114 



they may be due partly to the influence of laboratory conditions. I have 

 not found the larvae in the free, so I cannot ascertain whether such varia- 

 tions occur also under natural conditions, but I should expect so. The 

 main thing is, however, that the metamorphosis goes on quite normally 

 in all the various forms, showing that the variations are of no real im- 

 portance to the larvae. All the arms are supported by a fenestrated skeletal 

 rod. There is no vibratile band differentiated, only a general ciliation all 

 over the body, by means of which the larva swims. Only in some ex- 

 ceptional cases a distinct ciliated band is formed (comp. p. 116). There is 

 no pigment. 



Turning now to the development of the internal organs the first thing 

 to arouse attention is the larval esophagus. In PI. XV, Figs. 19 22 a 

 series of sagittal sections through an embryo 18 20 hours old are repre- 

 sented. There is no connection between the lumen of the pharynx and 

 that of the entoderm, only their walls are closely apposed. From the base 

 of the pharynx a sac-shaped prolongation proceeds along the dorsal side 

 of the entoderm reaching about halfway backwards. In PI. XVI, Fig. 5, 

 representing a median sagittal section of an embryo 21 22 hours old, 

 this prolongation is seen to reach to the posterior end of the body, thus 

 covering the whole dorsal side of the entoderm. Corresponding series of 

 frontal and transverse sections of embryos of the same age (PI. XV, Figs. 

 15 18; PI. XVI, Figs. 11 13) show both the pharynx and its posterior 

 prolongation to be flattened; the latter is seen to be confined to the dorsal 

 side of the larval body, the ventral side of the entoderm remaining un- 

 covered. The ventral wall of the posterior prolongation is distinctly thick- 

 ened, the dorsal wall remaining thin. In a slightly older stage, PI. XVI, 

 Figs. 14 16, representing transverse sections of an embryo 23 24 hours 

 old, the ventral wall of the sac is considerably thickened and begins to 

 fold, the underlying hydrocoel wall following the folds. The dorsal wall of 

 the sac on the other hand becomes thinner and gradually assumes a per- 

 fectly endothelial character. Thus it is clear already at this stage, and is 

 further definitely proved by the following stages (PI. XVI), that this 

 posterior prolongation from the pharynx is the amnion, which 

 is accordingly formed here in a way quite unique among Echinoids, as 

 far as hitherto known. The pharynx itself gradually shortens and soon 

 completely disappears, as becomes evident from a comparison of PI. XV, 

 Figs. 15 18 and PI. XVI, Figs. 1 4, or still more so on comparing the 

 sagittal sections PL XV, Figs. 19-22 and PI. XVI, Figs. 58 and 10. 

 The larval mouth, which never assumes the normal functions of a 

 mouth, since the pharynx never opens into the stomach, the larva being 

 accordingly entirely unable to feed, thus directly becomes the open- 



