others . Groups other than diatoms, greens, and blue- 

 greens were rare in both years; so rare that the counting methods 

 employed could not determine their abundance vdth accuracy. Hence, 

 it is not worth while to attempt to trace seasonal changes in 

 abundance. These groups made negligible contributions to the bulk of 

 the plankton. 



Adequacy of the sampling program 



In the section on horizontal distribution, it was pointed 

 out that the plankton is not uniformly distributed. It may be 

 well to consider, now, whether the sampling program as indicated in 

 Tables $1 and 53 was adequate to show seasonal changes in abundance 

 with an acceptable degree of accuracy, 



A study of the available data from individual stations shows 

 that the program was adequate to bring out the seasonal trends in 

 abundance. That is, the records of individual stations agree closely 

 as to the times of abundance and rarity of plankton. Each station 

 showed spring and autumn maxima, and a summer minimum of diatoms. 

 Also, each station showed maxima of greens and blue-greens in autumn 

 only. 



However, the records show some marked discrepancies in the 

 actual abundance of organisms at different stations in the same period 

 of time. In some cases the discrepancies are so large that caution 

 must be used in drawing conclusions from the averages. Certainly a 

 small difference between the averages of the same group of algae in 

 two periods, or between two groups of algae in the same period should 

 not be regarded as indicative of real difference in abundance. For 

 example, it would be improper to conclude that the autumnal maximum of 

 blue-greens was greater in 1930 than in 1929 (Fig* 17), or that greens 

 were more abundant than diatoms in July, 1930 (Fig. lU). However, 

 if small differences appear consistently over a period of time, it 

 appears proper to consider them as significant. For example, prior 

 to August in both years, blue-greens were never more abundant than 

 greens and were equal to them in only three periods, (Figs, 13 and 

 lU) , Although the superiority in numbers of greens over blue-greens 

 before August was never great, it was too consistent to be regarded 

 as accidental. 



In spite of the rather large differences in counts between 

 stations during the same period of time, there are good reasons for 

 believing that the averages give a fairly accurate record of the 

 seasonal changes in abundance. The first reason for believing so is that 



171 



