130 Mr. PTa. Buxton o)i the 



Viallanes). This is siijigested by its <reneral direction and 

 by the fact that it ends on the surface of the fibrillar part 

 of the brain, under a thin portion of the fjanfilion cell laver. 

 With this we may compare the similar " free " ending of 

 the tubercule anterierrr in Acridians (Viallanes, 1887, p. 42, 

 fig. 46), and of the vardere Wnrzel in Apis (Jonescu, p. 137, 

 Text fig. 10a) and in Vespa (Viallanes, 1886). The author 

 remarks : " La premiere se porte directement en avont pour 

 gagner la surface anterieure dn renjiement primaire ; c'est 

 la come anterieure.^' 



If, then, the ascending branch is the equivalent of the 

 vordere Wurzel, we must homologise the posterior branch 

 with the riXcMaufige Wurzel, or backward root, in considera- 

 tion of its backward direction and deeply buried termination : 

 this would probably be accepted were it not that Kiihnle 

 has asserted that the tid)erciile anterieur of Viallanes is the 

 homologue of the vordere Wurzel and also of the rucUaufige 

 Wurzel. Against this I must enter a most emphatic protest. 

 In the first place, there is an inherent improbability about 

 it ; we cannot willingly believe that vordere and anterieur 

 refer to an organ which is described in other insects as 

 " running back " {rilcJdaufge). We surely need good 

 evidence before we can accept such a statement ? In the 

 face of that improbability Kiihnle was dangerously bold in 

 asserting the homology. When a great many more types 

 have been investigated we shall be able to bridge many of 

 the gaps which at present exist in our knowledge. Till 

 then we can none of us be certain of any but the most 

 obvious homologies, partly because the described types 

 are so few, partly because no living man has first-hand 

 knowledge of more than half a dozen insect brains. 



Kiihnle was probably led into this error by the fact that 

 the majority of insect brains show only two roots, some of 

 them lacking the forward, others the backward root. 



If, however, Kiihnle's homology be accepted, the one 

 which I have suggested must fall ; for clearly I cannot 

 give the terms vordere and riicklaufge to two structures if, 

 as Kiihnle says, they are in this case synonymous. I do 

 not wish to press my own convictions unduly ; but at any 

 rate they are based on considerations of relative position, 

 that is to say on actual fact. 



If, then, Kiihnle is right and I am wrong, we may either 

 assume that the ascending trunk and its branches (ascend- 

 ing and posterior) together form the backward {rilchlcmfige. 



