OBSERVATIONS ON POLYZOA. 199 



without fear of misunderstanding to the consideration of 

 its general features. There are six characters; (1) single 

 character of the crencecium, (2) a character in the diffu- 

 sion of the branches, (3) a character in the tubularity of 

 the branches, (4) a character in the perfection of the 

 branches, (5) a character in the distinctness of the cells, 

 and (6) a character in the entireness of the walls of the 

 cells, all of which are decremental, the decrement being 

 traceable directly to the activity of the reproductive func- 

 tion, which increases, (1) the radiatory character of the 

 coencecia, (2) the crowding and adherence of the branches, 

 (3) the lobiform character of the branches, (4) the con- 

 solidation of the branches with the cosncecial trunk, (5) 

 the adherence of the cells to each other, and (6) the 

 merging of the cell-walls into the common surface of the 

 branch. All of the latter are respectively antagonistic to 

 the former, and although produced by the same agent, and 

 occurring primarily either upon different branches of the 

 same colony, different varieties of the same species or dif- 

 ferent species of the same genus, finally become of ge- 

 neric value,* exclude the former, and change the whole 

 aspect of the cceno3cium in Cristatella. 



Secondly, there is a series of characters; (1) a char- 

 acter of adherence of the ectocyst to surfaces, and of the 

 endocyst to the ectocyst, (2) the parchment-like character 

 of the ectocyst, (3) its permanency throughout life, all of 

 which decrease as the ectocyst becomes more gelatinous 

 and retrogrades from a protective covering to a transient 

 excretion. Thirdly, there is a series, (1) the erectness 

 of the free branches, (2) the envelopment of the ccence- 

 cium, (3) the small size and depressed form of the cells 

 of the outei- membrane, all characteristics which disappear 

 when the ectocyst retires from the upper side of the coenoe- . 

 cium. Fourthly, (1) the number of branches in a single 

 colony, a decremental character, (2) the permanency of 



* By " generic value " it is not meant that a variety may become a 

 species, or a species a genus by evolution, but that characteristics ap- 

 parently of accidental growth in the distortion of a branch, or in the 

 variety of a species, may finally become peculiar to some species, and 

 then an essential characteristic of a higher genus. 



