Paraphyletic genera

[Jeffrey Martz <jeffmartz@earthlink.net>]

      Since the subjects of parapyletic genera and using speciemns instead of 
species and genera has come up: what happens when you have several species of a 
genus that branch off consecutively below a clade?  Wouldn't this indicate that 
the genus WAS ancestral to that clade?  Wouldn't that be a hypothesis as well 
supported as any other phylogenetic relationship offered by the analysis?   
      I can't name an example off the top of my head for dinosaurs but here is 
one from mosasaurs by Gordon Bell.  He did two (one in the Marine Reptiles 
book), but I'll use the analysis he used in his thesis. One part of his 
cladogram, the tribe Plioplatecarpini, looks like this (to save space, 
_Platecarpus_ is appreviated "P.":

                                                                Plioplatecarpus
                                                                             *
                                       P. af P. somenensis *******
                                                                             *
                    P. tympaniticus (DMNH)******************
                                                                             *
        P. tympaniticus (FMNH)****************************
                                                                             *
P. planifrons************************************************

     Another mosasaur worker, Theagarten Lingham-Soliar, published a cladogram 
for plioplatecarpine mosasaurs of his own a while back which also made out 
_Platecarpus_ as being paraphyletic with Plioplatecarpus higher up, although it 
was pretty different in other regards. 
     Since the common ancestor of all the _Platecarpus_ species must have been 
_Platecarpus_, wouldn't that mean that this cladogram is hypothesizing that 
_Platecarpus is ancestral to _Plioplatecarpus_?  Moreover, isn't it in fact 
hypothesizing that the species _P. tympaniticus_ is ancestral to 
Plioplatecarpus?  Can ancestral relationships be hypothesized by coding 
specimens rather then higher level taxa?

LN Jeff