4 o DISCOVERY REPORTS 



The relative importance of the four groups of the phytoplankton met with in our 

 survey may be judged from the comparison of the totals of their estimated numbers in 

 our samples : 



Diatomales 1,846,000,000 



Protococcoideae (Halosphaera viridis) ... ... 84,000,000 



Silicoflagellata (Distephanns spp.) 420,000 



Dinoflagellata (Peridinium, Dinophysis and Ceratium) 283,000 



It is seen that the Dinoflagellates form but o-ooi per cent of the total phytoplankton. 

 Before passing to a general discussion of the phytoplankton as a whole and its pro- 

 duction in relation to hydrology, we will give an account of the occurrence of the 

 different species. The estimated numbers of the commoner forms are shown in 

 Appendix I, and those of the less common and rarer forms are recorded in the text. 

 At the top of each column in the Appendix is also shown the fraction of the sample 

 examined, so that the actual number of each organism observed, on which the 

 estimated total is based, can be seen. In the text the figure in brackets following the 

 estimated number shows the actual number observed. In many of the instances where 

 rarer forms are recorded it will be noted that the figures representing the estimated totals 

 have only a most hypothetical value and can mean no more than an indication of their 

 presence in small numbers in the sample. Similarly, many samples must contain small 

 numbers of rarer forms which have not been recorded because they have not been taken 

 in the fraction of the sample examined. The commoner species often occurred in such 

 large numbers that only a very small fraction of the sample could be examined. 



The significance of the Phytoplankton numbers 



Throughout the survey we are comparing the quantities of phytoplankton at different 

 stations as measured by the 50 cm. diameter net hauled from 100 m. to the surface, and 

 in the report, unless otherwise stated, the relative quantities of the different organisms 

 are expressed as numbers per haul of 100 m. No attempt has been made to express 

 these quantities as numbers per cubic metre of sea water for two reasons. Firstly, 

 because it is felt that estimates per volume of water by a calculation of the filtration of 

 the net must be widely out owing to the clogging of the meshes of the net in the heavier 

 hauls, and, secondly, because the diatoms are so much more numerous in the upper part 

 of the column of water sampled as compared with the lower part that an average figure 

 of diatoms per cubic metre for the whole column would have quite an artificial meaning. 



The column of water which would pass through the ring of the net if it was hauled 

 vertically from 100 m. to the surface without the net being attached would be approxi- 

 mately 20 cubic metres. Thus we can say that the average number of diatoms per cubic 

 metre in the sea per column sampled is not less than one-twentieth of the number given 

 in the sample. Actually it would be very much more, because not all the water in the 

 column will pass through the net, and the amount it will be more will vary according to 

 the number of diatoms reducing the filtration. We are concerned in this survey, as 

 already explained on p. 27, with only the really large differences in the numbers of 



