322 



In P. compressa, for instance, the phases of the cell- 

 growth are, in all respects, similar to those which occur 

 in P. concinna ; and to separate these two forms would be 



Fig. 14. 



Porella compressa. 



to make more of the erect habit than of the essential 

 elements of the structure. If this course were adopted, 

 there would be an indefinite multiplication of genera; 

 for the same type of cell is found associated with very 

 diverse modes of growth. Unless we are prepared to 

 hold to the old method and allow the most external and 

 superficial characters to rule, to the exclusion of those 

 which are the real indications of affinity, we must, I 

 think, found our genera in this division on the structure 

 of the zoo3cium, and treat the ordinary variations of 

 habit as merely subsectional. 



