28 



EDGAR DAVIDSON CONGDON. 



the end of the branch. Near the terminal hydranth they are so 

 large as to crowd the neighboring entoderm cells, bulging the 

 entoderm inward and mesoglcea outward. Their shape and posi- 

 tion are such as most easily to force their way along the branch. 

 They are flattened between mesoglcea and entoderm and their thin 

 edges are inserted, wedge-like, between the two layers. In the 

 fixed material I could find no indication of pseudopods. The 

 entoderm cells about them are compressed and distorted in the 

 way we would expect if the egg should move along under them. 

 There are two types of female gonophores differing so much 

 in size and position that one might reasonably consider them en- 

 tirely independent, were it not for the evidence of their common 

 origin. In one type the outline of the spherical egg can be 

 readily seen. An entodermal thickening passes half way round 

 it extending from its attachment to hydranth, up over the top, to 



a point opposite its origin (Fig. 

 i). Here the ridge divides into 

 a pair of branches which form 

 open coils on opposite sides of the 

 egg. Other gonophores have a 

 distinctly different position in the 

 colony. They are ovoid and the 

 egg is largely hidden by a sheath 

 of ectoderm and entoderm which 

 on account of its development 

 may be looked upon as a straight 

 spadix (Fig. 2). The two kinds 

 have been termed orthospadiceous 

 and streptospadiceous according 

 as the spadix is straight or coiled. 

 Streptospadiceous gonophores 



form whorls around the bodies of hydranths. Two, four, five 

 or six gonophores are usually associated. Either the single 

 terminal hydranth of a branchlet bears the gonophores or 

 several succeeding hydranths may do so. The appearance of a 

 gonophore terminates any further growth of a branchlet. 



The earliest indication of the gonophore is an outwardly di- 

 rected pocket of ectoderm and mesoglcea containing the egg 



