DISTRIBUTION OF NON-MEDULLATED NERVE-FIBRES. 31 



edges with oil. After a short time I was occasionally able 

 to hit upon a single pigment-cell, which had, in the language 

 of the authors, retracted its processes almost completely. I 

 was then able, in an hour's time or more, to convince 

 myself with positive certainty of the existence of pale and, in 

 parts, clearly striated processes, although the pigmented 

 portion had undergone no kind of change. 



I believe, then, that I have justified the statements made 

 above, and am, thereby, in agreement with Lister (34), who, 

 as is well known, attributes the alterations in the pigment- 

 cells of the web of the frog's foot, in part at least, to a trans- 

 ference of pigment-granules from the processes to the body of 

 the cell, and vice versa. 



Let us now return to our gold preparation, and more espe- 

 cially to the three pigment-cells. There is one other fact 

 beside the relations mentioned above, which we are able to 

 establish ; it is this : the pigment-cells are connected by their 

 unpigmented processes, not only with one another, but also 

 with the ordinary smaller branched cells of the nictitating 

 membrane. I have seen this relation, it is true, with 

 perfect certainty only in a single instance (and I Avish to 

 state this before going further) ; the instance was that 

 of the cells marked e and h in figure 9, but this with 

 such distinctness that I cannot hesitate to give it a general 

 application. 



We occasionally come, in gold preparations, upon a pigment- 

 cell from which there is given off a very thin and pale 

 process, appearing to taper to a point, and containing, at the 

 same time, pigment-granules arranged at tolerably regular 

 intervals behind one another. Considering that a fine nerve- 

 fibre may be seen running close by, and this also possesses 

 dark granular enlargements, we might easily fall into the 

 error of mistaking this process for a detached nervous 

 thread. With a higher magnifying power we can, however, 

 easily convince ourselves of its very different colour. 



All pigment-cells of the nictitating membrane are not, how- 

 ever, identical in appearance with those which have been just 

 described. In the superficial layers of many such membranes, 

 and especially in the neighbourhood of the pigmented border, 

 there occur pigment-cells of the type represented in fig. 8. 

 They are of a bluish-gray or bluish-green colour, contain very 

 minute and regularly distributed pigment-granules ; their 

 processes (in opposition to those of the pigment- cells formerly 

 mentioned, as well as to those of the ordinary branched 

 cells) possess nodular swellings of very characteristic appear- 

 ance. In point of size these cells are intermediate between the 



