SMITH : EYES OF I'ULMOXATE GASTEROPODS. 2G3 



borne out by the few longitudinal sections of this part of the cell which 

 I have seen. Moreover, the size of the meshes is visibly greater in 

 Figure 43, even though it represents the second section proximal to 

 that of Figure 42. As in vom Rath preparations, sections still more 

 proximal — those in the region of the nuclei (Fig. 44) — show that the 

 nuclei are more or less eccentric, so that the fibrillar network passes 

 down the cell on one side, — a condition which is maintained even 

 proximal to the nucleus (Fig. 50). 



In another respect Bethe's method supplements vom Kath's. Toluidin- 

 blue preparations decolored from all but a few fibrils do not exhibit that 

 confusion of fibrils which results from the use of vom Rath's fluid ; thus 

 it is easier to see in toluidin-blue preparations that the fibrils make 

 completely closed meshes and not a distally branching system. These 

 conditions can be observed in a study of some of the cells of the chief 

 retina (Fig. 53), but the cells of the accessory retina offer the more 

 favorable opportunity. Figures 52 and 55 are from the accessory retina, 



— the former a Bethe preparation, the latter a vom Rath preparation. 

 Comparison of the two will make my statement clear. The latter figure 

 has been described already. In Figure 52 a few loose meshes are to be 

 seen in the neurite process {pre. n't.), from which they are traceable 

 distally directly toward the nucleus. Here they unite with meshes on 

 one side of the nucleus. From these meshes fibrils pass toward the 

 periphery of the cell and also distad, probably on their way to the rod. 

 Thus the meshes in the neurite process of this cell correspond, except in 

 number, with those in the process of the cell shown in Figure 55, while 

 the networks in the vicinity of the two nuclei also correspond. The 

 evidence from the Bethe preparations supports the idea of a diffuse net- 

 work of neurofibrils in the cell body. 



Fibrillae by the Intra-vitam Method of Prentiss. Doubt as to the 

 reliability of Bethe's method as a specific neurofibrillar stain led me to 

 the use of methylen blue. This method has the advantage over vom 

 Rath's that only a few of the fibrils remain stained, precisely as in 

 Bethe's method. The method brings to view, in properly differentiated 

 cells, a network which repeats all of the essential features described for 

 the vom Rath preparations. Except for the number of fibrils, Figure 37 



— which is a section of a cell prepared by the method of Prentiss — 

 might pass for a cell prepared by the vom Rath method; e.g., the one 

 shown in Figure 5E The fibrils are fewer and they are less massed 

 than in the latter figure. The periphery of the cell is too far decolored 

 to show the fibrils distinctly. The larger of the two protrusions inconi- 



