i8o DISCOVERY REPORTS 



A study of the predominant genera composing the catches was made at the time 

 of their collection by Mr G. W. Rayner to whom I am indebted for Fig. 53 and the 

 data in Table XII and Appendix I. They show that although a cosmopolitan genus 

 like Chaetoceros was represented almost universally, many of the other genera were 

 abundant only in certain regions. Thus we may distinguish between oceanic species 

 like Rhizosolenia and Planktoniella, and neritic species : and we may distinguish between 

 species relatively more abundant in the north such as of Coscinodiscus and of Thalas- 

 siosira and those in the south, Synedra and Coretfiroti. To what extent these distribu- 

 tions change from season to season we have no knowledge. 



PHOSPHATE CONTENT 



Analyses of the phosphate content of the sea water were made on nine of the lines. 

 In illustrating the upwelling of inshore waters the results fall into line with temperature 

 and salinity records. The general characteristics of phosphate distribution in the oceans 

 are a high concentration in deep water and a low concentration at the surface, where in 

 well-lit layers inorganic salts may be consumed during photosynthetic activity of the 

 plankton. With this distribution our results agree except in those areas close to the shore 

 where upwelling of water rich in salts has taken place, and here phosphates are in high 

 concentration at the very surface. 



The phosphate analyses were carried out by Mr A. H. Laurie who has plotted the results 

 in Figs. 40 and 54-61 as the number of milligrams of PoOj per cubic metre of sea water.^ 

 Each figure illustrates the phosphate values in a vertical section running transversely 

 across the current. The sections may be arranged serially to illustrate a gradual change 

 from the condition at Pichidanque Bay, where surface phosphates are everywhere at 

 a minimum, to that at San Juan where rich phosphates occur at the surface for as far as 

 50 miles offshore. All these sections betray some trace of upwelling close inshore by 

 the direction of the boundary between the surface waters, where depletion of phosphate 

 has occurred, and the rich stores of phosphate in the deeper water ; this boundary line 

 is usually a layer with medium phosphate content which runs horizontally at about 

 50-200 m. but rises towards the surface near the shore. 



Unevenness in phosphate distribution probably depends upon three cardinal factors: 

 upwelling or subsidence of rich concentrations to and from the surface according to 

 wind, seiche or other conditions ; depletion brought about by propagation of phytoplank- 

 ton ; and vertical mixing of the richer and poorer waters through the action of gales and 

 currents. To these must be added the decomposition of organic remains and salts of 

 terrestrial origin. The common plan underlying phosphate distribution in the first 

 seven of the nine sections just described points to upwelling phenomena as the most 

 influential of these factors, and consequently a correlation is to be expected between the 

 phosphate distributions and the temperature curves. 



^ In this work it was found impracticable to filter the samples before analysis. The data were not 

 salt corrected and presumably include arsenates if present: they are therefore probably on the high side. 



