three days later in the laboratory. Ammonium molybdate and 

 sulphuric acid was then added to the sample, converting the 

 silicate to a yellow-colored compound which was compared 

 visually with standards of known silicate concentration. The 

 results of the silicate-silicon determinations were not 

 corrected for salt error. 



Phosphate -phosphcrus was determined by the Robinson and 

 Thompson (1948b) modification of the method of Deniges. A 

 solution of sulphuric acid and ammonium molybdate was added 

 to the sea water sample, forming a complex phosphomolybdic 

 acid. Stannous chloride was then added to reduce the complex 

 acid to a blue substance whose color is proportional to the 

 concentration of phosphate. The water samples were treated 

 aboard ship just after removal from the Nansen bottles and a 

 colorimetric comparison was made in Nessler tubes. 



The nitrate-nitrogen concentration was determined colori- 

 metrically by comparing the color produced in sea water with 

 a reduced strychnine reagent against a previously calibrated 

 standard of methyl orange in N/'lOO HCL (Wattenberg, 1937). 

 The methyl orange solution is standardized by comparison with 

 known concentrations of nitrate in sea water. To avoid the 

 salt effect, nitrate-free sea water was used. As soon as 

 possible after collection aboard ship, the strychnine reagent 

 was added to the sample of sea water, and 24 hours later the 

 sample was visually compared with the methyl orange standard. 

 Both nitrite and nitrate are measured by this method; however, 

 the former is in small enough quantities in ordinary surface 

 sea water to fall within the experimental error of the method. 



