A Chloride and Oxygen Analysis Kit for Pond Waters 1 
R. B. Dean and R. L. Hawley 2 
INTRODUCTION 
Fishponds in the Hawaiian Islands are 
shallow brackish ponds which are frequently 
fed by underwater springs as well as surface 
flows of fresh water and sea water. The 
chloride content of the water is an indication 
of the quantity of sea water present and is 
one factor which affects the growth and sur¬ 
vival of many of the plants and animals 
utilized as food by the fish. The presence of 
dissolved oxygen is absolutely essential for 
animal life in the ponds. Low oxygen con¬ 
centrations in some regions indicate stagnant 
waters and bottom putrefaction which make 
these areas unfit for fish. In the course of 
fisheries studies in the Hawaiian Islands it 
was therefore necessary. to make a large 
number of salinity and oxygen determina¬ 
tions on fishpond waters. 
Since both chloride and oxygen may vary 
severalfold in a single pond, and since most 
living organisms are insensitive to concentra¬ 
tion changes which are less than 1 per cent 
of the normal value, it is not necessary 
to make extremely accurate determinations. 
Methods were desired for the rapid analysis 
of chloride and oxygen at the side of the 
pond so that any unusual observations could 
be checked before one left the pond. 
The apparatus described in this paper was 
designed to provide sufficient equipment and 
reagents for 25 chloride, and 50 dissolved 
1 Research Paper 5, Cooperative Fisheries Re¬ 
search Staff of the Territorial Board of Agriculture 
and Forestry and the University of Hawaii. 
2 Department of Chemistry, University of Ha¬ 
waii. Dr. Dean is now at the University of Ore¬ 
gon; Mr. Hawley is a student at George Washing¬ 
ton University Medical School. Manuscript re¬ 
ceived December 2, 1946. 
oxygen determinations. Two auxiliary half¬ 
liter bottles contain enough reagents for 
an additional 50 chloride determinations. 
Chloride up to 25 gm. per liter (sea water 
is about 20 gm. per liter) can be determined 
with an accuracy of 0.06 gm. of Cl per liter. 
Dissolved oxygen up to 20 cc. per liter can 
be determined with an accuracy of 0.1 cc. 
per liter on a sample of 10 cc. of water. The 
apparatus weighs about 25 lb. and can be 
used in a rowboat, if necessary. A single 
chloride determination requires about 3 
minutes; an oxygen determination requires 
about 8 minutes. 
Although there are several well-known 
titrimetric methods for chlorides, none of 
them is suitable for use on micro-samples 
out-of-doors where it is difficult to observe 
a colorimetric end point. An electrometric 
method of detecting the end point which 
substituted a galvanometer needle for the 
color change appeared to be more suitable 
for our purposes. None of the published 
electrometric methods was quite satisfactory 
and an entirely new method was developed 
which will be dealt with in another publi¬ 
cation (Dean and Hawley: unpublished). 
This method makes use of two wire elec¬ 
trodes, a galvanometer, a few radio re¬ 
sistances, and a dry cell. A relatively low 
resistance circuit is formed which is not 
sensitive to high humidities, although it 
should not be soaked in water. 
Krogh (1935: 131-133) has described a 
modification of the Winkler method for dis¬ 
solved oxygen which was well suited to our 
purpose, with slight modifications. Krogh 
used the conventional starch indicator to de¬ 
tect the end point. We have substituted 
108 
