1922] Allen: Studies on Marine Phytoplankton at La Jolla 335 



than it is to find the neutral or contradictory evidence of the other 

 two months. Even the reverse showing of the dinofiagellates is 

 suggestive of diurnal rhythm in that group, although it is character- 

 istically more erratic in distribution than the diatoms. At any rate 

 the records suggest possibilities which are very interesting and which 

 certainly open up a wide field for investigation. 



In closing this general discussion of the material, I would remark 

 that in the study of these two groups it was soon found that relatively 

 few species showed marked prominence at any time, and that very 

 few species had such prominence either frequently or continuously. 

 Fortunately, most of these species can be satisfactorily identified 

 under ordinary conditions of study. 



A table giving the total numbers of phytoplankton cells per liter 

 in all catches throughout the four months is given on pages 346-347. 



Detailed Discussion 

 DIATOMS 



Throughout most of the year and in most localities diatoms are 

 both numerically and volumetrically the most prominent photo- 

 synthetic organisms in the open sea. The records for this four month 

 period show characteristic conditions in this respect, November being 

 the only month in which dinofiagellates appeared in superior numbers 

 for the month as a whole. 



Except for four catches in early September diatoms were always 

 found to be present, though often in very small numbers. Since it is 

 probable that larger samples would have revealed some in the four 

 catches in which they were found lacking, it may be safely assumed 

 that some diatoms might have been found at any time in the whole 

 period. 



Distribution of diatoms throughout the period was quite irregular, 

 the most remarkable feature being very high productivity in the last 

 three weeks of September following a season of extremely scant pro- 

 duction and preceding a time of very light production. The accom- 

 panying graph (fig. 1) giving production in terms of averages of five 

 day periods shows this condition with astonishing clearness. The areas 

 filled in solid black indicate the relative size of average numbers 

 found in each five day period. 



In a general way the distribution agrees with that reported for 

 the last four months of the vear in other localities, both marine and 



