THE PRIMARY SEGMENTATION. 123 



vertebrates contained an unsegmented anterior part, and the 

 question, How far forward does the segmentation extend ? 

 is a very important one. 



The evidence to elucidate this point has been accumulating, 

 and I think we are now in a position to answer that fundamen- 

 tal question. Already the olfactory nerves have been shown 

 to have a similar history to the cranial nerves, and, to all 

 appearances, the optic nerves are soon to be included with the 

 others. The Director of this laboratory has recently shown 

 conclusively that the head of annelids is metameric throughout. 

 The cerebral ganglion or brain of these animals is segmented 

 in the same manner as the ventral nerve-cord, and consequently 

 there is no non-metameric part of the nervous system, as has 

 been so long assumed. This must necessarily change the views 

 regarding the ancestral derivation of the vertebrate head. 

 Waters, Zimmerman, and others have shown the existence of 

 segments in the fore-brain of vertebrates, and further evidence 

 on that point will be brought out in this lecture. 



The recent endeavors to solve the problem of metamerism 

 of the head are based on observations on cranial nerves and 

 branchial clefts, mesoblastic head cavities and neural segments. 

 These are the cephalic structures that exhibit segmental ar- 

 rangement, and we must depend upon them for evidence. It 

 is an open question to which of these the most importance is 

 to be attached. The first-mentioned basis, namely, cranial 

 nerves and branchial clefts, is the least favorable, as it involves 

 too much conjecture. Dr. Strong, of Columbia University, has 

 shown that the cranial nerves do not appear in the positions 

 they come to occupy, and McClure has stated the case against 

 the cranial nerves as follows : "We have positive proof that 

 the degeneration of certain branches has taken place. This 

 being the case, we have every reason to assume that whole 

 segmental nerves may have once existed, which have completely 

 degenerated, leaving no trace whatever of their previous exist- 

 ence. If such be the case, the segments originally connected 

 with these degenerated nerves must necessarily be overlooked, 

 if the existing nerves are made use of as a means of determin- 

 ing the original number of segments. 



