43 
of Edinburgh, Session 1884 - 85 . 
particles; while, when four or five determinations were made on 
water contained in a bottle, each quarter litre would contain more 
sediment than the one before, and a minute shell makes a distinct 
increase in the alkalinity. Looking at the eight cases epitomised 
Table III. — Alkalinity Determinations. 
Ho. 
Sample. 
I. 
II. 
III. 
IY. 
Y. 
Mean. 
a 
68 
11‘75 
12-0 
12*1 
11-95 
b 
69 
11*5 
11 *1 
10-25 
11-5 
11-09 
c 
70 
14-05 
13-9 
14*65 
14-21 
d 
71 
11*9 
12*15 
11-75 
11*93 
e 
72 
11*0 
11-55 
11*9 
11-48 
f 
73 
11*65 
11-55 
11-61 
9 
74 
12*1 
11*75 
11*92 
h 
92 
12*8 
12-4 
11*7 
12 "4 
12*31 
i 
93 
12‘2 
11-8 
12-01 
j 
261 
9-2 
10-65 
9-9 
10'65 
9*8 
10-04 
k 
262 
10-05 
10-1 
9-85 
11*0 
10-25 
Table IY. — Deviations from Means . 
a. 
b. 
c. 
d. 
e. 
/• 
9‘ 
h. 
i. 
J- 
k. 
I. 
*2 
*4 
*15 
•04 
*48 
•05 
•18 
•5 
•2 
•84 
•2 
II. 
*05 
•o 
*3 
*22 
*07 
•05 
•18 
•1 
•2 
•61 
•15 
III. 
*15 
*8 
•45 
*18 
*41 
•5 
•14 
•4 
IY. 
*4 
•1 
•61 
•75 
Y. 
•15 
Mean deviation 0 ’29. 
in the preceding table (III.), where three or more determinations 
were made on one water, it is evident that in four (a, c, e, and &), 
the later determinations show an increased alkalinity. 
The quantity represented by an alkalinity is very small, although 
the number used to express it is large. An alkalinity of 50, 
means that in a litre, say 1026 grammes, there is 0 05 gramme of 
carbonic acid as calcium carbonate ; that is, a percentage of 0‘00487 ; 
which, taking the worst view of the inaccuracy of the determina- 
tions, might vary from 0*00498 to 0*00476. 
In the discussion of the alkalinity of the Firth of Forth in this 
paper, the integral figures may be viewed as correct, the decimal 
part being uncertain. 
It would impart a greatly increased certainty to the figures if 
the water had always been filtered before the alkalinity was 
determined, and if duplicate analyses were invariably made and 
