8. Compare to desired units by com- 

 paring sample results with those obtained by 

 concurrently analyzing a set of standards. 



Reagents: 



a. Alpha-naphthylcimine . Boil 

 0.2 gm. of alpha-naphthyl amine in 40- 

 milliliter distilled water. Separate 

 the clear portion from the dark- 

 colored precipitate and then add the 

 clear portion to 15 ml. of glacial 

 acetic acid. Add 285 ml. of distilled 

 water and stir. 



b. Sulfanilic acid. Add 1 gra. 

 of sulfanilic acid to 30 ml. of a 

 50-50 mixture of glacial acetic acid 

 and distilled water. Stir thoroughly, 

 gradually adding more water until 270 

 ml. in all have been added. During 

 this addition, gently warm until 

 solution is complete. 



Salinity 



We used the Knudsen (1901) method of 

 salinity determination incorporating a 

 slight modification. If coastal and river 

 waters (which require an accuracy of only 

 0.2 to 0.3 parts per thousand) were being 

 cinalyzed, the Knudsen burette was replaced 

 with a 50-milliliter self-leveling burette. 

 For the greater accuracy required for off- 

 shore samples, the end point was determined 

 by an electrometric rather than a colori- 

 metric method, and the entire operation con- 

 ducted in a constant temperature room. All 

 final determinations requiring an accuracy 

 of at least 0.1 "/„„ were based on the aver- 

 age of duplicate titrations. Also, the cal- 

 ibration for off-shore samples was achieved 

 by titrating a Copenhagen standard sea water 

 sample every sixth time. We recommend that 

 the effects of systematic errors be mini- 

 mized by making each titration with a sepa- 

 rate set of equipment (including reagent and 

 standard), and, if possible, by different 

 analysts. If only one set of equipment and 

 one analyst are available, make individual 

 titrations on separate days using two sets 

 of reagents and standards. For greatest 

 accuracy, use Copenhagen standard sea water 

 for standardization. Store all samples at 

 the same temperature as the reagents and 

 apparatus for a period of 8 hours before 

 analysis. 



The step-by-step procedure used for 



the analysis of coastal and river waters is 

 as follows: 



1. Using a l5-milliliter Knudsen 

 pipette, draw out a 15-mil liliter portion 



of sample and dispense into a 100-milliliter 

 beaker. 



2. Add 1 ml . of 8 percent solution 

 of K2Cr04 in water to the beaker. 



3. Place beaker on the magnetic stir- 

 ring assembly. 



4. Stir at a slow rate of speed 

 while simultaneously adding the silver 

 nitrate reagent from a self-leveling bu- 

 rette. 



5. As the end point approaches, 

 increase the stirring speed. 



6. Record the end-point reading in 

 the record book. 



7. Maintain a record of the strength 

 of the silver nitrate by checking with a 

 standard sea water sample that has been 

 standardized against the Copenhagen standard 

 water. 



8. Convert burette readings to salin- 

 ity units (parts per thousand) by means of 

 the Knudsen tables or a suitable alignment 

 chart (Marvin 1957). 



Silver nitrate reagent: 



Dissolved 705 gm. of AgNO in 19 

 liters of triple-distilled water and 

 then adjust the solution so that 38.75 

 ml. of the solution will be required 

 to titrate 15 ml. of standard sea 

 water (cl = 19.375 'A.). 



Copper 



We have used the method of Hoste, 

 Eeckhout, and Gillis (1953) and also that 

 of Chow and Thompson (1952) for the deter- 

 mination of copper. The former is preferred 

 because the reagent used is more selective 

 for copper at the pH range encountered in 

 river and coastal waters. In addition, the 

 copper complex extract of the first method 

 has less tendency to cloud in such waters. 



The procedure used in carrying out the 

 preferred method follows: 



