526 



CYCLES OF ORGANIC AxND INORGANIC SUBSTANCES 



3 4 S 



Phoiphori.^ (/U9at/i) 



Plan* (.gC/m') 



Fig. 2. Theoretically derixed relations between phosphate-phosphorus, plants, 

 and zooplankton. 



from discussion with Mr. Holmes of the Scripps Institution it 

 seems possible that similar spatial changes, in the plant-phosphate 

 picture in particular, occur outwards from an upwelling or "dome" 

 region in the tropical Pacific. 



Thus the commoner methods of statistical analysis, although 

 they are useful in indicating, initially, factors which may be af- 

 fecting production, are not so useful in more detailed work. This 

 is because in, say, a multiple correlation, the variables representing 

 the difTerent ecological factors must be capable of transformation 

 into linear (or additive) relations. But the various mathematical 

 forms for the causal relations, although they are rather dubious, 

 have one common property — the relations are always nonlinear 

 owing to the complexity of the interactions in the biological sys- 



