212 HYATT, 



muscular, and fifth membranous layers to form the reticu- 

 lated walls, (5) the increasing growth of the arms. 



These are the fundamental characteristics, and yet it 

 cannot be denied that they may first make their appear- 

 ance as a variable characteristic of diflTerent brandies of 

 the same colony become common to the varieties of other 

 species, and eventually of generic value. Thus, as before 

 described, a single branch of P. vesicularis assumed the 

 aspect which afterwards distinguished variety c of P. 

 vitrea, and approximated to the generic characteristic of 

 Lopbopus. The obliteration of the cell-walls in this 

 branch and in the colonies of variety c of P. vitrea were 

 traced to the increase in the number of buds growing up 

 side by side from the same cell, just as in Pectinatella and 

 Cristatella the generic form of the colony was traced to 

 the number of buds developing simultaneously from the 

 parent cell. 



The second, also, as has been stated, appears as a rare 

 variety in Fredericella rcgina and Walcotti, an ordinary 

 character in F. pulcherrima, and among the Plumatellse 

 ending with becoming of specific vame in P. vitrea, per- 

 haps on account of the short time which that species lives, 

 and is finally common to all the species of the remaining 

 genera of the group. It is evidently a local characteristic, 

 due partly to age and partly to physical causes in these 

 two genera; but in Lophop us, Pectinatella, and Crista- 

 tella, it is an essential peculiarity of the generic structure, 

 apparently not affected by physical causes, since wherever 

 these genera have been found the ectocyst is gelatinous. 



With regard to the third, it is possible that the en- 

 largement of the cells in Pectinatella indicates a loss of 

 the execretory power, and consequently causes the re- 

 moval of the ectocyst. As previously stated, this change 

 is begun in variety c of Plumatella vitrea, where the action 

 of alcohol creates a gap between the ectocyst and endo- 

 cyst. 



The fourth is even more inexplicable than the third ; 

 the walls are there, and are evidently necessary to keep 

 the ccenoecium firm and shapely, but they are apparently 

 without precedent. They do not begin, like other char- 



