316 



the remainder of the year. All of the collections were made 

 with the plankton pump. This was a year of high winter floods, 

 a normal March rise, a belated June rise, but prolonged low 

 water throughout late summer and autumn. The collections 

 afford a good opportunity to observe the result of prolonged 

 low water and a late autumn. 



From February to July the collections are too infrequent 

 to enable us to trace the curve of plankton production or de- 

 tect any cyclic movement. Of the two collections in February 

 the first was made under the ice sheet and yielded little plank- 

 ton or silt. The second, made while the ice was going out, 

 contained much more silt — the result of the rising flood. Both 

 were very poor in plankton (.03 and .05 respectively), but 

 neither showed the least evidence of stagnation conditions such 

 as obtained in 1895. This was due to the larger volume of wa- 

 ter and the swifter current and greater dilution of sewage, as 

 well as to the briefer ice blockade and the direct connection of 

 channel waters with the vegetation-rich bottom-land lakes and 

 forests when the ice was full of air-holes, so that the equilibrium 

 of gases in the water did not undergo so violent a disturbance 

 as in 1895. 



In the March collection there is evidence of the increas- 

 ing production as vernal temperatures approach. The col- 

 lections of April and May were both made on the decline of 

 the March flood, and both lie at temperatures between 60° 

 and 70°. This spring presented ideal conditions for a very 

 large plankton production, namely, uninterrupted decline, with 

 run-off of impounded backwaters in which the plankton had 

 had abundant time to breed. Neither of our vernal collections 

 shows any large production, though that of April lies in the 

 period in which the vernal maximum may be expected. It is 

 not improbable that a maximum occurred but was not detected. 

 The June collection lies in the midst of turbulent flood waters, 

 as the great proportion of silt (26.33 cm. 3 per m. 3 ) indicates. 



From this point until the close of our operations in March, 

 1899, the weekly interval of collection was adopted, and the 



