MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 529 



of the expanded J^olyp as opaque white bodies, moving freelj' about in the gastro-ccelomic and 

 tentacular cavities; upon retraction individual larviv often remained within the tentacular cavity, 

 distending it and forming small protuberances on the surface of the colon3^ The usual occurrence 

 of the larvffi within the tentacles woidd indicate that they made their exit through the tips of 

 these organs, but although large numbers escaped while the colonies were "nder observation 

 the actual point of extrusion was never determined. They were either shot out suddenly, 

 with force enough to send them some distance, or merely escaped and fell on the general surface 

 of the parent colonv. '' 



Some of the larvaj were able to swim about immediately on extrusion; others remained 

 motionless for a few moments either on the surface of the water or the bottom of the vessel, and 

 then commenced vigorous gyratory movements. The rotation was clock-wise when the lar^-aj 

 wei'e viewed with the narrow oral extremity upward. When first liberated, the larvaj show 

 considerable power of adhesion at any part of their surface; on transferring them from one 

 vessel to another they would often tix themselves within the pipette, and recpiire a considerable 

 force of water to dislodge them. Specimens might adhere either by their anterior or posterior 

 extremit}^ for a time, and then commence moving again. 



When first expelled some of the larvw were rod-shaped bodies, about 2 mm. in length, and 

 rounded at each end; others were pear-shaped, the broader pole being directed forward and the 

 narrow end backward in translation; others again were oval or spherodial. The individuals, 

 however, were able to change from one form to another. Seen with the naked eye, or by means 

 of a lens, the larva? were sti'ongly opaque, an internal yellowish mass 1)eing distinguished from 

 a colorless or slightlj^ green external layer. The narrow, posterior, oral pole was for a long time 

 more darkly colored than the rest of the larva. 



Under the microscope also the larvie were perfectly opaque, the ciliation was uniform, and 

 when first extruded no oral aperture nor mesenterial divisions could be discerned. The surface 

 appeared minutely granular, white dots being irregularly distributed over the ectoderm. The 

 denser coloration at the narrow oral extremitj' was seen to be produced by an accunuilation of 

 yellow cells within the ectoderm, which gradually diminished in number away from the extremity. 

 It was possible to determine that the coloration of the internal endoderm was also due to the 

 presence of zooxanthelhe. Soon after liberation, extrusions of yolk granules and zooxanthellte 

 from the oral extremity took place, and continued from time to time (fig. 96). 



Many of the larvie underwent nmch alteration in shape. The oral extremity became swollen, 

 and the aboral narrow, a reversal of the primary condition (fig. 100); ))ut when first extruded 

 the larvffi sometimes exhibited the swollen oral extremity and narrow aboral. 



Within a day or two certain of the larva? had settled to the sides of the vessel, becoming 

 flattened both orally and aboral ly; a few zooxanthellaj were still present around the oral 

 aperture, which had now become functional. Some specimens would again detach themselves 

 and move slowly around. After fixation the larvse were more transparent, and at first four 

 pairs of mesenterial divisions were visil)le from the outside, and later six pairs. The settled 

 larvae were soon able to extend themselves, and assume the coluninai' form, appearing greenish 

 in color. 



The various stages of mesenterial development were quickly passed through, until all the 

 protocnemes were present, the Edwardsian mesenteries complete and the fifth and sixth pairs 

 incomplete. Beyond this no increase in the number of mesenteries took place during a period 

 of three weeks. Within four days six tentacular prominences were apparent, the larvic at this 

 stage usually appearing flask-shaped, with a broad base (fig. 107). During the early stages the 

 tentacles often became involved in the expanded discal tissues of which they were outgrowths, 

 and as a consequence were indistinguishable as separate organs. In a young polyp from another 

 batch of larva?, the six memlters of the inner entoco?lic cycle also appeared in advance of the 

 members of the outer exoccelic cycle (fig. 106). 



One larva was secured attached to a fragment of glass, and could thus be examined as a 

 transparent object, and its later development observed. In seven or eight days the mesenterial 

 filameuts were visible on the first and second pairs of mesenteries as darker internal organs, and 

 in about fourteen days six small, clear, oval areas were recognizable within the entocoele of the 



