Echinoderm Larvm with Two Water-Vascular Systems. 331 



this pit becomes thickened and pressed against the hydroccele bud, whilst the 

 opening of the pit becomes narrowed and eventually closed : the pit is then 

 termed the amniotic sac and its roof the amnion. 



The complex structure consisting of the adpressed hydrocoele and floor of 

 the amniotic sac is known as the Echinus-rudiment, and the subsequent 

 development of the larva consists mainly in the growth and elaboration of the 

 Echinus-rudiment {Ech., fig. 5). 



Before, however, we deal with the further development of this structure, 

 we may glance at the changes which take place in other parts of the larva. 

 A fourth pair of arms which are termed the prae-oral arms grow out from 

 that portion of the ciliated band which extends between the bases of the two 

 antero-lateral arms. This pair of arms is extremely apt to be imperfectly 

 developed in larvte which have been artificially reared ; most frequently one 

 of the pair is longer than the other, but sometimes one and • sometimes even 

 both remain entirely undeveloped. They are supported by calcareous rods 

 which are outgrowths of a median Y-shaped ossicle the dorsal arch {d.a., 

 fig. 3), which lies above the oesophagus and which makes its appearance 

 about the 18th day, before indeed any trace of the arms themselves has 

 appeared. The stem of the Y extends backwards and develops into a 

 calcareous network whicli surrounds the madreporic pore and is the rudiment 

 of the madreporic plate or madreporite of the adult. From this plate 

 about the 30th day a knob sprouts out which becomes one of the adult 

 spines and is termed by me the madreporic spine {m. sp., fig. 22). 



About the 15th day the ciliated epaulettes make their appearance. 

 These rapidly develop into great crescentic lobes, horizontally placed, which 

 together almost encircle the larva. They bear numerous long cilia and in 

 later larval life constitute the most important locomotor organ of the larva. 

 They were originally loops of the ciliated band, the dorsal ciliated epaulettes 

 being formed from the loops between the antero-lateral and the postero-dorsal 

 arms, whilst the ventral epaulettes are formed from that part of the ciliated 

 band which stretches between the bases of the post- oral arms. These loops 

 become entirely separated from the ciliated band, but the breaches of 

 continuity in the band so caused quickly heal up (d.c.ep., v.c.ep., fig. 5). So 

 far as is yet known, ciliated epaulettes are confined to comparatively few 

 genera and are not by any means characteristic of all Echinoid larvae. 



We have mentioned above that the separation of the right anterior coelom 

 from the right posterior coelom is not completed before the 12th day. Aftei- it 

 is complete the right anterior coelom sends out from its posterior end a tongue 

 of cells which arches over the oesophagus towards the mid-dorsal line. From 

 the apex of this tongue a cellular bud is detached, which becomes hollow and 



