STUDIES IN MARINE ECOLOGY. 1 85 



Again the analysis follows the records through the season of 1920. 

 The comparison is based on total distribution records as was 

 the last. That is, for the purposes of this table the finding of a 

 given animal once in a given type of habitat is as significant as 

 though it were abundant there. There were 242 species in the 

 catalogue when this study was made. 



It is immediately apparent that this type of analysis serves 

 only to call attention to the larger number of animals taken 

 from the fiats and beyond indicating the higher specificity of the 

 flats, shows no evidence of relationships that may exist between 

 different habitats. The species taken from any given type of 

 habitat are found to be approximately equally distributed in the 

 other habitats of the region. Such experience has led casual 

 observers to conclude that relationships between different animal 

 associations can not be analyzed. 



In order to make such an analysis, the records were studied 

 from another angle. Species approximately equally distributed 

 through the different associations and those reported for one 

 season only were eliminated. Then the remaining species were 

 listed according to the habitat in which they are most abundant. 

 When a species was found equally abundant in two habitats it 

 was listed from both. The association in which the animal was 

 next most abundant was also estimated. Unfortunately these 

 records are based on "experience" cards that seldom give 

 specific figures and on the number of collecting teams that have 

 reported the species from the different associations, as in the 

 data given in the check list, rather than on strictly quantitative 

 grounds and while substantially correct, they lack the finality 

 that statistical treatment would give. 



This type of analysis gives a real basis for comparisons of the 

 relationship between the different associations. It is of interest 

 that the relationships shown by Table II. are practically the 

 same as given by an analysis of the entire catalog in 1917. In 

 other words, the early collecting gave the typical forms char- 

 acteristic of the environment while much of the later work has 

 yielded in addition to these characteristic species, a number of 

 accidental or incidental records. 



13 



