568 PROCJ^EDINGS OF THE NATIONAL MUSEUM. vol. xxvi. 



o-enerally one side of the zooecial cavity as a series of superimposed 

 vesicles. When tangential sections cut these vesicles, they show 

 within and extending- across the zooecial cavity, a curved line, the 

 amount of curvature depending- upon the gibbosity of the cystiphragm. 

 In If. callosa Ulrich, and //. rfmmilosa (see Plate XXV, lig. 3) this line is 

 but slio-htly curved, showing that the vesicle was little rounded. In 

 the type species the vesicle is of such a shape and occupies so much 

 space as to cause the cut edge shown in tangential sections to extend 

 around about two-thirds of the circumference of the zorecial chamber. 

 A diti'ercMit style of cystiphragm occurs in such species as IT. jndehm 

 and //. cincinnatiemis. Here, occasionally, the cystiphragm extends 

 entirely around the bounding wall l)ut leaves the central portion of 

 the zooecial cavity unoccupied, and here ordinary horizontal diaphragms 

 are developed. In this case, tangential sections (Plate XX, lig. 12, Plate • 

 XXI, lig. 7) show the cystiphragm as a more or less rounded, central 

 ring. In vertical sections the cystiphragms appear as semicircular 

 lines lining usually one side of the zocecial tube, but when the vesicles 

 extend entirely around the cell cavity both sides show a series of 

 curved lines. 



The portion of the peripheral region of the zooecial tube not occu- 

 pied l)y the cystiphragms is generally intersected by transverse par- 

 titions, the diaphragms, which may or may not be as numerous as the 

 cystiphragms, seldom however exceeding them in rmmber. It is also 

 to be noticed that in forms showing no diaphragms in the zooecial 

 tubes the cystiphragms seldom overlap, the lower end of one not 

 reaching to the next below. If this interpretation is true to nature it 

 would imply that the cystiphragms were open at the bottom. How- 

 ever, one or both of the following explanations may account for thisi 

 appearance. The cystiphragms are seen in vertical sections to bei 

 thickest at the upper end, and often are reduced in thickness to ex- 

 treme thinness at the point of overlap. In the forms under discussion 

 the lower part of the cystiphragm may have been so thin that sections 

 do not reveal it at all. Another interpretation is that suggested by 

 Nickles in the description of II. hasderl, that calcitication in the living 

 state was more or less incomplete. 



Upon the basis of the distribution of the cystiphragms the II. com- 

 nmnk group of the genus maybe further subdivided into two sections, 

 one in which cystiphragms line the peripheral region of all the zocecia. 

 and another in which the zoacial tubes of the maculae only are pro- 

 vided with cystiphragms, as seen in sections of //. austiiri. Tht 

 latter, an unusual occurrence, may throw some light upon the func- 

 tions of l)()th macuhe and cystiphragms, Ulrich has considered th( 

 macula' (including the monticules and groups of larger zocecia, all OJ 

 which are evidently identical in function) of trepostomatous bryozoj 

 to l)e coimected in some way with reproduction, these groups perhapi 



