940 
ME. A. SANDERS ON THE ANATOMY OF THE 
0’004 millim. long by 0*003 millim. wide, and the nucleus is 0*002 millim. in diameter; 
the other is 0*0034 millim. in diameter with a nucleus of the same size as that of the 
one first mentioned. The cells of the larger size seem to be transitional forms to the 
Purkinje cells. The one figured (fig. 17, c.) measures 0*006 millim. long by 0*0054 
wide, the nucleus is 0*003 in diameter, and the nucleolus, which in this case is rather 
larger than is usual in this species of cell, measures nearly 0*001 millim. in diameter. 
These two kinds of cells are irregularly mingled with each other, but the smaller sort 
are more often found nearer the molecular layer and extend some distance into it. 
The basis of this granular layer contains a fine network of fibrillse (fig. 17, h.) which 
connects the cells together and extends into the molecular layer. 
The ridges are bounded on the other side by a single row of cells which appear to cor¬ 
respond to the Purkinje cells of the cerebellum; their stratum may therefore have the 
same denomination of intermediate layer (fig. 17), and the term is still appropriate, for 
although they are not intermediate between the granular layer and the molecular— 
the former having encroached more on the latter than in the cerebellum—they are still 
intermediate between the granular and the fibrous layer, which latter is common to two 
contiguous ridges. As just mentioned, these cells are arranged in a single row in 
ordinarily formed ridges, i.e ., in ridges which occupy the straight parts of the wings, 
and are bounded on one side by the granular layer, and on the other by a space 
which separates one ridge from the other, and in which the fibrils from the fibrous 
layer are found, which presumably join the axis cylinder processes of these cells, 
although I have not actually been able to see the junction. 
These cells are generally pear-shaped or oval, and usually have two processes, one at 
each end ; one of these is the protoplasmic process, which can often be traced through 
the granular into the molecular layer. This process usually runs straight; but in the 
cell which formed the subject of the drawing (fig. 17, a.) it had accidentally become 
bent; the other process is probably the axis cylinder process, and it is not difficult to 
understand that it may join the fibrillse from the fibrous layer. Occasionally cells 
occur which are more or less triangular in shape, in which case the apex is turned 
towards the granular layer, and two processes are given olf from the broad end, which 
probably join the fibrils. All these Purkinje cells are of larger size than the largest 
of those belonging to the granular layer, although smaller than the Purkinje cells of 
the cerebellum. Their measurements are as follows (fig. 17, a.) :—length, 0*01 millim.; 
width, 0*0068 millim.; nucleus, 0*0048 millim. in diameter; and nucleolus 0*0015 
millim. Another cell is rather longer and larger, being 0*016 millim. long and 
0*012 millim. wide, with a nucleus 0*006 millim. in diameter; but the nucleolus is 
smaller, being only 0*0005 millim. in diameter. The arrangement of these cells in 
those ridges which are placed on the straight part (either horizontal or perpendicular) 
of the wings is shown in the figure (fig. 17) ; but where there are curves and foldings, 
as at the angles of junction of horizontal with perpendicular wings, other cells are 
added to the single row of Purkinje cells so as to form groups filling up the depressions 
