242 NIKOLAS KLEINENBERG. 



^ /nd thus begins the repetition of the successive variations 

 just described. 



According to these observations the mode of development 

 of the ganglionic chain would be the following : — Some of 

 the ectoderm cells, situated on the two sides of the ciliated 

 furrow, divide and form two parallel thickenings. One, or 

 sometimes two or three, cells, of the newly formed deep layer 

 acquire special characters, and separate themselves distinctly 

 from the superficial layer and from the lateral parts of the 

 ectoderm from which they originated. In this way are 

 developed two cords, completely separated from each other 

 by the cells of the ventral furrow. This is the original 

 double rudiment of the subintestinal central nervous appa- 

 ratus. Then the neighbouring margins of the cords raise 

 themselves, and approach each other, forming between them 

 a groove, sometimes very deep, but having only a temporary 

 existence. The upper cells of the neighbouring sides force 

 themselves above the groove towards the median line, where 

 they meet and unite with each other ; their number increasing 

 the groove becomes little by little flattened out and finally dis- 

 appears. The primitive cords are thus united into a median 

 plate. As soon as the union has taken place, the cells 

 group themselves into a series of swellings and constrictions. 



The first represent the ganglia, the second the connecting 

 trunks. Now, certain cells placed beneath the dorsal surface 

 on each side are transformed into an ill-defined granular 

 substance, Avhich gradually extends to the median line and 

 forms the fibrous commissural cord. This developes sepa- 

 rately for each segment of the chain, before the single 

 segments become united among themselves by a special con- 

 ducting tissue. The connecting trunks, which run through 

 the whole length of the ganglionic chain, are formed later, 

 simply by the fusion of the commissural trunks of the suc- 

 cessive segments ; hence the first rudiments play the part of 

 a common foundation for the transverse and longitudinal 

 commissures, in which afterwards, by the development of 

 Lerve fibres, a regular apparatus>,of conducting threads is 

 established. In consequence of the formation of the fibril- 

 lar substance, the parts of the nervous plate, which are 

 destined to become changed into ganglion cells, extend on 

 the sides and ventral surface in a more or less thickened, 

 but everywhere continuous, layer. 



At intervals, thickenings of this cellular covering pene- 

 trate deeply into the interior of the plate. They are formed 

 in part by the central cells, which are' not transformed into 

 the substance of the commissure, in part by cells, which 



