OF ARTS AND SCIENCES, 103 



varies considerably both in the different regions of the same embryo 

 and in the corresponding regions of different stages, as may be seen 

 from the outlines (Pis. I. and II.) ; in some sections the space between 

 them is so narrow as to leave room for only one or two fibres. A few 

 nuclei are constantly found in the fibrous substance occupying this 

 intercolumnar space while there are as yet none in the columns them- 

 selves. 



Already at this stage there are indications of fibres running across 

 the median plane, just ventral to the posterior columns, and serving to 

 connect the posterior horns. These fibres are probably connective 

 tissue, and not nervous elements. 



In PI. III. fig. 4, fi'om the lumbar enlargement of the same embryo, 

 is shown the relation the nuclei bear to the posterior horn fibres. The 

 latter expand into a brushlike enlargement, one end of which is merged 

 in the cuticula interna, while the corresponding nuclei are situated at 

 a considerable distance from the cuticula. 



Cross sections through the posterior lumbar region of embryos 

 60 mm. long (PI. III. fig. 5) show a pair of cords (G), composed of 

 a coarse network of fibres, occupying the space between the posterior 

 columns. They are more firmly connected with the pia mater than is 

 the white substance of the columns, as is proved by the fact that they 

 maintain their connection with the pia, while the white substance is sepa- 

 rated from it by the shrinkage due to the hardening fluids. These cords 

 contain nuclei which have a granular appearance. I am inclined to 

 consider this apparently granular condition the result of the nuclear 

 metaraorpliosis which immediately precedes cell division, and conse- 

 quently that the cells composing this tissue are still undergoing rapid 

 proliferation. The columns diverge at their dorsal edges, embracing 

 between them a triangular space filled with a network of more delicate 

 fibres than those of the cords themselves. Undoubtedly these cords 

 are merely the result of a further modification of the cells embraced 

 between the columns in the preceding stages. The position of the 

 nuclei indicates at least that they are genetically connected with the 

 nervous cord rather than its pia-mater sheath. 



The connection between the posterior horns is very plainly marked 

 at this stage, the round homogeneous nuclei of the gray substance being 

 in striking contrast to the oblong reticulated nuclei of the epithelium. 

 In this stage nuclei are also found in the posterior columns ; they 

 make their appearance first near the deeper margin of the white sub- 

 stance, as may be seen in PI. III. fig. 5. 



Sections from the lateral columns (PI. III. fig. 7) indicate clearly, 



