422 BULLETIN 82, UNITED STATES NATIONAL. MUSEUM. 



closed over. In the center of the regenerating visceral mass is a small opaque 

 papilla, probably due to the thickening of the epidermis preparatory to the forma- 

 tion of a month, but unfortunately Dendj^'s sections were not sufficiently well pre- 

 served to decide the point. 



Sections sliow that regeneration has proceeded as far as the formation of a 

 cushion of loose connective tissue, much thicker than at two days, and covered by 

 the deeply staining epidermic laj'er before mentioned. The growth of connective 

 tissue is especially strong underneath the point where the mouth is subsequently 

 to be found, so that we have here a pillar of denser connective tissue reaching from 

 the top of the rosette to the roof of the visceral basin and thickest just beneath the 

 epidermal layer. 



The central plexus is very well shown; but in these sections and in those of 

 later date there is no means of telling how much of it, if anj', has regenerated and 

 how much was left after evisceration. 



The ambulacral system has begun to regenerate, as is shown by the presence 

 of new ambulacral pores. On examining one of these more closely it is seen to be 

 vesicular, lined with columnar epithelium and opening to the outside; no internal 

 opening is yet visible. 



Another specimen supposedly, like the preceding, of three days' growth, is 

 much more advanced, the mouth and alimentary canal being present. In other 

 respects it is similar. In the center of the visceral mass is a depression; the 

 epidermis covering the depressed area is thickened, and in the center is a small 

 mouth leading obliquely downward into a feebly developed alimentary canal. The 

 alimentary canal has already the characteristic form found in normal specimens, 

 taking one turn in a horizontal plane and ending in an interradial anus. The 

 most remarkable feature about it is the rudimentary condition of its walls. In 

 histological characters the walls of the alimentary canal are almost indistinguish- 

 able from the roof of the visceral basin, and to all appearances have been formed 

 by an invagination of this roof. 



The roof of the visceral basin is composed of two layers — externally a thin 

 layer of minute, rounded, deeply staining cells, which are probably epidermal, and 

 beneath this a thicker layer of connective tissue cells, which stain less deeply. 



In the floor of the central depression already mentioned the epidermal cells 

 show a slight tendency to become columnar. The walls of the alimentary canal are 

 composed of exactly the same two layers — a layer of more deeply staining cells 

 which corresponds to the epidermal layer of the roof, but is now, of course, on the 

 inside, and outside this a layer of connective tissue. The cells of the inner layer 

 show a tendency to become columnar, and this layer is more distinct than the 

 epidermal laj^er of the roof. The lumen of the alimentary canal is very small and 

 its walls are very thin and but slightly folded. There is no trace of an anal cone ; 

 the anus itself is represented by a minute perforation in the roof of the visceral 

 basin. The walls of the last part of the intestine just before reaching the anus are 

 quite indistinguishable from this roof. There is no decisive evidence as to the 

 manner in which the alimentary canal is formed ; but there appears to Dendy to be 

 a strong probability in favor of the view that it is formed by invagination, the 



