OP ARTS AND SCIENCES. 105 



The growth of the Burdach'sche Keilstrange towards the central 

 canal does nut take place with uniform rapidity in all parts of the cord, 

 neither is there an uninterrupted gradation in the advancement of this 

 process from one end of the cord to the other ; they grow much more 

 rapidly in the cervical and lumbar enhirgements than elsewhere. 



There seems to be no causal connection between the closure of the 

 canal and the development of the posterior fissure, since the dorso- 

 ventral diameter of the canal is very frequently reduced to nearly its 

 minimum length before the white substance of the posterior columns, 

 on the morphological condition of which the presence or absence of the 

 fissure depends, begins to grow veutrahvards. 



In embryos measuring 125 mm. and upwards the canal is so reduced 

 as to appear in cross sections like a round or slightly oval opening, but 

 in the dorsal and sacral regions of many of these embryos there is 

 scarcely a trace of an ingrowth on the part of the posterior columns. 

 (Compare PI. II. figs. 30, 32.) Later stages in the development are 

 shown in PI. II. figs. 33-36. 



Until the embryos have grown to the length of 186 mm., the latest 

 stage I have sectioned, the posterior fissure remains closed, i. e. the 

 walls of the GoU'sche Keilstrange remain close together, being only 

 separated by the horn fibres which completely fill the space between 

 tliein. In the dorsal and sacral regions, even at this advanced age, the 

 Keilstrange have made very little progress towards the canal. 



In no stage is there any evidence of a direct median coalescence of 

 the facing walls of the canal. If such fusion actually occurred, one 

 should find, in some sections at least, a keel-like blade of cuticula 

 projecting dorsally from the cuticula interna which lines the roof of 

 the canal at all stages ; or, in the event that the fusion were to occur 

 more promptly near the centre of the canal than along its doi'sal edge, 

 there might even be a vertical blade of the cuticular substance entirely 

 separated from the cuticula lining the roof of the persistent portion of 

 the canal. But such structures are never to be observed, the cuticular 

 lining of the roof of the canal always presenting an even curved out- 

 line, not only for the face which looks inward toward the lumen of the 

 canal, but also for that which is directed away from it. 



It is therefore evident that the lumen of the canal is restricted from 

 the dorsal towards the ventral side by a gradual and continuous pro- 

 cess, which causes a shifting of the epithelial cells and makes their long 

 axes pass from a direction perpendicular to the median plane into one 

 more nearly parallel with it, as the cells are successively brought into 

 the region of coalescence. 



