388 On the Frequency Distributions of Phagocytic Counts 
with the other examples. It is specially to be borne in mind that the change of 
type occurs at the highest mean value of the series. The interest of this will be 
clear when we have examined the distributions of Dr Fleming's counts. 
Just as before, we find the distributions for Dr Fleming's counts are excellent 
examples of skew variation, and, with one exception (Norm. S.A.), the fits are 
good, the values of P being but little inferior to those obtained for the S. counts. 
The exception (N.S.A.) is, we think, more apparent than real. The graph is 
a very fair fit, and one group has been mainly influential on P, adding more than 
10 to There appears to be definite evidence of anomaly in the group of cells 
with 7 bacilli. 
The mean of this count was high, and the sensible frequencies extended to 
9 and 10 bacilli per cell. It is notoriously diflScult accurately to determine the 
TABLE V. 
Equations to Curves. 
Strangeways : 
I 
II S. C. 
II c.&s. 
Ill 
IV 
V 
VI 
VII 
IV 2000 
Fleming : 
T. C. 
10 Norm. 
N.S.A. 
N.S.B. 
No. 2. 
T.A. 
^/ = 299'93816 
?/ = 348-0489 
?/ = 315-8.3835 
2/ = antilog 16 
v/ = 3.35-8664 
3/ = 258-7307 
?/ = 278-51616 
2/ = 284-2268 
?/ = 643-9147 
?/ = 264-7097 
?/ = 237 -05324 
«/ = 173-91942 
1 + 
5715 
X 
12-7037 
X 
3466 
r^, -| -7283 r 
^"^-91 76 J [^"ST-^ 
''"-9692 J |_ ] 62-257 1_ 
3831727 X .t— 12-7635 X e-67-8953/a:_ 
-7902J L 33-3574J 
L 30-1 S 
Graph No. 
1 
1 + 
.J. -|1.0!)37 
r6084j 
^. -11.2202 
-| 20.4840 
228j 
X 
-1 
1-6537 J \} 
■*'r0594j L 14-1413J 
l+^^^l II 
18-6085J 
^ -17 -3627 
I4T4I;; 
X 
■ 10397 J L 80-801 2 J 
^0578j L 19-2729J 
-| 2-8984 r 
u L 
2-9510 
rg -|48-1737 
49 -0480 J 
+ 
-12-59 
2-9578J 
^, -121.9550 
24^981 J 
y = antilog 29-9855876 x - ^i-^i^s e - i55-3i38/x 
?/=antilog 22-603153 x.:^- 17-6050 ^ e-oi-i68i/a:. 
3/ = antilog 27-0100681 x:j7-i8-2'»69 gi6i.ii34/x 
10 
11 
12 
15 
13 
14 
