130 FAMILY MILIOLIDA. 



considerable variety, too, in regard to the proportion whicli tlie central space bears to the 

 breadth of the annulus ; the inner circle being for the most part relatively smaller in the 

 flattest annuli, and increasing in diameter as the annulus itself swells out and diminishes in 

 breadth. It is a remarkable feature in this type, which may be distinguished as D. annulus, 

 that the annulus is usually divided with great regularity into chambers of equal size ; a 

 regularity 7or which it is difficult to account on the supposition that the entire ring 

 has been developed by the successive growth of independent chambers. After a careful 

 examination of numerous specimens, I have only met with one which gives any distinct 

 indication of a line of junction, such as might be expected to result from the meeting 

 of the two ends of an imperfect ring completed by the successional addition of new 

 chambers. And it may therefore be fairly questioned whether the uniform annulus, which 

 is certainly the normal type, did not originate in a radiating outgrowth of segments from one 

 large primordial mass of sarcode occupying its centre, in the same manner as the first 

 annulus of chamberlets in the typical Orhitolites (^ 161) is formed around the primitive disk; 

 the space at first occupied by the primordial segment being not covered-in by shelly w^alls, 

 and being perhaps vacated in these annular forms so soon as the surrounding ring of 

 chambers has been consolidated. We shall hereafter see reason to believe that in the more 

 complex type of Dadijlopora the central cavity continues to be occupied b)" the sarcode-body 

 through life, and that it is in this portion that all new annuli have their origin (f 197). 



191. In each of the two forms now described a curious variety is occasionally met 

 with, resulting from what may be termed a " wild "' growth of the segments (such as we shall 

 frequently encounter in certain forms of the vitreous series), which tends to convert the 

 closely set, flask-shaped chambers into elongated, divaricating tubes. This is most strongly 

 marked in the pupoid type, the chambers of which occasionally undergo such an elongation 

 as to be converted into cylinders (fig. 16) so closely resembling the cells of Ti/biiUpora or 

 Celh'pora, that, if detached from each other, they might easily be mistaken for fragments of 

 those Polyzoa ; and, in fact, the determination of this form as a variety of Bactyhpora, under 

 the designation B. dic/itaia, chiefly rests on its evident relationship to that to be next described. 

 No specimen has yet been found sufficiently perfect to exhibit the unbroken terminations of 

 the cells ; but there can be no reasonable doubt that they are closed at their diverging 

 F XWII extremities (as shown diagrammatically in Fig. XXVII), 



hke those of the forms to which they are related. — A 



similar tendency occasionally shows itself in a less 



marked degree in the annular type (fig. 15); giving 



rise to the form which has been described and figured 



by Michelin ('Icon. Zooph.,' p. 177, pi. xlvi, fig. 27), 



under the name of Cli/peina mar^inoporella, as a member 



of the family TuhuUporida. The chambers here also 



are elongated and subcyhndrical ; but they remain in 



Diagrammatic seetior7z)../^/o^.m^,>,-^«/a,- adhesion to each other laterally, so as to lie obliquely 



—A, horizontal; b, vertical; 1, 1, line of to the axis of the ring, and to form a sort of inverted 



^™'"'^'^' . funnel, as is shown in the diagrammatic sections in 



Fig. XXVIII. The pores represented by Michelin along the external margin have no real 



