H. W. Acton and W. F. Harvey 
283 
table for testing the goodness of fit of theory to observations, Elderton, Biometrika 
Vol. I. p. 155) we find that \ 2 = 7-987 and P = '43, which means that the two 
distributions cannot be considered significantly different (Fig. 2 shows graphically 
the two distributions). From the foregoing we may assume that our obtained 
results conform more or less to theoretical requirements. 
We will now leave the subject of errors of technique and random sampling 
and proceed to consider , the observations themselves. If an analysis is made of 
the total first counts (Table II and Fig. 3), it will be seen that the distribution of 
our cases is a long drawn out one starting at 2,700,000 erythrocytes per c.mm. and 
Fig. 3. Total First Counts. 
Fig. 2. Firm Line. Actual Observations. 
Broken Line. Calculated Values. 
110 
100. 
90. 
80. 
70. 
t*3 
60. 
50. 
40_ 
30_ 
20. 
10_ 
27_ 
26_ 
25. 
24 _ 
23_ 
22_ 
21 _ 
20- 
19- 
18_ 
17_ 
16_ 
1 5 
14_ 
S 
en 
13_ 
12_ 
11 _ 
10_ 
9_ 
8_ 
7_ 
6_ 
5_ 
4_ 
3_ 
2_ 
1_ 
0 
0 1 2 3 4 5 
Cells per Square. 
2-5 30 35 4 0 4 5 5-0 
Erythrocytes 
5-5 6-0 6-5 7 0 7-5 80 8-5 
in Millions. 
36—2 
