depths under conditions of natural illumination, reduced by neutral or 

 selective (blue) filters. The degree of attenuation of light 

 corresponds to its attenuation in the sea at the depth at which the 

 samples were taken (samples taken from the surface were exposed to 

 natural illumination). A fourth plan for determination of primary 

 production is the "chlorophyll" method of Ryther and Yentsch, and is 

 essentially the same as the plan of Sorokin. The heterogeneity of the 

 vertical distribution of phytoplankton is considered by determining the 

 concentration of chlorophyll, while the correction for light is 

 introduced using the curve of variation of photosynthesis with light 

 intensity. The relationship of photosynthesis to concentration of 

 chlorophyll with optimal illumination (AN^p^) is assumed constant. 



All of these plans, naturally, are imperfect. Measurements in situ 

 most closely reflect the phenomena which occur in nature, and can~5e 

 used as a standard. 



A distinction can be made between random and systematic errors for 

 the various techniques of measurement. Random errors, based on the 

 results of the 5?nd voyage of the VITYAZ', amounted to about 20% of the 

 total deviations of data, resulting from errors in methods and the 

 natural fluctuations. The relative total deviations increase 

 exponentially with decreasing production (Fig. 1). With a production of 

 0.1 mg C/m"^ per day, the total deviation is 100%; this value may be 

 considered the limit of sensitivity of the method (Koblentz-Mi shke, 

 1Q61). 



AP/P 



/o 



20 



30 



'iO 



.,50 



cO 



mg C/m-^ per day 



Fig. 1. Relative deviations (%) of primary production (P) at the 

 surface, based on data from the 29th voyage of the VITYAZ' (1), the 46th 

 voyage of the HUGH SMITH (2) and the 46th voyage of the OSHORO-MARU (3). 



230 



