Simply put, fossil sequences are based on morphological similarities between any two fossils in the sequence. However, this isn't the whole of it by any means. Diagnostic features, which are features limited to related extant species and fossil species are the main criteria for morphological comparisons. On top of the morphological similarities, the stratum the fossil is found in determines the age of the fossils, which has to be sequential in direct lineages but can be 'out' to some degree in sister species. Ecologies are also considered when deriving sequences.
Here is a quick list of features found in the Cetartiodactyl sequence taken from an earlier post of mine.
The transition in fossils in a number of lines, including those of the cetartiodactyls, show much more than just 'possible' connections. If those fossils just showed one or two transitional features, there would be room for doubt. In the case of whales, we have a continuous line of fossils that show 1) an elongation of the head, 2) movement of the head/neck joint from the lower rear to rear position, 3) movement of the nostrils from the front of the snout to the top of the head, 4) change in the ear from above water use to below water use, 5) change in leg length from long to short in the front, and gone in the back, 6) change in back leg/pelvis connection from connected to unconnected, 7) change in spine from rigid to flexible. There are a few more shared features that I won't bother to list, I think this is enough for a start.
"2. I read the "Chihuahua (sp) versus Great Dane argument regarding species. I also read here in this thread "Groups that can interbreed to some degree can still be separate species. Consider lions and tigers, for example. A better definition is that species are groups that *don't* interbreed to any large degree. A more technical way to put it is that they are independent breeding populations. "
Would this not make these dogs into two different species, because they are (almost certainly) unlikey to breed?
Although different areas of evolutionary study use slightly different definitions of species, I find the most useful on these threads is the 'gene flow' definition. If there is potential gene flow between the two subspecies in question, then they are not separate species. In the case of Great Danes and chihuahuas, they would definitely not breed normally so could be considered different species, except for the possibility of genes to flow from one to the other through the intermediate sized dogs. Stop that flow somehow and you would have two separate species.
"Could physical geography be used as a reason to seperate a species?
That is the usual reason for separation. You might check out (Google) Allopatric and Sympatric speciation.
But what CAUSED these things to occur?
Using the Schroedinger's cat illustration, these critters either lived in the water or they didn't, as being equally efficient on both is illogical.
Anyone who came across these, who had no preconceived ideas about them would say they were all different creatures an none gave rise to the other.