THE DEVELOPMENT OF THE EARTH-WORM. 23f 



and above a thick layer of ganglionic cells remains. As the 

 extremities of the arch descend^ these commissural cords 

 lengthen proportionately, and the ganglionic cells on their 

 external sides become scarcer, so that it may be said that the 

 CBSophagus is not embraced by the entire ganglion, but rather 

 by the elongated branches of the commissure. These 

 branches must themselves descend to the ventral wall to 

 meet the first ganglion of the ventral chain, for 1 have never 

 seen prolongations directed upwards from the latter. But 

 the investigation of this point is extremely difficulty since 

 the lateral parts of the collar are very closely enveloped by 

 the mesoderm, whose cells resemble so closely those of the 

 nervous ring that it is not easy to distinguish them with 

 exactness. Hence I cannot say definitely that mesoderm 

 cells do not at this time enter the lower extremities of the 

 collar (this applies only to the lower extremities, since all the 

 remainder is clearly separated from the middle layer) to take 

 part in the formation of the commissure, but it would be 

 still less possible to prove that they do so, and I think it is 

 most improbable. When, at a relatively very late period, the 

 definite union of the cerebral ganglion with the first ganglion 

 of the ventral chain takes place, this first ganglion, as well 

 as those following, possesses a well-developed commissural 

 trunk with which the cord from the cephalic ganglion 

 appears to be directly united. 



The Development of the Ventral Chain of Ganglia. 



I began the account of the development of the central 

 nervous apparatus with the cephalic ganglion, because, even if 

 it is not, as I believe, the first part formed, it certainly appears 

 at least contemporaneously with the earliest traces of the 

 ventral chain. It is known that the development of the 

 latter progresses from before backwards, but its first rudi- 

 ment extends rapidly along the whole length of the embryo, 

 as far as the caudal extremity. On the other hand, the 

 separation of the individual ganglia and their histological 

 development takes place gradually, and much later in the 

 posterior than in the anterior part ; while the first ganglia 

 have already attained a state of great perfection, those fur- 

 ther back exhibit all imaginable gradations, till we reach the 

 condition of the undifferentiated rudiment. Hence, for the 

 investigation of the first changes, it is best to take early em- 

 bryos ; for that of the following stages much older embryos 

 answer very well, because the most different stages of deve- 

 lopment, united by the minutest gradations, are found in a 

 single individual. 



VOL. XIX. NEW SER. Q 



