278 Trans. Acad. Sci. of St. Louis. 
tion from the type is easily calculated; for example in the 
series of Table No. 1, the chances are even that any deviation 
will fall between +1.0 d and—1.0 d, for it has been seen that 
half the whole number of observations fall within these limits. 
Quetelet demonstrated that the method of grouping related 
measurements could be used not only for showing the most 
probable height of one individual but also for showing the 
most probable or typical height, i. e., that most often found, of 
a number of individuals. The Belgian Astronomer Royal 
pointed out that the series obtained by measuring a number 
of individuals of the same type, e. g., men of the same nation- 
ality, was characterized by the peculiarities that distinguish 
a series in which the deviations from the typical or most 
probable value are due to accidental causes. The truth of 
this statement is obvious when the two are compared. 
TABLE No. 2.* 
HeiGuTs OF UNITED States RECRUITS. 
Height at Intervals of One Inch. Number of Recruits. 
Between 78 and 79 inches - 
‘“ 7 “ —— 6 
“ 6 ts Foo 9 
és 15S ae 42 
“c Th 8 ee 118 
és 7g ke bee 343 
“ y 6 eee: 680 
ry; “ ye 1485 
6s ya “ 2075 
“6 69 «6 pe 3133 
‘ 6s << y ee 3631 
“ 67 « 68. 4054 
‘“ 66 * 67 «6 3475 
ts cs. « 6a « 3019 
Pt} C4 tk 1947 
“ 6B ¢ ‘“ 1237 
“a 62 * 63 «6 526 
“ r a ‘“ 50 
“ 60 “* 61 « 15 
“ BO © 60 ti 10 
“ 58 * 59 6 6 
“ 67 5a 8 7 
i. 2a a ee ee 3 
rd 55 BBC 1 
“55 and less 4 
Total. ..25878 
* E. B. Elliott’s table in Medical Statistics of the Provost-Marshal-Gen- 
eral’s Bureau, by J. H. Baxter, 1875, Vol. I. Introduction, page xxx. 
