The interpretation of bubble and oyster Feynman diagrams? I am reading 'A Guide To Feynman Diagrams in the Many Body Problem' By R.D.Mattuck, in which their is reference to oyster and bubble Feynman diagrams, shown respectively below.

In these diagrams I am confused of what is meant to be a 'hole' or anti-particle and what is meant to be a normal particle. 
Here is my (clearly wrong) reasoning on what I know so far:

At 'a' and 'c' we have arrows pointing in the -ve time direction,
  therefore we must have holes here. But at d and b the lines are
  pointing in the positive time direction and therefore we must have
  particles.
By the same reasoning we have particles at 'e' and 'h' but holes at
  'f' and 'g'.

As already stated this reasoning does not work. So how do explain what is going  on in these diagrams in terms of holes and particles?
 A: As far as I am concerned, I think you shouldn't take the question particle/anti-particle too literally. Here is why:
Take you first picture. The bubble on the right hand side does represents a process which does not have a well defined stop or starting point. You could interpret it as (an this is how it is typically done) a particle anti-particle pair being created spontaneously from the vacuum (c is the anti-particle,d is the particle), then the anti particle interacts with some external particle and both particles annihilate each other (a,b). But you could as well say that at some other point of the loop a particle pair (or multiple at different points) are created and the particles and anti-particles travel forwards and backwards in time. Threre is really no reason why you have to pick the point between c & d as the starting point.
Now to your second picture. You could interpret this as a particle-anti-particle pair is created at the point between e and f. The antiparticle e annihilates the incoming particle. At the same time a particle anti-particle pair is created at the cusp at h. The created particles becomes the outgoing particle while the anti-particle h annihilates g, which is the particle f. Again, you could interpret it like this, but:


*

*As you may know, the position of the interaction vertices is integrated over all space and thus the ordering in which events happen may change, similarly the interpretations of internal lines in terms of particles and anti-particles.

*In this example the vertices are spacelike separated. This effectively means that you cannot say one event happens before or after the other. There are reference frames for either case. Thus, for spacelike separated events it does not make sense to talk about particles or anti-particles being exchanged. This can easily be seen if we interpret anti-particles as particles traveling backwards in time. Since "backwards in time" depends on the observer for spacelike separated events we can not assign the label "particle" or "antiparticle" to an excitation which mediates between these events.
Just one more remark. Feynman diagrams are a tool to simplify & organize calculations. I believe that one should not interpret them too literally, i.e. thinking that virtual particles are actually created (whatever that means) might be a bit too much...
