126 Mr. P. A. Buxton on the 



inwards past the side of the outer capsule of the central body. 

 This band gives off a few fibres to the outer capsule (tract n), 

 and then passes into that region of the protocerebral lobes 

 which lies on each side of the central body (tract r) ; there 

 the band divides and is lost to sight (PL VIII, fig. 12, and 

 PI. X). Such a tract has not been described in other 

 insects. 



The stem {st.) of the mushroom body leaves the ventral 

 side of the head and passes downwards and forwards and 

 also slightly inwards ; it is a single cylindrical rod of parallel 

 fibres ; and whatever may be the case in other insects it 

 is not penetrated by a canal. It may also be noticed that 

 it is not surrounded by a sheath. (Even if it were I should 

 not follow Kiihnle in calling the sheath a neurilemma. 

 A neurilemma is a syncytial layer covering a brain or 

 ganglion.) 



The stem is one of the most striking features of any 

 section in which it occurs ; it stains more heavily than 

 the surrounding protocerebral lobes, and is a most use- 

 ful landmark. The stem is a cylindrical structure, and 

 well-fixed material shows that the greater part of its 

 shaft is surrounded by a space which is not developed 

 at its top or bottom; at these points the stem simply 

 pierces the surrounding parts of the protocerebral lobes. 

 At its lower and anterior end the stem is swollen and be- 

 comes superficial, that is to say it is no longer buried in 

 the protocerebral lobes, but reaches the surface in the 

 interval between the dorsal lobe and the anterior part 

 of the ventrolateral lobe. At this point it is covered by a 

 few nuclei ; some of these are merely nuclei of neuroglia 

 cells, some are nuclei of true nerve cells, which send their 

 axons into the foot of the stem. This is certainly remark- 

 able, but I have satisfied myself that it is the case by 

 examining a large number of sections through this region. 



At the bottom of the stem we should expect the roots to 

 arise, and actually we find two processes of rather uncertain 

 homologies, both of which make a marked angle with the 

 stem. The first runs inwards and slightly backwards, and 

 is the inner root {in. r. ; innere Wurzel of Kiihnle).' It is 

 straight and cylindrical and separated from its fellow of 

 the opposite side by a very distinct ])art of the proto- 

 cerebrum, the anterior part of the middle lobe. At its 

 termination the inner root is somewhat swollen. 



The second structure which leaves the anterior end of 



