MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 427 



tentacles. These four, two on each side of the median plane, are appi-oxiniatcly eiiual. Later, 

 in describing the relationships of the mesenteries of Madrepora, it will be seen that the large 

 anterior abaxial tentacle is dorsal or sulcular in position, while the opposite axial tentacle is 

 ventral or sulcar as regards the polyp as a whole (p. i44). 



Of the smaller alternating series of six tentacles the abaxial laterals (one on each side of the 

 large at)axial tentacle) are always the smallest, and are generally colorless throughout. The 

 middle laterals come next in size, and the axial laterals may be a little smaller than these. The 

 difference in size between the middle and axial laterals is, however, often scarcely perceptible; 

 l>ut the four are always larger than the two abaxial laterals, and are more deeply colored. In 

 polyps near the apex of growing branches all the tentacles may be colorless. 



As far as can be made out in the living state, the tentacles of the very minute, intercalary 

 polyps are uniform in size, and in regions where the corallites possess a circular, free edge 

 the tentacles tend to become more uniform in size. The large abaxial tentacle is always best 

 developed in polyps where the corallite has the most marked nariform projection, as in M. 

 ccri'icornis; undoubtedly, there is a relationship between the form of the mouth of the corallite 

 and the amount of inequality among the tentacles. 



In L. Agassiz's Report on the Florida Reefs (PI. XVIII) an outline figure of an expanded 

 terminal polyp of Madrepora cerL'lcomls is given, in which six large equal tentacles alternate with 

 six much smaller tentacles, likewise equal. Such a stage is occasionally met with on young 

 branches, but is to lie regarded as transitional to the stage in older branches with onh' six equal 

 tentacles. On the same plate are also outline figures of expanded lateral polyps from near the 

 tiji of a branch; as there represented the abaxial aspect is uppermost. 



Prof. A. E. Verrill (18^9) was the first to draw attention to this vai'iation in the external 

 characters of the axial and radial polyps of Madrejwnt, and regarded it as the only instance of 

 dimoi-phism among the Madreporaria. The aj^ical jwlyps are seen, however, to be derived by 

 modification of the radial, and, as will be shown later, the internal anatomy of the apical and 

 radial polyps presents no differences corresponding with those of the tentacles, so that the 

 dimorphism is not very deep seated. 



The tentacles in all the West Indian species of Porites are, like those of Madrepora, usually 

 twelve in number. Developing polyps exhibit a less number, and others occasional!}" occur in 

 which the number may be fourteen, sixteen, or as many as twenty-four. On the colonies no 

 distinction is to l)e made between apical and radial polyps. The tentacles of all the poh'ps 

 are exti'emely small, smooth-walled, and digitiform, rarely exceeding 1 or 2 mm. in length. 

 Viewed with a lens, in their fully expanded condition, or even when introverted, they appear to 

 constitute but one cycle, and very often differences in size are recognizable of the same character 

 as in Madrepiira (PI. IV, fig. 32). The two tentacles in the longer oral axis are somewhat 

 larger than the others, and one of these, corresponding with the abaxial in Madrepora^ is 

 somewhat longer than the other; the tentacles situated one on each side of the largest are likewise 

 the smallest of the twelve. Both P. clamria and P. furcata exhibit this bilateral arrangement, 

 but in such minute polyps the differences are not so decided as on the larger polyps of Madnpora^ 

 and are not obvious on all the i)olyps of a colony. In P. astrseoides the twelve tentacles are 

 usually equal in size. 



It is shown later (p. 431), that this regular variation in the size of the tentacles of Madrepora 

 and Por'dcH is to be explained as the retention in the adult of a well-known larval stage passed 

 through in the development of the tentacles of certain Actiniaria, and is also associated with a 

 primitive condition of the internal mesenteries. 



TENTACLES OF SIDERASTR^A AND AGARICIA. 



The tentacles on the polyps of the genus SideraMrxa are so small as to be scarcely distin- 

 guishable with the naked eye, especially when retracted; but by careful examination with a lens 

 their disposition and character can be made out. Observations have been made upon the organs 

 in both S. rad'uDtx and S. Kiderea. Instead of being closely arranged in a narrow^ peripheral 

 zone, as in most corals and anemones, the individual tentacles arc widelj' separated from one 



