NUCLEAR CHANGES IN RANA CLAMITANS. 1 27 



Three hours after the operation there are fewer of the angular 

 nuclei than at one hour and more of the round deeply-staining 

 nuclei. The latter vary from the size of similar ones in the 

 earlier stages down to fragments. Moreover some of the larger 

 of these seem to be in the process of fragmentation, that is, ap- 

 pearances indicating stages in direct division are seen. The 

 gradation in size and depth of stain at one hour from normal 

 nuclei nearly to rounded ones, and the gradation down to frag- 

 ments at three hours, as well as the appearances of fragmentation, 

 make it fairly clear that normal nuclei just in front of the cut 

 edge may contract, become rounded, and fragment. This must 

 be a degenerative process. Even finer intermediate steps are 

 seen in preparations of later stages. 



Sections of one individual at this period appear very much 

 like those immediately after the operation. The deeply-stained 

 nuclei are similar, and the spinal cord is not covered either by 

 epidermis or plasma, so that a recent outbreak of the cerebro- 

 spinal fluid must have taken place. In this case a second contact 

 with the exterior has again started the degenerative process. 



At five and a half hours the spinal cord is entirely covered by 

 the thickened plasma layer, in which is a group of fragmenting 

 globular nuclei. In one preparation at this time, the epidermis 

 has closed-in over the entire wound, and there is a series of stages 

 in the degeneration of nuclei. Some are only slightly smaller and 

 darker than normal nuclei; others have the angular appearance 

 characteristic of nuclei one hour after the operation, while still 

 others are round and fragmenting. At this stage there is another 

 evidence of the presence of a cerebrospinal fluid. The plasma 

 covering the end of the spinal cord is pushed outward, making a 

 knob-like extension of the central canal similar to that shown in 

 Fig. 8. This did not appear in earlier stages either because not 

 enough cerebrospinal fluid was present, or because the plasma 

 layer had not coagulated sufficiently to resist the outward 

 pressure of this fluid. 



Of the two preparations of tadpoles killed after a nine and a 

 half hour interval, one shows the epidermis and plasma covering 

 all the wound except the neural tube; the other shows this part 

 also covered. In the former, the sides of the neural tube are 



