BRAINS OF THREE GENERA bF ANTS 553 



other parts of the brain are as follows: connected by three sets 

 of fibers with the optic lobes, Kenyon, connected by fibers with 

 the protocerebrum, Bellonci, Viallanes, while all writers are agreed 

 that the main bulk of the fibers pass into the stalks, thence in 

 smaller part to the anterior roots ending on the anterior surface of 

 the protocerebral lobes, in larger part to the 'inner roots' which 

 were said to end abruptly beneath the central body without 

 visible connection with other parts of the brain. 



I have shown that the anterior dorsal protocerebral commissure 

 and the median, protocerebral tissue are connected with fibers 

 from the mushroom body tracts, d, b, e, and probably others; that 

 the tracts, c, p, n, connect with the lateral protocerebral tissue, 

 but that the largest, most prominent fiber tracts, h, gl., gr., 

 originating in the largest cells, those of Groups I and II, and 

 certainly efferent in nature, together with the smaller tracts 7n, and 

 r, from Group III, form the main bulk of the stalks, and that 

 the greater part of their fibers can be traced into the so-called 

 'inner roots,' which are said to end beneath the central body, but 

 which, however, do not so end, but are distributed in part to the 

 central body, in part, as the posterior roots of the mushroom 

 bodies, to the protocerebral lobes at the point where these are 

 connected with the subesophageal ganglion and consequently 

 with the ventral nerve cord. Granted that the fibers of the stalk 

 are efferent, motor, fibers, the question at once arises, what part 

 could they play by merely running into the 'inner roots' and 

 ending there? Is it reasonable that the chief efferent fibers of 

 the body should end abruptly leaving no trace behind and without 

 connection with other nerve centers? With the desire to trace 

 some connection with the inner ends of the mushroom body stalks 

 I have studied this region with great care and have returned to 

 the problem again and again, until I am now confident that the 

 following description is correct. The great efferent tracts from 

 the mushroom bodies which may be traced into the stalks, and 

 which go in small part to the anterior roots, do not terminate 

 beneath the central body, but there each stalk divides into two 

 roots. One root, the dorsal bundle, goes to the central body 

 and through the central body connects by fibers with the proto- 



