April 28, 1911] 



SCIENCE 



665 



each other. Any object then throws two shadows, 

 one from each window. The one shadow coming 

 from the colorless window ought to be simply 

 darh, but appears, if the other window contains, 

 bay, red glass, of a saturated red color because at 

 this shaded spot no white light is mixed with the 

 red light. Everywhere else this mixed illumina- 

 tion takes place and makes the observer overlook 

 the fact that the whole room with everything in 

 it appears actually reddish. The other shadow, 

 coming from the red window, might be expected 

 to appear simply white, since this shaded spot is 

 illuminated only by white light. As a matter of 

 fact, it appears, by induction, green because the 

 whole room excepting this spot stimulates on our 

 retina the red process. The area on the retina 

 corresponding to this shaded spot is therefore the 

 only one where the disturbed equilibrium of the 

 components of the red-green substance can be to 

 some extent restored by spontaneous action of the 

 sense organ. (Words in italics are blanks.) 

 The maximum number of points which any 

 student could make in the quiz shown is fif- 

 teen. In case the word inserted was not quite 

 correct, but indicative of the right approach 

 to the solution of the problem, one half point 

 was given. A class of twenty-six students 

 made the number of points shown in the table 

 below. Students having the same number of 

 points were given precedence according to the 

 order in which they had been able to hand 

 over their quizzes to the instructor. Banking 

 these papers took about thirty minutes, and 

 having completed the task the instructor felt 

 sure that he had done no appreciable injustice 

 to any of the students. Ranking twenty-six 

 different essays on the same subject, the Her- 

 ing window, would have taken several hours, 

 and the instructor would have finished his 

 task with the unpleasant conviction that his 

 ranking in many cases was arbitrary. 



Points 



Points 



1 15 10 12i 19 



2 15 11 12J 20 



3 14J 12 12 21 



4 14i 13 Hi 22 



5 14 14 10 23 



6 134 15 81 24 



7 13 16 7 25 



8 13 17 6 26 



9 121 l6 6 



I do not record the number of points made 

 in any quiz by any student, but only his rank, 

 placing those who " cut " at the bottom of the 

 class. I find that this is an excellent way of 

 insuring class attendance. At the end of the 

 semester I sum up the rank numbers of each 

 student and, by the smallness of the total, 

 determine the rank for the whole semester. 

 By recording in each quiz, not the points 

 made, but only the rank, we overcome the diffi- 

 culties in grading caused by the fact that 

 examinations are sometimes too easy, some- 

 times too hard. In either case the best stu- 

 dents come out on top, the poor students at 

 the bottom, especially when we take into ac- 

 count, as mentioned above, the time within 

 which each student completed his task. 



Let us take up another problem, which, it 

 seems to me, is dependent on our system of 

 grading, but which is by no means a part of 

 this or any system of grading. I believe that 

 Professor Cattell was the first to point out the 

 justice of giving more credit towards gradua- 

 tion to those students who ranked high in 

 scholarship than to those who ranked low. 

 Now, the first and absolute condition of giving 

 varying credit is a method of controlling the 

 uniformity of grading in the institution. 

 Without this the injustice done by diversity 

 of grading on the part of different instructors 

 is only multiplied. In the University of 

 Missouri we credit an E with 30 per cent, 

 plus, an S with 15 per cent, plus, an M with 

 the normal credit, an I with 20 per cent, 

 minus and an F with nothing. It must be 

 understood that the choice of these quantities 

 is for the most part experimental. We are 

 gathering data to put our credits on a more 

 strictly scientific basis. 



I have heard of only one real argument 

 questioning the justice of varying credit. 

 Some believe that by letting a student of 

 excellent scholarship graduate after having 

 taken a somewhat smaller variety of courses 

 than the average student, we give the degree 

 to one who does not have the breadth of 

 training which a college graduate ought to 



