Prentiss, Pcrip Jural Netzvorks. 105 



be discovered. In the palate of Nccturiis there was some evi- 

 dence of a fibrillar basket-work about the nuclei of certain cells 

 (Fig. 7). Only a few meshes were observed and these may have 

 been formed by the crossing of fibrillae; by far the greater num- 

 ber pass directly through the cell without branching, as may be 

 seen in the figure. I therefore maintain that these cells, if 

 nervous structures, are to be compared, not as Bethe would 

 have us, to the ganglion cells of invertebrates (in which a com- 

 plete basket work is formed by the fibrillae), but rather to the 

 central nerve cells of vertebrates, through which, as a rule, 

 most of the neurofibrillae pass entirely independent of each 

 other. 



In the palate of the frog the cells of the networks are usually 

 located centrally with reference to the fibrillae, the nucleus being 

 surrounded by the latter. It is difficult to see how such nuclei 

 can be interpreted as belonging to sheath cells. Often, how- 

 ever, a cell lies eccentric 10 the fiber, the fibrillae all passing 

 to one side of the nucleus; or two nuclei may be found in close 

 proximity to each other, the one surrounded by fibrillae, the 

 other being eccentric in position. The form of the 

 nuclei also varies with their location. If surrounded by 

 fibrillae the nucleus may be of spheroidal or pyramidal shape; 

 if lying eccentric they are usually flattened and elongate in form. 



I have described in some detail these networks of cells and 

 fibers both because my observations substantiate more or less 

 completely the results of Dogiel, Bethe, and Leontowitsch, 

 and in order that the significance of the following degeneration 

 experiments might be more easily understood. From the 

 histological evidence one must conclude that the non-medullat- 

 ed fibers of all these networks are true nervous structures, and 

 not connective tissue as has been maintained. They are com- 

 posed of neurofibrillae, are connected with meduUated fibers, 

 and are not demonstrated by specific connective tissue stains. 

 The presence of neurofibrillae about the nuclei of some of the 

 cells would indicate that they also are of nervous character, as 

 Bethe and Leonowitsch assert. The evidence as to the pres- 

 ence of neuro-fibrillar networks in these cells is not conclusive, 



