JUNE 7, 1912] 
of the developed fields has also been rapid, 
it has probably not been relatively more 
rapid. Hence the mathematical investi- 
gator of to-day can pursue his work with 
the greatest confidence as regards his serv- 
ices to the general uplift both in thought 
and in material betterment of the human 
race. All of his real advances may reason- 
_ably be expected to be enduring elements 
of a structure whose permanence is even 
more assured than that of granite pillars. 
G. A. MiuuER 
UNIVERSITY OF ILLINOIS 
SCIENTIFIC DISTRIBUTION OF GRADES 
AT REED COLLEGE 
A SCIENTIFIC rather than a personal basis 
for awarding grades in courses of study, if 
grades are to be used at all, together with 
definite credit for quality as well as for quan- 
SCIENCE 
887 
tute one vague expression for another. With- 
out scientific definition, any set of symbols is 
inevitably used in personal, variable, erratic 
ways. To award scholarships, degrees and 
other honors, as if an A in one course repre- 
sented the same distinction as an A in another 
course is to administer the curriculum on a 
patently false assumption.” 
Until all school work can be measured by 
scales, made up of units that are equal in a 
defined sense, the best available grading is one 
of relative position in a series. The nearest 
approach to such a scientific basis for award- 
ing college credits appears to be a distribu- 
tion following the normal probability curve, 
skewed to take account of the effect of select- 
ing the student body. 
The Reed College system is shown in the 
illustration. The outer curve, partly dotted, 
is the normal probability curve. The inner 
curve, partly coinciding with the other, shows 
tity of work, seems desirable, especially for 
institutions that are more than theoretically 
devoted to scholarship, and that are willing 
to make what sacrifices such ideals may 
involve. 
The common grades A to H have no defined 
meaning. To call them such and such a per 
cent. of an undefined something is to substi- 
1Tt is impossible, here, to give a detailed ex- 
planation of the ‘‘credit for quality’’ system and 
the scientific distribution of grades adopted by 
Reed College. The underlying principles of both 
are fully explained in chapters 12 and 13 of the 
‘¢Administration of the College Curriculum,’’ 
Houghton Mifflin Company, 1911. ‘‘The Distribu- 
tion of Grades on a Scientific Basis’’ is presented 
in Popular Science Monthly for April, 1911. 
