40 Hjalmar Théel, 



to be traced by the fact that the cells, where the band develops, present 

 themselves columnar in profile and, when viewed from the surface, some- 

 what smaller and polygonal from mutual compression. Of course, those 

 cells still remain highest, which originally belonged to the animal pole 

 of the Gastrula but have directly passed into the cihated band. 



In the meantime, great changes have taken place with regard to 

 the shape of the body. The dorsal surface has been highly con vexed, 

 which is also the case with the under part of the ventral, while the 

 rest of that surface has become strongly concave. Thus, two areas are 

 to be observed on the ventral surface, of which the anterior concave one 

 is bounded by the ciliated band. The whole animal gradually assumes 

 the shape of a bell turned upside down, which becomes more conspi- 

 cuous in somewhat older larvae. The cells at 'the original vegetative 

 pole have passed into the entodermic sack during the invagination, and 

 the blastopore itself has gradually changed place in front on the convex 

 part of the ventral surface, and consequent!}^ the posterior end of the 

 Pluteus in no respects corresponds to the vegetative pole of the 

 Gastrula. 



In the meantime the first formed pair of calcareous stars have 

 increased and changed considerably, their three processes or rather 

 rods having lengthened obviously and obtained spines or small bran- 

 ches. One of the rods, PL V, fig. 82 — 85 c, runs on the central 

 surface of the larva towards its middle-line, crossing the corresponding- 

 rod of .the opposite, star just at the transverse portion of the cihated 

 band. Often they not only cross but also coalesce. The second rod, 

 PI. V, fig. 82 — 85 6, passes along the ventral surface towards the poste- 

 rior end of the larva, and the third runs towards the dorsal surface, PL 

 F, fig. 83 f, and divides into two, of which one branch, PL V, fig. 

 82— '85 e, extends forwards towards the anterior lobe of the larva in 

 order to support subsequently the anterior ventral arms, while the other 

 passes towards the posterior end of the larva and there obtains consi- 

 derable branches, which meet those from the posterior ventral rod, PL 

 F, fig. 82 — 85 d. However, the arrangement of these latter branches 

 is subject to variations of more or less importance, which may be best 

 understood from the figures. Lastly we have to pay attention to the 

 latticed rods, which support the posterior ventral arms of the larva, 

 PL F, fig. 82 — 85 a. As before mentioned, they begin to arise du- 

 ring the gastrula stage as three small processes, one on each rod_ of 

 the star close to its centre, PL IIL fig. 38. These processes stretch in 



