THE 8TAEFISH SOLASTEE ENDECA. 25 



bending the uppermost arms and the body outwards on the surface of the water, as 

 grown-up Starfish do when seeking a new attachment. It is difficult, however, to see 

 how a young Solaster in its natural surroundings could ever have the opportunity 

 of practising the trick of surface-floating which I have described above. The same 

 faculty has been observed in other very young Echinoderms. 



Chronology. — The following details regarding time supplement those given on 

 pp. 13, 17:— 



4-5 weeks Larval sucker no longer functional; attachment by the first-formed tube- 

 feet. 



6-7 „ Formation of mouth; formation of radial pouch IX of the hydrocoele. 



7-8 „ Formation of anus ; appearance of aboral arm-rudiments 8 and 9 ; 



diameter of disc and arms about 1*8 mm. 



12 „ Three pairs of sucker-feet on each ray; arm-rudiments 8 and 9 still a 



little smaller than the rest ; diameter about 2*2 mm. 



5 months Five pairs of sucker-feet ; arms all alike ; diameter about 2"6 mm. 



7 „ A sixth pair of sucker-feet beginniug to appear ; diameter 3 mm. 



VI. Development of Cavities and Oegans. (Pis. III.-V. figs. 24-56.) 



The mode of formation of the gastrula has already been described. At the end of 

 gastrulation the archenteron is a dome-shaped cavity with a uniform lining of columnar 

 ciliated cells. For the next few days this cavity follows in general the shape of the 

 growing larva (see p. 14), becoming elongated as the larva lengthens, sending out 

 hollow processes into the arms as they develop, and flattening from side to side in its 

 posterior half like what we have called the body of the larva. At the same time the 

 cavity becomes gradually marked out into three regions, although there is no external 

 indication that such a change is taking place. The regions are respectively anterior, 

 middle, and posterior, the anterior being much the largest, while the posterior opens 

 on the external surface by the blastopore (PI. IV. fig. 36). The middle region is the 

 smallest and is much compressed from side to side. It is at first ill-defined, appearing 

 rather as an isthmus between the other two than as a distinct region. Its mode of 

 separation is described in detail under the heading " Enteron " on p. 37. The division is 

 completed by the tenth day. The first and third of the chambers constitute respec- 

 tively the anterior and posterior cceloms of the Solaster larva, while the middle one 

 becomes the enteron or gut. 



In the further description under this heading I shall give as far as possible a con- 

 secutive history of each of these cavities and its derivatives, both in the larval condition 

 and throughout metamorphosis. It is hoped that such difficulties as may arise through 

 having to speak of the later development of one set of structures much in advance of 

 others may be minimised by the cross-references throughout the text, as well as by 

 the illustrations, and the general summary on pp. 50-52. 



VOL. XX. — pabt I. No. 4. — February, 1912. e 



