498 



per centimeter, a residue of .3 of one per cent, which had escaped 

 through the fiher-paper. (d) A Fox River collection made Sep- 

 temher 2, containing 3 161 organisms per cubic centimeter, yielded 

 in the filtrate 12.6 organisms per centimeter, less than .4 of one per 

 cent. 



There was, of course, always some undeterminable loss through 

 adhesion to the filter-paper, but this was minimized by the large 

 amount filtered (50 cubic centimeters), and by thorough washing of 

 the filter during the removal of the lO-centimeter residue. To all 

 appearance the loss was very slight, and, the error being practically 

 uniform since the same methods were used throughout, it could not 

 invalidate comparisons. 



Definition of Terms. — As a considerable part of this paper will 

 be given to a description of the criteria and effects of different de- 

 grees of contamination of the natural waters of the Illinois, we will 

 distinguish, as clearly as convenient, three stages of impurity, by 

 the use of the following terms applicable both to the waters them- 

 selves and to the characteristic organisms, given here in the order 

 of a diminishing impurity, namely, (i) septic or saprobic, (2) pol- 

 luted or pollutional, and (3) contaminatecl or contaminate; and to 

 these we will add "clean water" to indicate the conditions and organ- 

 isms substantially equivalent to those of the natural, uncontaminated 

 stream. These expressions seem sufficiently definite and significant 

 to distinguish between stages which at best can not be sharply marked 

 off; and we prefer them for our purposes to the equivalent highly 

 technical terms polysaprobic, mesosaprobic a, mesosaprobic j8, and 

 oligosaprobic, introduced by Kolkwitz and Marsson. 



Only a very general account of the product of these collections 

 will be given at this time, the subject being reserved for more de- 

 tailed report in a later paper. 



For a merely general and preliminary presentation of our data 

 of 191 1 relating to the midsummer microplankton of the upper Illi- 

 nois, we may conveniently divide the waters from which our collec- 

 tions were made into four sections or general situations, correspond- 

 ing to distinguishable stages of pollution and self -purification; (i) 

 the sanitary canal at Lockport, (2) the Des Plaines River at Dresden 

 Heights and the Illinois River from that point to Marseilles, and 

 (3) the Illinois River from Marseilles to Starved Rock, and (4) 

 from. Starved Rock to Lockport. 



In the sanitary canal at Lockport we found, in September, 191 1, 

 an abundant Lake Michigan plankton with little admixture of septic 

 organisms. It was largely composed of characteristic lake diatoms 



