these the averages were:

"For sugar, 1.3 per cent., M. V. 0.4, or 1 part to 77; for sal 0.48 per cent., M. V. 0.19, or 1 part to 208; for acid, 0.41 per cent M. V. 0.40, or 1 part to 244; for quinine, 0.0177 per cent., M. V 0.0062, or 1 part to 5,694.

"The strength of solutions for normal people as determined b Bailey and Nichols is for sugar 1 part to 199, for salt 1 part to 2,24 for sulphuric acid 1 part to 2,080 and for quinine 1 part to 390,00 A marked dullness throughout, and especially so for bitter.

"Six of those tested were unable to recognize the taste of one more of these substances even when used full strength. With tw of them salt was wanting, with three of them sour, and with two them bitter, in every case being identified with another taste.

"Hence, with the feeble-minded, as with criminals, we find marked dullness of taste. With a few some of the tastes are lac ing, while with the lowest grades only one or two may be presen

"The reaction times were taken with a Hipp's chronoscope an the common touch key; for sound an electric hammer was use The current through the instrument was reversed after each readin Of those making the test twenty-two-sixteen boys and six girls were considered to have made the test properly. Their ages vari from eight to thirty-seven years, the average being twenty year The average number of tests for each was twenty-one. The avera reaction time was 0.338 seconds, with a mean variation of 0.08 se onds. The normal reaction time lies between one tenth and tv tenths seconds, say 0.148 seconds, and the mean variations can put at 0.04 seconds.

"In the number of tests made the effects of fatigue and practi seemed to about balance each other. Twelve other children we tested. Of these five were unable to perform the test, and sev of them had an average reaction time of 0.538 seconds, with a me variation of 0.164 seconds.

"Sixteen children made the tests for sound reaction. The avera number of tests was twenty-four. The average reaction time w 0.293 seconds, with a mean variation of 0.085 seconds. The norm time has been found to lie between 0.120 and 0.180 seconds, with mean variation of 0.022 seconds. However, fifty-three men in t University of Minnesota gave an average reaction time of 0.2 seconds. The reaction time of the insane to sound has been fou to vary from 0.200 and 0.340 seconds, with a high mean variatio On being directed to be quick, eleven children, in a series of fi tests, succeeded in reducing their reaction time by an average 0.012 seconds. Five who did the touch reaction tests were unable do the sound reaction tests.

Of

"Among the children tested there were fourteen Mongols. these eight were regarded as having made the tests successfully. Their average touch reaction time was 0.396 seconds, with a mean variation of 0.095 seconds. The average of the other six was 0.570, with a mean variation of 0.160 seconds. Seven of the Mongols performed the sound tests with an average of 0.360 seconds, and with a mean variation of 0.113 seconds.

"In order to test the visual memory of our children the author made use of form, color and letters. For the first, ten forms were cut from card board, the more common geometrical forms as well as some irregular ones being chosen. Two sets were made, one for the child and one for the experimenter. For the color tests, ten colors were selected from Bradley's colored papers, the endeavor being made to get them as unlike as possible, and affixed to cards. For the last test paper letters mounted upon cards were used, the consonants being chosen so as to exclude syllable formation. The child being ready for the experiment, a set of forms, colors or letters were arranged before him on the table; five of the same series were then shown to him by the experimenter for two seconds, as indicated by a metronome, then they were covered from view and the child was required to select the same five as he remembered them from the sets of objects before him, the order of selection being disregarded. Five trials of each sort were made, and the number of correct selections noted. The results were as follows:

"Thus out of five objects these were the average number correctly remembered. The ages of the children ranged from fifteen to thirty, and we think that we are not wrong in expecting almost perfect answers to our tests from normal people of the same age. The memory error for five objects of our children is:

by their teacher.

"Grouping the children according to mental ability, as estimated A being the brightest, we have:

"Interpreting the mean variation as the uncertainty of memory and deducing the memory error from the other columns, we see tha the error and uncertainty of memory increase with mental dullness In regard to the influence of the knowledge of the names on th memory of the objects, we find that the average number of name known by each child is:

This shows a slight influence in colors and letters with the girls. "The tests so far having to do with visual memory, we next tool up for consideration the auditory memory. To this end, we mad use of the following tests. First a series of nonsense syllables wer made, endeavoring to get some as free from association as possible These were read to the child at the rate of one per second, fiv making one test. Next were selected groups of six associated words which were likewise read to the child at one per second. He wa required to repeat them immediately, and both the words remem bered and the number of associations were noted.

Lastly, group
These were o

of sentences were selected from their school readers. various lengths from five to thirty-five words. They were read t the child at the rate of a word per second and he was required t repeat it immediately. The number of words correctly remembere was noted. The results were as follows:

"The small number of associations and the large number o words is very noticeable. Chance we consider to be excluded by the method of work of the children.

"Grouping our results according to the mental ability of th children, we have:

"Here, as above, we find the error and uncertainty of memory increase with the mental dullness. These tests in auditory memory seemed easier to perform than those for visual memory, for four girl

and two boys graded C while being able to perform the auditory tests, could not do the visual tests. Grouping the results to show the average number remembered in each way, we have:

"This shows practically no difference. Yet five girls and two boys said that they remembered by 'looks,' and seven girls and nine boys by 'name.' However, we have seen some influence from the knowledge of the names in the visual tests.

"The same children were given a test on the memory of muscular movements. A graduated rod was fixed on suitable supports and two adjustable stops or rings were placed on the rod so that any distance could be set off by them. Three distances were used; ten, thirty and fifty centimeters. In performing the test, the child was seated so that one of the stops on the rod was in front of him, the other stop was set at the required distance; with eyes shut, the child then moved the index finger of his right hand from one stop to the other and then back. The outer stop was then removed and he was required to move his hand over the original distance as he remembered it. In the first series this was done immediately, then after a wait of ten seconds, then of twenty seconds. The results for 100, 300 and 500 millimeters were as follows:

"The averages found by Scripture for the same distances:

"Comparing these results from normal people with the ones we have given, we find the memory error of the feeble-minded to be from two to ten times the normal, and the uncertainty of memory to be from five to fifteen times the normal. The memory error, as with normal people, increases with increased distance while the uncertainty steadily increases. Grouping the results according to mental ability, we have, for 100, 300 and 500 millimeters: