as follows: 
THE STATISTICAL STUDY OF VARIATION 33 
individual beans varied considerably in size. Furthermore if one were 
to obtain the average weight of the whole million, it would not differ, 
essentially, from the average weights of the smaller groups. 
The prin- 
ciple involved here may be stated in various ways. 
Weld expresses it 
M gh 102 
99 
ie et 
B65 
gee 
GoW SE OOSCEn GER OO OGOL OL PEHECCBOONE, GOL LOCOO. OF PHOECE ESC Rae 
36 
COURS Ce"OSS O30 LG093R1ENG SEI BAT 
Opec HEOLese ceevesEsed HOeCorec GeOceussed: C@GCCEO! 
8 
8 8 
3 8 
es , 
3 F 
$ a | 
» x | 
L g : 
2 x i 
¢ ‘> 4 4 A 
of 2 3 = 343 7 8 GB 
pome 2S 2 218 8 BS 8 BS! 
(Ome 3) ISS (GST SAG ie eO) vel etet 
Fic. 14.—Frequency distribution of 500 Broad Beans arranged in classes according to 
width. 
“Tf a number of different events are equally possible as regards 
constant conditions (that is, if there is no persistent reason why one should 
occur rather than another), and all are repeatedly given opportunity to 
occur, they will in the long run occur with equal average frequency.” While 
this is a satisfactory general statement of the law of probability, the same 
3 
Digitized by Microsoft® 
