PART 5 A MONOGRAPH OF THE EXISTING CR1XOIDS 137 



by breaking through a hole in the egg membrane. The empty membranes may easily 

 be found in the dishes, and form very characteristic objects. 



The newly hatched embryo is uniformly ciliated, with a tuft of longer cilia at the 

 anterior end. In the course of the next day the ciliated bands, only 4 in number, be- 

 come differentiated, the yellow cells appear in considerable numbers, and the larva, 

 seen in diffused light, forms quite a striking object. The yellow cells are restricted to 

 the interstices between the ciliated bands which appear as broad white lines; only very 

 rarely may a yellow cell be seen lying in one of the bands. The vestilnilar invagination 

 appears after about 3 days as a wide shallow depression much like that in the other 

 species of the genus. There is no suctorial disk. The formation of the skeleton begins 

 at the age of 6 days. 



It was repeatedly observed that the embryos had considerable difficulty in ruptur- 

 ing the egg membrane. Mortensen supposed that the absence of motion in the water 

 in the dishes containing the developing eggs was the chief cause of this, and this sup- 

 position was strengthened by the accidental observation that by sucking the embryos 

 up with a pipette and then squirting them out again with some force they were greatly 

 aided in rupturing the membrane and becoming free. The embryos which did not 

 succeed in rupturing the egg membrane did not die at once but continued developing 

 within the egg, the vibratile bands and the yellow cells appearing at the same age as 

 in the free embryos. In such cases the embr3 r os thus reached the same stage within the 

 egg membrane as do normally the embryos of A. mediterranea and A. bifida before they 

 become free. If not liberated in the way described above they gradually died; but 

 some of them remained alive up to the age of 6 days. 



Although excellent cultures of the larvae were repeatedly obtained both in 1918 

 and 1919 Mortensen never succeeded in rearing them through the metamorphosis. At 

 about the tune of the commencement of the formation of the vestibular invagination 

 they became abnormal, with a median constriction, or with the anterior end swollen. 

 They might live for many days in this condition, but ultimately died without attaching 

 themselves. Only in relatively few cases did the vestibular invagination develop nor- 

 mally, but even these larvae did not attach themselves. The normal development of 

 the crinoid skeleton within the larva was not observed; in some cases a few calcareous 

 pieces were formed which could not with certainty be identified. 



As soon as the blastula has formed, after about 6 hours, the formation of the mesen- 

 chyme begins, and the blastocoele cavity is soon filled with a dense mass of mesenchyme 

 cells. The gastrular invagination does not occur until a much later stage, about the 

 time when the embryo is about to leave the egg membrane, and the blastopore, which is 

 very small, docs not close until the embryo has become free. The iiivaginated portion 

 is not a simple sac, but, while still remaining in open connection with the exterior, is 

 IhiUened in the longitudinal axis. Mortensen was unable to trace with certainty the 

 later subdivisions of the invaginated portion. He remarks; that he intentionally does 

 not refer to the, invaginated portion as the archenteron, as he is not at all sure that the 

 stomach is derived from this part. He believes that there are indications of more than 

 one invagination taking place, in which case probably the enterocoele vesicles alone 

 develop from the invagination described. The enterocoele vesicles do not alter their 

 position in quite the same way as in A. mediterranea and in the other crinoids thus far 

 studied; at the time of the formation of the vestibular invagination the two vesicles are 



