SEA-FISHERIES LABORATORY. 14] 
Then we plot the summed frequencies, as in the following 
gure -— 
tle 
“3 
if ~ ce | 
JO | “bp - Rane £ ey 
= 
>= ‘ 
é/3- Range $1 Ty 
7a 
ee... ‘ 
45 — Range 
q a 
O ; G! | SS fi 
Si Ft@) 75 2O ‘ee 5 a O Spey 4-O 
The point 1 being plotted against 41 on the horizontal 
axis, 2 against 36, 2 against 31, 4 against 26, and so on. A 
smooth curve is then drawn as evenly as possible and as near 
as possible to all these + points. This is an approximation 
to the integral curve, that is, if f(x) be the function representing 
the frequencies, / f(x)dz is the function representing the 
summed frequencies, and it is given approximately by the 
summational curve. 
Now it is easy to see, from the construction of the figure, 
that any point on this curve gives us the sum of the frequencies 
from the right-hand limit up to that point, and so we can 
interpolate graphically for any desired range of counts of 
colonies. The mode, or greatest frequency, is at the extreme 
left of the curve. Now the range of the vertical scale (No. of 
colonies on a plate) is 20, and we take 4, %, }, #, etc., of this 
and measure off these distances from above downwards, thus 
