MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



44U 



secoiul ami third cycle.s. The nuinbor of mesenteries associated with the stoniodical .s^ysteius is 

 very variable, and in any colony, however lai'ge, there are probably only two pairs of directives, 

 situated at what may l>e regarded as the morphological extremities (p. 507). 



Compared with the cyclical complexity in Faula, Manlclna, etc., the mesenteries of 

 Jfxa/idri/K/ and C'ii/poj>/ii////'(i exhil)it a remarkable simplicity of arrangement (PI. XXI, fig. 

 141). Practically all the i)airs are complete, and may be regarded as Ixdonging to a single order; 

 here and there alternating incomplete pairs are met with, but it is impossible to establish a 

 regular succession of complete; and incomplete pairs. Never more than one or two alternations 

 occur together, while frequently six or seven consecutive complete pairs may be passed in i-eview, 

 all apparently of equal value, and all bearing filaments. The incomplete pairs in all probability 

 represent new pairs in process of development, which will ultimately become complete like the 

 rest. From eight to sixteen mesenteries may be inserted on each stomodteum, according to the 

 size of the oral aperture; apparent!}^ there is no regularity in the number of mesenteries which 

 upon fission may be apportioned to each daughter stomodteum. In the living colony a small 

 oral aperture with only seven or eight complete stomod;eal ridges, representing so many mesen- 

 teries, may appear as if just cut off from another large aperture with a dozen or more stomodaeal 

 ridges. 



Fig. 5. 

 IsopliylUa dipsacett,— Diagrammatic representation of the mesenteries in two polyps. The hexameral plan is departed from, the cyclic 



character is irregular, and no directives occur. 



The mesenterial development of Mdidciua has been traced from polyps with only one oral 

 aperture to small colonies with four stomodteal sj'stems (p. 503, et seq.), and similar early stages 

 have been followed in Fcmia (p. 508, etseq.). It is found that in simple polyps with only one oral 

 aperture the hexameral cyclical character of the mesenteriivs i.s as regular as in any other 

 sexually-produced polyp, or in adult poh'ps where columnar budding predominates; two pairs 

 of directives also occur. It is only after fission has been established that irregularities begin 

 to appear, and the hexameral plan is altogether lost. The subject of polypal fission is fully 

 described, and from the manner in which the proces.s takes place it can be easih' understood 

 how the above mesenterial irregularities come to be established. 



It may therefore be taken as a general rule among corals reproducing l)y fission, that whether 

 they give rise to distinct polyps, or form meandering systems, their mesenteries tend to lose the 

 hexameral cyclical arrangement chai'acteristic of the earliest stages, and probably never more 

 than two primary pairs of directives arc found in an}- colonv, however large. Where the fission 

 polyps are distinct, a C3'clical disposition continues to be more or less recognizable, but where 

 complex meandering systems ai"e formed only a single order, including some developing pairs, can 

 be established. 



