28 CRYPTOLARIA COXFERTA. 



sarc of the branch. It is possible that this floor disappears with age, and 

 that the older h3'^drothec8e, where they are immersed in the fascicled stem, 

 are without it. In Cryptolaria longitheca, another species occurring in the pres- 

 ent collection, the hydrothecge appear to pass continuously into the tubes of 

 the hydrocaulus without the intervention of a perforated floor. I have had 

 no opportunity of examining the nearly allied genus Grammaria, but accord- 

 ing to Sars the hydrothecae in this genus form continuous tubes passing un- 

 interruptedly into the tubes of the fascicled stem and allowing of the entire 

 retraction of the hydranth from the hydro thecaB into the tubes of the stem. 



On the branches of the specimen here described there occurred here and 

 there certain very remarkable bodies, the real nature of which I have not 

 succeeded in placing beyond doubt. They are of an irregularly fusiform 

 shape, and at the spots where they occur surround the branch like minute 

 sponges. A closer examination shows them to consist of a multitude of flask- 

 shaped, apparently chitinous receptacles (Figs. 9, 10), adnate to one another 

 by their sides, and springing by a narrow base from an irregular network of 

 tubes which encircles the branch. The distal extremity of each is prolonged 

 into a free neck-like extension which terminates in an even circular orifice. 



Each receptacle gives exit after a time to a single spherical body, which 

 is retained for a period in an external membranous sac connected by a nar- 

 row neck to the orifice of the flask-shaped receptacle (Fig. 9, a, a). 



It is scarcely possible not to recognize in these bodies an assemblage of 

 true hydroid gonangia, each giving origin within it to a single ovum, which 

 is subsequently expelled from its cavity and lodged in an acrocyst in which 

 it continues to be for some time retained. 



With the exception, indeed, of there being no apparent hydro theca3 in- 

 tercalated among the gonangia, the bodies in question resemble in all 

 essential points a colony of Coppinia. For, just as in Coppinia, we have 

 here a colony of mutually adherent gonangia, each containing a sporosac 

 with a single large ovum, which after a time is carried out and retained 

 within an acrocyst. The absence of apparent hydrotheca^, liowever, will 

 not allow us to make too close a comparison with Coppinia or to regard 

 these enigmatical bodies as constituting a hydroid colony complete in itself 



Another view, however, suggests itself May they not represent the 

 gonosome of the hydroid with which they are associated ? ' In favor of this 

 interpretation it may be urged that nothing else which can be regai-ded 

 as a gonosome occurs in the specimen, and that if we look upon them as 



