Herrick, Morphology of Brain of Bony Fishes. 63 



gions of this segment, i. e. they are switch cells or reflex sta- 

 tions. In this case we are certainly dealing primarily with sen- 

 sory fibres, though some of the processes given off may pass to 

 the motor peduncles. These cells are most numerous near the 

 border but are scattered throughout the nidulus. They vary 

 greatly in form and size, but do not become small enough to be 

 confused with the second type. In some cases they are pale 

 and granular instead of deeply stained and rather homogeneous 

 and in such cases it might be supposed that we have a transition 

 to the splotches which give the characteristic appearance to the 

 ruber. This is not the case, however, for these bodies are larger 

 and tend to assume an irregularly polygonal or subspherical 

 shape. The most common cellular elements are scattered in 

 great numbers throughout the whole mass. They have the ap- 

 pearance of the cells of the cerebrum, but vary greatly in size. 

 They are fusiform, with a strong ape.K process and large, clear 

 nucleus and simple nucleolus. These cells lie in all directions 

 and their nervous processes make up a large part of the reticu- 

 lum of the body and radiate from it in various more or less defi- 

 nite tracts. One is struck by the fact that these cells vary 

 greatly m the intensity of stain and that they tend to aggregate 

 in clusters with the fibres in the same axis. Using higher pow- 

 ers and becoming aware of the faintness of the stain, in many 

 cases the number of such cells continues to increase until there 

 remain only a few very dense masses which can still be regarded 

 as the representatives of the opaque areas, which are so conspic- 

 uous under lower powers. Addressing ourselves to these bodies 

 it is possible to show that these also consist of a cluster of the 

 cells above mentioned, compacted into a mass and ensheathed 

 with neuroglia. In no case which we have examined, where the 

 conservation and stainmg were perfect, have we failed to demon- 

 strate these cells with nuclei and processes. The color is very 

 faint and we suppose that the staining portion of the protoplasm 

 has been partly extracted and suffused through the cluster. It 

 might also be supposed that these cells, by reason of their com- 

 pactness, are less accessible to preservatives and stain and there- 

 fore shrink more and stain less. Such large vacuoles, as de- 

 scribed by Sanders, do not develop when properly treated with 

 chrom-acetic, which seems to be a much better reagent for this 



