MEASURING BY LIMITED IMPRESSIONS 323 freshly in mind, would encourage a deliberate instead of a hasty expression, would serve as a check against emotional prejudices, and would overcome the serious lapses to which man’s memory is subject. Too much can not be said about the unreliability of the faculty of memory and the great danger of depending for our opinions of people upon so uncertain a quantity. A periodic written estimate would do much to reduce the sporadic character of so many of our impressions regarding people. However, the impressionistic method can be still further improved upon by arbitrarily limiting the range over which our impressions shall be spread. Without some arbitrary limitation, the qualities with which a worker may be credited or debited will be limited only by the descriptive vocabulary of the person making the estimate. The result will be a collection of biographies, too clumsy and involved to be of any value in comparing people. Therefore, although men may be capable of as many virtues as the human tongue can find names for, practical considerations make it necessary to concentrate upon a few of the more important qualities. Only by means of such a limitation will it become possible to find a uniform basis upon which to compare people. The selection of the few most important qualities presents ­ a most complex problem. To begin with, we have no concrete manifestation or summation of qualities such as is offered by the comparative productiveness of the piece-worker. Probably the only mathematical or concrete ­ summation of qualities which we have is that signified ­ by the attendance record. The factor of attendance, therefore, presents itself as the most definite and obvious one to be selected for this purpose. All other qualities must be selected after a more or less arbitrary method.