6 Beecher — Observations on the Genus Romingeria. 



same plane and thus aborting manj of the buds, which only 

 grow to be a few millimeters in length or are entirely sup- 

 pressed ; and (J) the inability of many of the corallites to 

 reach the next zone of budding before all the available space 

 is taken up by verticels from other rapid growing and more 

 favored corallites. Whenever the growth and budding are not 

 seriously restricted as to space, it is found that from ten to 

 twelve buds are almost invariably given off. 



Altogether, the probabilities against the successful attainment 

 of maturity with the growth of a single cycle of buds are 

 seldom equaled among corals. In ordinary proliferation in a 

 compound coral, a bud is developed when by a divergence of 

 the corallites through growth a space is made to receive it, but 

 in Romingeria the buds are thrown off without any reference 

 to their future possibilities. If the pores of Favosites are 

 considered as potential buds, then the amount of suppression 

 in that genus is inlinitely greater than in Romingeria. It 

 should be noted, however, that in Favosites whenever a coral- 

 lite once appeared it usually continued to grow until the death 

 of the whole colony. 



A diagram of two successive periods of proliferation for a 

 single corallite of R. umhelUfera is shown in Plate Y, figure 

 1. The third zone of buds would number 2,197 corallites, the 

 fourth 28,561, and so on, according to the permutation of the 

 number 13. 



Corallites subjected to pressure in the matrix often exhibit 

 a tendency to split into longitudinal plates corresponding to 

 the twelve mesenteries or primary septal divisions, figure 5, 

 Plate I. The twelvefold nature of the walls may be devel- 

 oped in this way without any evidence of septa being pre- 

 served. 



Rotningeria commutata, sp. nov. 

 Plate V, Figures 4, 5. 

 Romingeria umbellifera Davis, Kentucky Fossil Corals, pi. 76, fig. 1, 1887. 



The specimens identified by Davis with the type of the 

 genus have already been referred to as probably distinct. The 

 growth of the corallum is lax and the budding of the cor- 

 allites is irregular, not producing the storied appearance 

 typical of R. umhelUfera. No regularly developed 10-12 

 rowed umbels are present in the Kentucky specimens, which 

 seem to give off bundles of buds rather than to form perfect 

 rosettes. There is also a much stronger development of the 

 8-10 rows of trabeculse, which are not confined wholly to the 

 budding region, but may extend for considerable distances 

 along the corallites and also appear at the accidental points of 

 contact of two corallites where adventitious mural pores are 



