Origin of the banana graph The graph with two vertices and $n > 1$ edges connecting them has been called the "banana graph" in a number of papers. For one example, see "Feynman Motives of Banana Graphs" by Aluffi and Marcoli, Comm. in Number Theory and Physics (2009) 1-57. (The short title of this paper is "Banana Motives", which I find endlessly entertaining.)
Does anyone know who coined the term "banana graph"?  
 A: 
These diagrams come by different names: "banana", "water melon", "basket ball". An early reference is M. Creutz - Feynman rules for lattice gauge theory, Rev. Mod. Phys. 50, 561–571 (1978). A more recent reference is S. Groote, J.G. Körner, A.A. Pivovarov - On the evaluation of sunset-type Feynman diagrams (1998).
There is a long tradition of giving fanciful names to Feynman diagrams. This is the sunset diagram:

A: I just had lunch with Oliver Schnetz and our conversation broached the topic whether any of us ever coined a name which stuck. He mentioned to be the first one to attach the word "banana" to the banana graph. As a witness serves his unpublished paper Calculation of the $\phi^4$ $6$-loop non-zeta transcendental
from 1999 which deals with "$n$-banana diagrams". Later he suggested the term to Marcolli. She modified it to "banana graph" in her paper with Aluffi from 2009.
Incidentally, Oliver mentioned also that the name sunset diagram in Carlo's answer is a misnomer. It should be sunrise diagram because "the sun never sets on Quantum Field Theory" (D. Broadhurst) (-:
