32 AMERICAN HYDRUIDS. 



A study of .serial longitudinal sections of Dip hasia fall aa> 1 shows that no true acrocyst is 

 found in this species. It is true that an examination of entire adult gonangia with transmitted 

 light, seems to reveal an inner globular chamber besides the outer one formed by the gonangial 

 leaves. That this is an optical illusion is seen when a median longitudinal section is studied 

 (see fig. !>7). 



The inner and outer walls of a gonangial leaf are seen to be widely separated and not strictly 

 parallel, the distal end being much thicker than the rest. The result is that the inner profile of 

 the leaf forms nearly a half circle and, in conjunction with its fellow on the opposite side, forms 

 nearly a complete circle which looks almost exact!}' like the outline of a sphere when seen from 

 the side. Thus it happens that we have the appearance of a sphere in the center of the marsupial 

 chamber, occupying, indeed, the exact position of an acrocyst. As there are eight of these leaves 

 in 1>. r<nii;-a their inner edges would thus simulate the outline of a .sphere when viewed from 

 any side. 



It might be argued that the acrocyst, according to my own statement, is but a temporary 

 structure, and might therefore have been absent in the specimens studied by me, but present in 

 those studied by Professor Alhnan. In some of my specimens there were ova or planuhe in the 

 marsupial chamber (rig. '>~i,j>). Under these conditions, if ever, the acrocyst would be present. 



The male colonies of I>/2>huia produce gonangia without the, marsupium, and hence of very 

 different appearance, and it was this fact, doubtless, that led the elder Agassiz to give the name 

 l ' Diji/taxia" to this genus. Fig. !s represents a longitudinal median section of a male gonangium 

 of I)ij>li<ixi<i jxini'iniiniii containing a blastostyle which bears gonophores in a series, each 

 consisting of a simple spadix surrounded by a mass of sperm cells. The gonangia are unusually 

 long and slender, and often contain a row of five or six gonophores. 



We have now considered all of the distinct types of gonophores found in the Sertularidse, so 

 far as known to the writer. 



THE GONANGIUM. 



This structure is much more diversified in the Sertularid.e than in the Pluumlarida 1 , due 

 probably to the fact already suggested that the gonangia in the latter family are often protected 

 by various forms of phylactocarps, and hence are not so much influenced by the immediate 

 environment of the species. 



Perhaps the most typical form of gonangium is the simple oblong oval, truncated at the top, 

 well illustrated by X ffnlnriu /nun/In, (fig. !!(), A'. tip, ,(///,/?,/ (fig. 10(1), .V. xtnnl-fijl (fig. 101), 

 Thiiiitrni ilnijii (fig. 102), Diphasia l-in<-<ti<li (fig. 112^), and Al>i<1in<iri<i </i</initea (fig. 103). The 

 main modification of this form consists of the narrowing of the distal end of the gonangium so 

 as to form a short tubular neck, as in A. cnxfnfii (fig. 1<>4), Tlin'mr'ni fnr<//<!<i (fig. 105), and 

 T/tuiaria tubulifoi'mix (fig. 100). This narrowing may be such as to form a short cone instead 

 of a tube, as in Dictyocladium //<///,<//*// (fig. 107); or it may form a frustum of a cone, as in 

 Abidi mi riii ///'(( /n't (fig. IDS). Where the aperture is large an operculum is usually present, as 

 in Scrtularia stookeyi (Plate V, fig. 6), and Sertuta/rella formosa (fig. 109). 



The most common form of ornamentation found in the gonangia of this group is brought, 

 about by annular rugosities which are often exceedingly graceful and beautiful. O'ne of the 

 most attractive structures in the whole family is the gonangium of X, rlnl<ir< l/n tricuspidata 

 (Plate XXV, figs. -t and 5). Similar gonangia are found in #._//'/ ifunn is (fig. 113), and the mo*t 

 excessive ornamentation along this line is seen in X, rfulm; -lln , In/mi* (fig. 114), where the 

 annulations take the form of greatly compressed ridges which are elaborately frilled so as to 

 resemble lace work (Plate XXIV, fig. 1). Sometimes these annulations are confined to the 

 distal or upper part of the gonangium, and are much broader and less incised, as in X. nil muni 

 (fig. 115), X. catena (rig. Hi), X. meridionals (fig. ll(i). and S. <ill>i<l<i (fig. 122). 



1 These sections, as well as the < it hers used in the study of the gonosome of the Sertularidie, were made for me 

 l>y Mr. William B. Bell, one of my students. 



