STREAM POLLUTION 
407 
in the actual case of pollution under consideration and the same chemical interre- 
actions were possible. When it was not feasible to use stream water at the site of the 
pollution, unpolluted water from a comparable stream was diluted or fortified until 
the salt content, the conductivity, the carbonates, and the pH were essentially the 
same as those of the stream water in question. This was made possible by the field 
and laboratory analyses of the waters of the stream. This method of compounding 
a synthetic stream water, although laborious at times, gave very satisfactory com- 
parative results. Control tests were also made in each investigation using glass- 
distilled water as the diluent for certain critical concentrations. 
In making the assays of pollutants, regardless of tbe diluent water, no material 
was removed after making the dilution; i. e., even if precipitates or colloidal suspen- 
sions were formed on the addition of the pollutant to the diluent water, the fishes or 
daphnia were exposed to the action of the entire mixture. Many substances which 
are precipitated in stream and river water are nevertheless slightly soluble in water, 
particularly if small quantities of carbon dioxide and other compounds be present, so 
that portions of the precipitate might continue to redissolve for hours or days, and 
insoluble substances even though precipitated to the bottom of the stream constitute 
supplies of various ions which may be drawn into the water as others are removed. 
Again, some of these precipitated materials are dissolved or chemically changed by 
the action of various substances in or on the body of the fish or daphnia, or by the 
excretory products of these animals, so that particles of precipitated material lodging 
temporarily on the gills might be dissolved by the carbon dioxide there, or other 
particles swallowed by the fish might be acted upon by the digestive fluids of the 
animal. In Hew of these complicating factors, the tests made in stream waters are 
much more valuable in obtaining the actual hazards to aquatic life than those made 
in distilled water. 
In the basic data tests (see tables 7, 8, 9, 10, 11, 12, and 13) four types of water 
were used as diluents, namely, glass-distilled water for the chemical controls in 
which the specific and uncomplicated action of the substance was studied, and soft, 
medium, and hard water (following the classification of Birge and Juday, 1911), i. e., 
with fixed carbonates less than 5 cubic centimeters per liter, 5 to 22 cubic centimeters 
per liter and over 22 cubic centimeters per liter, respectively, for the general lethality 
tests. Ranges of lethality were obtained in this way which are applicable within the 
limitations previously discussed to most stream waters of the United States, since 
the hardness of these natural stream waters is dependent almost entirely on the cal- 
cium and magnesium carbonates which exists in different degrees ol concentrations 
in the interior waters. 
The soft and hard natural waters were obtained at Columbia, Mo.; the hard 
water representing a typical limestone drainage, and the soft water a typical clay 
drainage. The medium water was obtained from the upper Mississippi River. In 
the tables the pH and specific conductance are given both for the original water and 
for the mixture as modified by the pollutant, so that the fluctuations in tbe diluent 
water itself and in the mixture as the result of the action of the pollutant on the salt 
and buffer balances may be followed in each case. 
