96 
gives the only statistics available for this 
purpose, so far as I know. He collected the 
printed programs of schools for the period 
just preceding 1894, when the report of the 
Committee of Ten would not yet be effective, 
and for that ten years later, and compared 
the two. In his own words: 
For the earlier portion of the study 80 schools 
were covered: 35 in the eastern section of the 
country, 25 in the middle west and 10 each in the 
south and far west. For the period a decade later 
the number of schools was 160: 49 being in the 
east, 46 in the middle west, 30 in the south and 
35 in the far west. 
Neither these numbers nor the particular schools 
studied were the result of arbitrary choice, but in 
most cases of dire necessity. Every available 
course of study for the years 1892 to 1894 was 
considered, and this was essentially true for the 
period ten years later. So far as possible, the 
same schools were considered at both periods; but, 
as indicated by the figures, many more schools 
were included in the later than in the earlier study. 
This was that errors due to accidental conditions 
might be reduced to a minimum. I have not 
thought it necessary in this paper to give the 
names of the particular schools studied, but will 
say that the list includes the high schools of nearly 
all the larger cities of the country; and that none 
of the smallest schools are covered is suggested by 
the fact that only those issuing printed courses of 
study are included. The part of the study covered 
by this paper has to do only with those recom- 
mendations of the special subcommittees (of the 
Committee of Ten) 
school curriculum. 
which bear upon the high- 
The second factor, the probability that the 
student will have the opportunity to take a 
study in the year in which he happens to be, 
can in most cases be computed from Dexter’s 
data, in an approximate sort of way. I will 
give the computation for German in some de- 
tail, as in it appear all the irregularities 
which show themselves in connection with 
other subjects, and further, the resulting 
table contains the only essential absurdity 
which developed in the preliminary computa- 
tions. 
First, Dexter’s table for German. 
SCIENCE 
[N.S. Vou. XX XV. No. 890 
TABLE I 
1894 1904 
Percentage of schools offering 2 years 34 25 
Percentage of schools offering 3 years 33 36 
Percentage of schools offering 4 years 33 23 
Percentage beginning in the first high- 
school year or earlier ............ 48 47 
Percentage beginning in the second 
high-school year’ ..:.............. 30 41 
Percentage beginning in the third 
ihioh-schoolnyeataurene tre tteitettelelst ters 22 12 
(I infer that 16 per cent. gave a 1-year 
course in 1904, beginning in year III.) 
This table I rearrange and extend as fol- 
lows: 
TABLE II 
1894 1904 
Begin eon Length Hae Begin ae Length Gok 
33 f 33 23 4 23 
I 48 I. 47 
15 24 
qe 1) SS 72 2 36 
II 20 II. 41 | 
12 29 
SH ey oe “4 2 28 
Tie 22 Ik 12 
0 16 
ine. @ Dy oh” ity NOM aime 
The percentages opposite the half-braces 
({) mean thus: For 1904 23 per cent. of the 
schools gave a 4-year course, while 47 per 
cent. began in year I.; hence 24 per cent. must 
begin a 8-year course in year I. Thirty-six 
per cent. gave a 8-year course, hence 12 per 
cent. begin the 3-year course in year IL, ete. 
That this course of reasoning is imperfect 
appears from the fact that according to it 
—4 per cent. begin a 2-year course in year 
T11., which is absurd. However, the difficulty 
lies in the original data being out of reach, 
and as the absurdity is not going to be of 
great influence on the computations, as it 
comes in the third and fourth years, I use the 
figures as they stand. The table for German 
is the only one in which any patent absurdity 
shows itself. 
From this table I obtain the percentages 
which express the random student’s oppor- 
tunity to take German, 7. e., the per cent. of 
students who would be taking German were 
it required wherever and whenever it is 
offered, thus: 
