376 



THE FURTHER DIFFERENTIATION OF THE 



by the following experiment. If a region of the spinal cord corre- 

 sponding to trunk- segments 1-3 is removed and grafted into the 

 spinal cord of another embryo in place of the region of segments 

 4-6, the amount of neuron proliferation shown by it remains 

 roughly the same as it would have exhibited in its normal position.^ 

 Other regions, however, show dependent differentiation in the 

 proliferation of neurons. The spinal cord of a newt tapers from front 

 to back, which means that the tube in the region of the more 

 posterior segments of the body contains fewer neurons and axons 



Fig. 181 



Diagram to show the operation of exchanging a region of the spinal cord of 

 Amhlystoma corresponding to segments 3-5 for a region corresponding to seg- 

 ments 7-9. (From Detwiler, Natunviss. xv, 1927.) 



than do the anterior regions. If a region of the spinal cord corre- 

 sponding to trunk-segments 3, 4 and 5 is cut out, rotated about its 

 long axis, and planted back again in the order 5, 4 and 3,^ the 

 result is the normal differentiation of the spinal cord, with the 

 proper taper and the proper number of neurons and axons in the 

 various regions. The same is true if a region of the cord corre- 

 sponding to trunk-segments 7, 8 and 9 is grafted in place of the 

 region of segments 3, 4 and 5^. (fig. 181). It is clear, therefore, that 

 the factors which govern the proliferation of neurons in these regions 

 of the spinal cord reside elsewhere. In other words, while the spinal 

 cord is qualitatively self-differentiating, it is still dependent- 



^ Detwiler, 1925 b, 1928 a. - Detwiler, 1923 b. 



^ Detwiler, 1923 a. 



