560 PROCEEDINGS OF THE AMERICAN ACADEMY. 



accidental staining of parts adjacent to the nerve fibres, — a sort 

 of extravasation, — accompanied by a failure to stain on the part of 

 the real cell body and the more distal portions of the sensory cell. 

 The spindle-shaped enlargement is sometimes stained uniformly, but 

 more often the staining is irregular and blotchy ; in some cases a 

 nucleus is to be distinguished near the middle of the cell body in 

 the widest part of the spindle, which it almost completely fills. In 

 one case (PI. 3, Fig. 14) the nucleus was sharply differentiated from 

 the cell body, which was not at all blotchy, but distinctly fibrous and 

 sparsely granular. 



From the distal end of the spindle-shaped cell body there passes off a 

 fibre that, I believe, breaks up into a number of fibrils, each of which 

 seems to me to end iu a disc (PI. l, Fig. 5 ; PL 3, Fig. 14). In Fig- 

 ure 31 (PI. 4), the fibrils are quite clearly recognizable ; in Figures 25 

 and 29 (PI. 4), though distinguishable, they are not so evident. The 

 terminal discs (PI. 3, Fig. 18; PI. 4, Figs. 25, 29) may, it is true, 

 be artefacts ; but the frequency of their occurrence and the similarity of 

 their appearance seem to me to be arguments against that supposition. 

 Sometimes the blue is deposited in great amount around this bunch of 

 fibrils (PI. 3, Figs. 12, 15, 18 ; PI. 4, Fig. 29), but in other cases it has 

 failed entirely to stain the portion of the sensory cell that is distal to the 

 spindle-shaped enlargement. On the other hand, there are cases in which 

 the peripheral part of the distal portion of the sense cell has been differ- 

 entiated by staining in haematoxylin (PI. 2, Fig. 11, not well brought out 

 in the figure). In the case in which I have seen fibrils with their terminal 

 discs most distinctly (PI. 3, Fig. 14), the discs at the ends of the fibrils 

 are at the surface of the papilla outside the cuticula ; in other prepara- 

 tions, the fibrils seem not to pass through the cuticula, but to end at its 

 deep surface. It is probable that in most cases the cuticula has been 

 artificially separated from the protoplasmic mass of the papilla, and that 

 normally the fibrils pass to the surface of the papilla. 



The connection of the cell body with one of the eighteen longitudinal 

 nerves of the proboscis is often to be traced on a single thick section. 

 The process which the cell body sends centripetally either joins a longi- 

 tudinal nerve directly, or enters the peripheral nerve plexus, which in 

 turn joins the longitudinal nerve (PI. 2, Fig. 10; PI. 3, Fig. 19). The 

 basal end of each of the two or three cell bodies of the papilla seen in 

 methylen-blue preparations (PI. 3, Figs. 12, 14, 20; PI. 4, Figs. 23, 

 25, 29) is prolonged into a slender nerve fibre. While the fibre be- 

 longing to one of the cells of a papilla bends to the left when it joins 



