xvi ECHINODEEMATA 461 



and becomes flattened on one side, which we shall term the posterior 

 side of the larva. In the middle of this flattened surface an invagina- 

 tion makes its appearance which is the beginning of the archenteron. 

 This invagination is of small diameter compared with the diameter 

 of the larva, and, in contradistinction to all the gastrulae so far 

 studied, a wide space, the primary body-cavity or blastocoele, 

 intervenes between endoderm and ectoderm. In most gastrulae a 

 slit-like blastocoele is present, but in Asterias it is enormous. 



As the invagination progresses or, according to Field, from 

 its very beginning cells are budded off from the invaginating 

 surface into the blastocoele. These cells are termed mesenchyme. 

 The " wandering " of these cells seems to be effected by their emitting 

 long filamentous pseudopodia which span the blastocoele, and along 

 these strands the body of the cell glides like a drop of dew on a spider's 

 web. From the mesenchyme an exceedingly delicate gelatinous 

 connective tissue is formed, since the mesenchyme cells and their 

 pseudopodia secrete a few intercrossing fibres ; but the fluid " ground- 

 substance," Which from the beginning has occupied the cavity of 

 this blastocoele, forms the great mass of the " tissue " until the 

 completion of metamorphosis. 



As the invagination proceeds the gastrula grows in length, 

 changing its shape from a hemispherical to a cylindrical form, and 

 when the archenteron has attained about two-thirds the length of 

 the larva, it develops at its end a thin-walled vesicle, and the process 

 of gastrulation may be said to be complete. From this vesicle 

 mesenchyme cells continue to be budded off. The ectoderm cells at 

 the anterior pole of the larva become rather more columnar than 

 elsewhere (ap, Fig. 356, C), and bear longer cilia. This thickening 

 we may regard as a rudimentary sensory apical plate, but no nerve- 

 fibres have as yet been detected at the base of these cells. 



In Asterias rubens, according to Gemmill, the formation of 

 mesenchyme does not begin so early as in Asterias vulgaris, no 

 mesenchyme at all being formed in the British species until invagina- 

 tion is well advanced. 



From the vesicle at the apex of the archenteron two lateral 

 pockets grow out (Fig. 356, D). These are the rudiments of the 

 coelomic sacs, and they soon become completely cut off from the 

 archenteron, which in this way becomes divided into coelom and 

 gut. It is not clear, from the accounts which we possess, whether 

 the two coelomic sacs are, or are not, at first united across the middle 

 line by the remains of the original vesicle ; the coelomic sacs certainly 

 are united in this way in the larvae of Ophiuroidea and Echinoidea. 

 From the walls of the coelomic sacs more mesenchyme cells are 

 given off. The pseudopodia of these cells, which, as we have seen, 

 span the blastocoele, become in many cases muscular, and confer on 

 the larva powers of bending and of changing its shape. 



The anus is nothing but the persistent opening of the blastopore, 

 but the mouth is the external opening of a wide funnel formed by an 



