543 



rections, such as coefficient, have been made, so that the data are 

 as comparable as the circumstances will permit. The data in 

 the table are derived from 642 catches, and cover a sufficient 

 range of seasons and years to yield averages significant of any 

 larger relations existing between the several bodies of water 

 examined, and also to illustrate the influence of certain com- 

 mon, as well as some contrasting, factors in this fluctuating en- 

 vironment. 



The backwaters in the list of stations examined by us are 

 typical of practically the whole range of the bottom-land wa- 

 ters of the Illinois, including, as they do, a reservoir and a 

 spring-fed lake, both tributary at practically all seasons, a 

 marsh, and a lake tributary for only a part of the season and 

 free from vegetation. Results derived from an examination 

 of these will therefore afford some conception of the relation- 

 ship between the plankton of the backwaters of the stream as 

 a whole and that of the channel. 



In the discussion of production in these backwaters ( pp. 

 350-454) I have noted the periods when hydrographic condi- 

 tions permitted a run-off of the backwaters impounded in the 

 several localities to the main channel, and instances in which 

 their plankton contents served to increase or dilute that of the 

 channel. It will suffice in the present connection to note that 

 production is less in the backwaters than in the channel in 

 only 48, or 26 per cent., of the 185 monthly averages — the total 

 number of all of the monthly averages of all stations but the 

 Illinois and Spoon rivers. In Quiver Lake we find 30 months 

 out of 53, or 57 per cent., deficient in production as compared 

 with the channel ; in Dogfish Lake, 6 out of 25, or 24 per cent.; 

 in Flag Lake, 2 out of 24, or 8 per cent.; in Thompson's Lake, 8 

 out of 52, or 15 per cent.; and in Phelps Lake, 2 out of 31, or 6 

 per cent. 



Lakes most free from access of tributary water and from 

 vegetation are most constant in their excess of production over 

 channel waters. Thus we find that of these 48 instances of 

 deficient production in the backwaters 30 occur in Quiver 



