300 



nant or more typical species. In the case of the total plankton 

 some obscurity results at times from the inclusion of unusual 

 proportions of -an adventitious population with flood waters. The 

 selection of particular organisms as representative is also subject to 

 some error, since seasonal changes in temperature and other more 

 subtile causes often deflect or suppress their development. The totals 

 of the ChlorophycecB, Bacillariacece, and chlorophyll-bearing Masti- 

 gophora, and of the Rotifera and Entomostraca (PL I. -IV.) probably 

 give as complete and accurate a delineation of the recurrent pulses 

 as the statistical data afford, since they include relatively few 

 adventitious organisms, cover the entire year, and swamp more or 

 less completely individual and temporary divergences of particular 

 species. The delineation of the pulses by statistical data is obviously 

 more significant than the volumetric method, since it more clearly 

 presents the results of the reproductive processes which lie at the 

 foundation of the phenomenon of recurrent pulses ; and this method 

 is also free from the unavoidable error arising from the presence of 

 silt in the collections. 



The interval between collections introduces an error of consid- 

 erable moment in any effort to determine with accuracy the duration 

 of individual pulses, that is, the length of time between their minima 

 or maxima. Daily collections would render this feasible, but with 

 an interval of a week or more, not only the duration, but in some 

 cases the probable separation of the pulses and location of their 

 maxima, is to some undetermined degree obscured. 



The duration of the pulses of the five groups of plankton organ- 

 isms shown graphically on Plates I. -IV., in the case of all chlorophyll- 

 bearing organisms considered as a whole, is in 29 out of 36 instances 

 between 21 and 35 days, less than 21 in 2 cases, and more than 35 

 in 5, reaching extreme limits of 14 and 49 days. They average 

 30.25 days between minima and 29.97 between maxima. 



The rotiferan data in the same months may be divided into 36 

 periods, in 33 of which pulses are traceable. The duration of pulses 

 between minima lies between 2 1 and 3 5 d ay s in 2 3 of the 3 6 instances , 

 falls below 21 in 5, and is above 35 in 8. The extreme limits are 

 14 and 49 days. 



In the case of the Entomostraca, where also the pulses are obscure 

 in a few of the intervals, we find that 22 of the 36 are between 21 

 and 35 days between minima, 5 are below 21, and 9 are above 35. 



