38 CHARLES F. W. MCCLURE, 



and sublimate solutions, it is evident that the spindles cannot be 

 artefacts. 



The following experiment further substantiates this view, by clearly- 

 showing that their appearance in sections is dependent upon not only 

 a particular mode of staining, but also upon the extent to which the 

 after-differentiation process is carried. 



For example, in those cells fixed by Flemming's solution, which 

 have remained in the iron-alum for about 3, and in the haematoxylin 

 for about 12 to 18 hours, the spindles are with difficulty decolorized 

 when differentiated in the iron-alum; and are even plainly seen in 

 sections which have remained in the differentiation fluid for 1 hour. 

 On the other hand, if the time of immersion is shortened to V2 to 

 1 hour for the mordant, and to 1 or 3 hours for the haematoxylin, 

 the spindles are easily decolorized when differentiated. 



By allowing such sections to remain in the differentiating medium 

 30 minutes, no structures are visible in the cell body 

 which one can definitely call spindles. 



Fig. 16 represents the ganglion cell of Helix which was fixed 

 by Flemming's solution and stained by the iron-alum-haematoxylin 

 method. The section remained in the mordant IV2 hours and in 

 the staining fluid 2 hours. It was then differentiated for 30 minutes 

 in iron-alum. An examination of the body of this cell, shows that no 

 spindle-structures are visible. Compare this figure with Fig. 15, which 

 represents a ganglion cell of Helix ^ in which a similar absence of 

 spindles is seen. 



The fact that the spindles are clearly defined in certain pre- 

 parations, which have been stained by the progressive iron-haemato- 

 xylin stain, seems to find its explanation in the circumstance that 

 the osmic acid probably stains these structures to some extent, and, 

 at the same time, renders them extremely susceptible to haematoxylin 

 staining, when iron-alum is used as a mordant. All of the chromo- 

 philous granules in the cell body are deeply stained by this method, 

 but by virtue of the fact that these spindles consist of compactly 

 arranged groups of the small granules, the latter, when progressively 

 differentiated, hold the stain more tenaciously than the more isolated 

 granules and, therefore, show up by contrast in such preparations. 



The spindles are not to be mistaken for pigment granules, which 

 also stain deeply in these FLEMMiNG-iron-haematoxylin preparations. 

 The pigment granules are most commonly found in the cell body, at 

 the base of the axis-cylinder process, although they frequently extend 



