Tlie Productimty of Fish Food in Oneida Lake 91 



of methods. The material collected was all sorted in glass 

 dishes, with the aid of hand lenses up to ten diameters. 



Petersen ('11, pp. 50-53) counted and weighed the animals 

 obtained, giving both "dried" and alcoholic weight. His 

 material varied so greatly in size that a measurement of mass 

 per unit area was necessary. In making the studies in Lower 

 South Bay all of the individuals were counted. It was thought 

 that counting would be mrtre practicable, and suggestive in 

 studies of the contents of the stomach and digestive tract of 

 fishes. By this means a uniformity of method is secured for 

 both stomach contents and for the estimates of the food supply. 



Comparison of the Bottom Habitats of Lower South Bay 



For the purposes of this statistical study Lower. South Bay 

 includes all of the bay ])roper and the area extending eastward 

 and northward within a line drawn eastward from Long Point 

 which meets another line drawn northward from Norcross 

 Point (see fig. i). This area embraces 881 acres and the total 

 area sampled, including certain territory in the vicinity of the 

 bay, aggregates 1,164 acres (see the map, fig. i). 



Petersen selected mollusks and echinoderms to characterize 

 his animal communities, because these groups comprise the 

 greater part of the bottom fauna in these marine waters and 

 also because they are not seasonal animals. Their ease of 

 preservation for future reference, because of the hard 

 exoskeleton, was also another feature in their favor. The 

 marine communities were divided into principal species, at- 

 tendant species, and casual species, those wuth the greatest 

 frequency being principal and those occurring but rarely being 

 casual species. For the reason that mollusks are more numer- 

 ous than any other single group they have been chosen to 

 characterize the difi^erent communities in Lower South Bay ; 

 they also usually contain the principal or dominant species. 

 Attendant species may be mollusks but are more often some 

 group of associated animals. For purposes of better compari- 

 son the quantitative data is arranged under several heads, each 

 based on the character of the bottom material — boulder, 

 gravel, sand, clay, mud, and vegetation. Certain areas in the 



