464 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



appeared. To the left of ray V is ray I, and to the right ray IV. The left 

 dorsal ray is ray II, and the right dorsal ray ray III. 



Thus the pore lies between radii I and V, but very near the former. 



The five primary hydrocnele evaginations coincide with the radii I-V. 



The ciliated bands disappear immediately after the fixation of the larva, and 

 the ectodermal epithelium shows a uniform structure throughout. The first (most 

 anterior) incomplete band is the first to go, followed a few hours later by all the 

 others. The disappearance of the first band may even occur before fixation. 

 Occasionally free larvae are found lying on the bottom of the containers which 

 have lost all of the ciliated bands. But it is possible that these larvae are incapable 

 of further development, and in general it is correct to consider the disappearance 

 of the cilia as the result of attachment. 



It has already been mentioned how in the intermediate zones of the older free- 

 swimming larva the cells begin to secrete a homogeneous hyaline substance while 

 their protoplasmic content decreases in volume so that the epithelial layer appears 

 loose. After fixation the same process takes place in the ciliated bands, so that 

 very soon their borders are no longer demonstrable. 



When this process has been completed the whole ectoderm is composed of 

 detached spindle-shaped cells, considerably elongated, which for the most part 

 stand at right angles to the surface, in section resembling somewhat the palisades 

 of a wooden fence. In addition there are found other types of cells, some of 

 pyramidlike shape with broadly expanded bases turned either outward or inward 

 and long drawn out apices, and others more rounded, almost spherical. 



The spindle-shaped forms predominate in the anterior section of the column, 

 while the rounded forms are more commonly met with in the region of the hydro- 

 ccele ring. In the middle of the cells are the nuclei, which are sometimes seen 

 in process of division. 



In general cell division in the ectoderm is rare at this stage, and an appreciable 

 enlargement of the number of cells no longer occurs. It is possible that with the 

 increase in the intermediate substance in the ectoderm the cell borders entirely 

 disappear. 



Some of the ectoderm cells have withdrawn from the surface and have assumed 

 a subepithelial position, a process the commencement of which may be observed 

 in the older free-swimming larvae. In addition to these cells, which have passed 

 out of the ectoderm in their entiety, there are others, originally subepithelial in 

 position, which have arisen through the division of ectoderm cells parallel to the 

 body surface instead of at right angles to it. 



In the free-swimming larvse the ectodermal body epithelium is everywhere 

 sharply defined, and nowhere can there be a doubt whether a given element belongs 

 to the ectoderm or to the mesenchyme. But soon after attachment the relation 

 between the ectoderm and the inner elements undergoes an important change. 

 Simultaneously with the inward migration of numerous ectoderm cells, and the 

 giving off from them of processes directed into the body cavity, individual star- 



