478 ADAM SEDGWICK. 



first and second somites with the inner portion of the cerebral 

 rudiment, which is a development of the whole surface of the 

 pre-oral part of the lateral thickening. It is true that at first 

 this part of the central nervous system is weaker than the parts 

 in front and behind it, but long before the separation from the 

 ectoderm occurs (PI. XXXV, fig. 19 6; PI. XXXVI, fig. 35,c.o.n.), 

 it forms a well-marked cord with a layer of white matter. 

 The whole central nervous system of Peripatus 

 capensis develops, therefore, as a continuous struc- 

 ture from the ectoderm, and the independent origin and 

 secondary connection of the cerebral ganglia and ventral 

 chain, which has been asserted for some Arthropoda, e.g. 

 for Spiders by Balfour (No. 3), for Annelida by Salensky 

 (No. 27), Lumbricus by Kleinenberg (No. 15), and for Mol- 

 lusca by various observers, does not hold for Peripatus.^ 

 Balfour held the same view with regard to this point in Peri- 

 patus (No. 2 p. 337). 



I entirely agree with the remarks of Hattchek (No. 12, p. 8) 

 on this subject. He holds, in opposition to Salensky, that the 

 circumoral part of the nervous system in the annelid larvse 

 which he has investigated, develops fron the ectoderm in con- 

 tinuity with the apical ganglion and ventral nerve-cords ; and 

 I am strongly inclined to think that a further and closer inves- 

 tigation will show the same fact to hold for other Annelids and 

 Molluscs. 



The Cerebral Grooves. — The cerebral ganglia gradually in- 

 crease in size (Pis. XXXIV, XXXV, figs. 14, 16 a, 19 a) and 

 their ventral surface becomes in Stage e markedly flattened (PI. 

 XXXV, fig. 22 a) and then invaginated (Part I, fig. 35), so as 

 to form two grooves, which rapidly become narrower and deeper, 

 until in Stage f they form two slits, longitudinally arranged, 

 ending blindly in front, but opening behind into the buccal 

 cavity (Part I, fig. 36, and PL XXXVI, fig. 33). Eventually, 

 in old embryos of Stage f, they lose their external openings, 

 become reduced in size, and form two vesicles in the ventral 



1 Kenuel apparently holds the opposite view ou this point. If he is right 

 I can only suppose that P. Edwardsii differs in ihis respect I'rom capensis. 



