174 



C.-E. A. WiNSLow AND Anne F. Rogers 



1 7° J (3) water, 95; (4) earth, 67; (5) air, 109. It is probable that 

 this deviates from a representative sampling of the cocci in nature 

 by laying too great stress on the saprophytic forms. It is difficult 

 to find cocci at all in earth and water, whereas they are present 

 on the surfaces of the body in enormous numbers. The majority 

 of this group appear to be parasitic or semiparasitic in habit. At 

 the same time, the fairly equal weight given to the saprophytic forms 

 helps to bring out the differences between the two main groups, 

 those living in or on the animal body (i and 2), and those living in 

 the outer world (3, 4, and many of 5). 



We have prepared tables to show the distribution of each charac- 

 ter among various habitats, and the relations shown are so suggestive 

 as to warrant rather full discussion. In Table i is given the cor- 

 relation between habitat and cell-grouping, and it is at once evident 

 that the sarcinae occur in greater proportion outside than inside the 

 body. 



In this and succeeding tables the figures represent the number of 

 cultures showing each combination of characters out of the 500 

 studied. 



TABLE I. 



Correlation between Habitat and Cell-Grouping. 



Whereas packets are more abundant in earth and water, the 

 other forms — chains, plates, and irregular groups without sarcinae 

 — make up two-thirds or more of the parasitic forms. 



TABLE 2. 

 Correlation between Habitat and Gram Stain. 



The distribution of our cultures according to their relation to 

 the Gram stain brings out a similar condition. The cultures giving 



