638 COMPARATIVE ANATOMY 



geny: the blastopore ('Urmund'), the neuropore (the Annelidan mouth), 

 and finally the definitive vertebrate mouth." 



In deriving the neural tube of vertebrates from a part of the alimentary 

 canal of pre-chordates, Delsman's theory may appear to resemble that of 

 Gaskell, which derives the ependymal lining of the neural tube from the 

 intestine of arthropods. The similarity, however, is superficial. Gas- 

 kell's theory assumes that the entire alimentary canal of the arthropod 

 becomes the lining of the neural tube, so that an endodermal structure 

 is converted into an ectodermal one. Delsman's hypothesis holds that 

 only a part, and that an ectodermal part, of the invertebrate intestine 

 becomes the chordate neural tube. 



The conditions under which an ectodermal stomodeum might be 

 converted into a neural tube have been stated by H. E. Ziegler (1908). 

 If we start with a gastrula, he says, we must assume that it was primarily 

 nourished through the blastopore. The ciliated dorsal or neural plate 

 swept the food towards the blastopore, and consequently might have 

 acquired the function of a sensory epithelium. As the sensory plate 

 was converted into a neural tube, the stream of water entered the neuro- 

 pore and was carried through the neurenteric canal to the enteron. Then 

 came the formation of an anus, and later that of the definitive mouth and 

 gill slits. The neuropore and neurenteric canal became superfluous and 

 disappeared. Thus the neural tube, which had acquired a sensory func- 

 tion, became exclusively nervous and served as the central nerve cord. 



In support of this speculation of Ziegler, Delsman calls attention to 

 the fact that in some gastropods the buccal ganglia of the nervous system 

 are derived as a proliferation of the stomodeum. 



Others also have been impressed by the evidence that the neural 

 tube was formerly a part of the digestive system. Huntsman (1913) 

 writes, "It may be noticed that the condition in the tunicate strongly 

 supports the view that the neural tube was originally not nervous but a 

 part of the digestive system." 



In agreement with most supporters of the annelid hypothesis, Delsman 

 compares the ventral side of the annelid with the dorsal side of the chor- 

 date. But, if the stomodeum of the annelid forms the nerve cord of 

 chordates, then what becomes of the chain of nerve ganglia of the annelid 

 when transformed into a chordate? According to Delsman, the annelid 

 ganglia are represented in the dorsal ganglia of vertebrates. These are 

 formed, as would be expected, between the stomodeum and the skin in 

 annelids, and in chordates from the neural crest. The fact that the 

 ganglia of chordates are purely sensory, while those of annelids contain 

 motor cells, is an objection to this homology which Delsman has attempted 

 to meet. He points out that dorsal nerves in chordates are primarily 

 mixed in function, and that the motor cells of the sympathetic ganglia are 



