1922] Allen: Studies on Marine PhytoplanMon at La Jolla Ml 



Although all other production was very inconspicuous as com- 

 pared with the September maximum there was a pulse covering 

 two days, November 17 and 18, at one time, and another covering 

 six days, December 17-23, at another time. Since there was little 

 current recorded at the November pulse its abrupt increase of some 

 fifteen-fold may simply indicate the presence of a swarm stranded 

 in quiet waters. While the December pulse at its highest showed 

 about the same increase in production, its development was much 

 less abrupt and so was its disappearance. In this ease also the 

 currents recorded- were weak and the difference in development as 

 compared with the November pulse might be considered due to 

 relatively rapid production induced by some favorable local factor or 

 favorable change in the factorial complex. 



Forty-six forms of diatoms are listed as occurring within the four 

 months in sufficient numbers to deserve separate mention. Of these 

 only twelve had specific characters sufficiently marked to make 

 identification easy. Fourteen could sometimes be satisfactorily 

 identified and usually approximated, i.e., confusion was limited to 

 only one or two other forms. Of the other twenty, some were always 

 doubtful and others were simply grouped under the genus name 

 because of small numbers or great uncertainty of identification. 



The Chaetoceras genus offers most constant difficulties of this sort 

 on account of its constant representation and on account of the fact 

 that many species characters are based on conditions at sporulation 

 or on terminal setae which may be missing. It is also true that many 

 Chaetoceras chains are more or less twisted in form and that they 

 lodge in all sorts of positions in the counting cell. Identification 

 might be possible with unlimited time and effort, but it is not 

 practicable to insure accuracy under counting conditions. Some 

 Chaetoceras species could not be identified because of small size, as 

 was also true of diatoms in other genera. In the main, however, the 

 more important forms numerically in this series of catches were at 

 least approximately identified as to species and almost all were 

 correctly placed as to genus. 



By selecting in each month the five species which showed the 

 highest averages it is possible to get a good general idea of the 

 character of the main diatom population throughout the four month 

 period. Eleven forms are recorded in the highest five in one or more 

 months. Chaetoceras sp. appears in this list in all four months. 

 Chaetoceras curvisetum Cleve in three months, Chaetoceras flcbile 



