40 DE. J. P. GEMMILL ON THE DEVELOPMENT OF 



Paired Badial Cceca. — The paired radial diverticula appear as folds on the roof of 

 the stomach which converge towards a central point, but, owing to the positions of the 

 anus and the axial complex, they are not disposed in a perfectly symmetrical manner. 

 The outer extremities of the folds, which, to begin with, fall some distance short 

 of the ends of the stomach-lobes mentioned above, become elevated from the surface 

 and grow out as blind pouches into the rays. Closure of the lips of each fold in 

 its outer part now occurs. The stalks of the pouches are lengthened thereby, and 

 their openings are all brought together into a small chamber — the pyloric sac — in 

 the roof the stomach. This chamber is U-shaped, the concavity of the U being 

 occupied by the anus (PI. I. fig. 11). The position of the madreporic interradius is 

 marked by an indentation near the middle of the convex side of the U. This 

 indentation causes the roots of the radial diverticula to fall into two groups, one 

 consisting of four pouches belonging to rays II to V inclusive, and the other of five 

 pouches belonging to the rest of the rays, viz. VI to IX and I. Fig. 11 illustrates 

 the relation of the roots of the various pouches to one another in a Solaster aged six 

 months. These relations remain practically the same in the adult, if one allows for 

 differences in the proportionate size of parts. 



During the fourth month the radial diverticula begin to broaden at their outer 

 ends. This is followed by bifurcation at these ends, with accompanying division 

 of their suspensory mesenteries and of the pockets of epigastric coelom contained 

 therein (p. 34). 



Interradial or Bectal Cceca. — The so-called rectal caeca are late in development. 

 They may, however, be recognised during the fifth month. In the six months' 

 specimen from which fig. 11 was taken they are represented by two outgrowths. 

 Of these the larger shows division into three lobes, and originates, as nearly as one 

 might judge, opposite interradius VI/VII. The other arises opposite interradius 

 II/III, and, though it has a somewhat crenated margin, is not yet divided into distinct 

 lobes. In the adult Solaster the caeca in question also come off by two roots occupying 

 positions similar to those just indicated. That which corresponds with the outgrowth 

 last described usually gives off five slightly lobulated caeca, 1 cm. in length, which lie 

 respectively in interradii IX/I, I/II, II/III, III/IV, IV/V, Avhile the other gives origin 

 to similar caeca for interradii V/VI, VI/VII, and VII/VIII. The remaining inter- 

 radius, namely VIII/IX, is oftenest left unprovided with a caecum. Apparently the 

 larger of the roots in the young Solaster is afterwards outstripped in growth by the 

 smaller, since it is the latter which provides five out of the total of eight caeca. 



It may be of some theoretical interest to point out that the distribution of this set of 

 five caeca is such that they might be considered as representing a set belonging to the 

 five-rayed ancestor of Solaster, leaving the rest as interpolations, like the four later- 

 formed rays. However, it w^ould be manifestly unsafe to lay stress on the relations 

 of parts so variable in Starfish morphology as are the rectal caeca. Of much greater 



