OBSERVATIONS ON POLYZOA. 213 



acteristics, in a few individuals or species, gradually be- 

 coming of generic value and general application as the 

 complication increases, but start into life and functional 

 activity without even a whispered warning of the coming 

 change in the organization of any preceding genus or 

 species. 



The fifth is characteristic in different degrees of the 

 individuals in Plumatella (the length of the arms always 

 varying with the number of tentacles, and these are differ- 

 ent in nearly every individual), and becomes much more 

 constant, if not invariable, in Cristatella. Four of these 

 five fundamental characters, therefore, are incremental. 

 These begin in a few individuals and become of more 

 general value and characterize more and more extended 

 divisions and more complicated animals. Thus the sec- 

 ond is an individual variety in the first genus, obtains a 

 more general application and finally a specific value in the 

 the second, and at last is sufficiently constant in the two 

 species of Cristatella already known, to be considered at 

 least of generic" value. 



The increased value, or constancy and applicability, 

 therefore, of these incremental characteristics is due en- 

 entirely to the increase in the number of the individuals 

 characterized by them in each succeeding and more com- 

 plicated division, with regard to the number of the indi- 

 viduals that remain unchanged in the same division. For 

 example, in P. arethusa the majority are brownish, 

 wherever found ; in P. vesicularis there are few brown 

 individuals in proportion to the transparent ; in P. vitrea 

 all yet found are transparent; and in the succeeding 

 genera every individual is transparent. 



The fourth character must be considered an exception 

 to this rule, unless indeed new discoveries should disclose 

 an intermediate form between Cristatella and Pectinatella, 

 or a lower form of the same genus that will give us 

 the clue to the abrupt introduction of the reticulated 

 walls. 



It is surprising that when viewed with regard to their 

 own young, the condition of the coencecial and lophophoric 

 regions in Cristatella, although more complicated, is nev- 



