The finer Structure of the Nerve Cells of Invertebrates. 37 



In sections of cells which have been fixed by Flemming's solution, 

 but stained by any other method than the progressive iron-haemato- 

 xylin (as by Delafield's haematoxylin, iron-alum and Delafield's 

 haematoxylin and safranin-light green), it is a question whether one 

 can see any structures in these cells which he can definitely say cor- 

 respond to these spindles. 



In such sections the same differentiation in regard to the staining 

 capacity of the axis-cylinder process and cell body is produced, as 

 previously mentioned in connection with other sections. See Fig. 15, 

 which is a ganglion cell of Helix fixed in Flemming's solution and 

 stained in iron-alum-ÜELAFiELü's haematoxylin. 



The general appearance, presented by the cell body of this cell, 

 is that it contains a large number of small, darkly stained granules 

 which are arranged chiefly in rows. In this respect, it agrees with 

 the general arrangement of these granules, as seen in sublimate- 

 methylen blue-eosin, etc., preparations (compare with Fig. 4), in which 

 a grouping of small chromophilous granules into spindle-shaped groups 

 is not to be made out. There are certain points in the cell body, 

 however, which one might regard as spindles, but the contrast 

 between their manner of staining and that of the sur- 

 rounding structures is so slight, that one cannot deter- 

 mine their character with sufficient definiteness to 

 regard them as such. 



In sublimate sections which have been stained by the progressive 

 iron-haematoxylin method, the spindles are never so clearly defined 

 as in the Flemming preparations which have been similarly stained 

 (Fig. 14). In fact, unless the differentiation process is carried to an 

 exact degree, the spindles cannot be definitely made out 

 at all; and the appearance then presented by such cells 

 is identical with those which have been stained in 

 Flemming's solution, but stained otherwise than by iron- 

 alum-haematoxylin (Fig. 15). 



In sections fixed by sublimate and stained by the progressive 

 iron-haematoxylin method, the spindles give up their stain much 

 more readily when differentiated, than in the Flemming preparations. 



Finally, sections of nerve cells fixed in sublimate, which have 

 been stained in any other manner than by iron-alum-haematoxylin, 

 show a similar absence of spindles in the cell body. 



In view of what has already been said concerning the clear de- 

 finition of the spindles in material which has been fixed in Flemming's 



