216 
Miscellanea 
The same value is obtained by the product moment method, the difference between this and 
the r="613 obtained by De Helguero being due to the sUp in his calculation of the standard 
deviation. It appears from the above that the calculation of r is a very simple process. 
TABLE I. 
Inflorescences of Cicorium. 
8 
9 
10 
11 
13 
1 
13 
1 
14 
15 
26' 
17 
Totals 
8 
9 
10 
11 
12 
13 
n 
15 
16 
17 
1 
1 
1 
2 
1 
2 
2 
15 
10 
1 
2 
1 
2 
8 
19 
11 
4 
1 
1 
15 
19 
134 
90 
18 
1 
2 
10 
11 
90 
114 
97 
9 
4 
1 
4 
18 
97 
122 
53 
2 
2 
1 
1 
9 
53 
96 
10 
4 
2 
10 
10 
3 
3 
2 
1 
1 
5 
33 
46 
278 
337 
297 
172 
29 
6 
1 
Totals 
5 
33 
46 
278 
337 
297 
172 
6 
1 
1204 
V = Difference of x and y. 
0 
1 
3 
5 
6 
7 
8 
9 
0 
1 
1 
1 
2 
0 
0 
0 
0 
0 
2 
2 
15 
10 
1 
2 
0 
0 
0 
8 
19 
11 
4 
1 
0 
0 
0 
0 
134 
90 
18 
1 
0 
0 
0 
0 
114 
97 
9 
4 
0 
0 
0 
122 
53 
2 
0 
0 
0 
276 
96 
10 
0 
0 
2 
224 
10 
3 
0 
4 
180 
2 
1 
20 
64 
0 
56 
50 
488 
276 
794 
iV^=1204, 
o-^ =1-3352* 0-^2=1-7828, 
o-,2 = ?i^ = 1.318937, 
(^ 1 '^A [^ 1 1-318937 \ 
In working with a symmetrical table we copy down only the diagonal cell and the positive 
differences. But it would be very easy to continue right down the columns : after passing 
the diagonal cell the differences would take the negative sign and we could calculate the 
standard deviation from all the differences instead of from a part of them. In the symmetrical 
table this would be merely a waste of time, but where the tables are not symmetrical (and the 
two variates have different means and standard deviations) this process can be used to advantage. 
* The value 1-3536 given by De Helguero was apparently obtained by taking the root of the second 
rough moment, i.e. without first subtracting the square of the first rough moment. 
