This was the musk deer (Moschus moschiferus). The most obviously strange thing about, for a deer, is that it doesn't have antlers. Less immediately apparent is the fact that it's the only kind of deer to have a gallbladder, and there are some other anatomical oddities, too. The former could be explained by it having lost them at some point in its evolution - and, indeed, there are other known species of deer where this has quite obviously happened. However, it's a bit trickier to explain why a gallbladder would come back after vanishing, so the assumption was that musk deer were a very primitive form of deer that had never lost it in the first place.
It turns out that this isn't true, but not for the reason that one might think. For one thing, it became apparent in the early 1980s that there were multiple living species of musk deer, and some people began proposing that they should be placed in a family all of their own, related to, but distinct from, the 'true' deer. And then, in 2003, molecular data dropped the real bombshell: not only are musk deer not deer, they're actually more closely related to antelopes and cattle than they are to true deer.
This has repeatedly confirmed since, and there's no longer any dispute that the musk deer aren't some weird form of true deer, but a family all of their own.
It turns out to be a family with quite a significant fossil history and, in retrospect, the forms of some of them do look more like early antelopes than they do like early deer. In fact, two extinct subfamilies have joined the living one, and there are a number of fossils that appear to lie outside even those, very early branches in what we now know is the musk deer family tree. Right at the base, likely the very earliest branch we know of is that occupied by Hispanomeryx.
Hispanomeryx had always been a bit odd, not least because some of its teeth really did look like those of antelopes. That's probably a derived feature, a case of convergent evolution, given how long ago they must have branched off from the real bovids, although a common starting point may well have helped. Although "it's a kind of musk deer" was the initial guess when it was first described in 1981, its many peculiarities, combined with very limited fossil remains meant that it was variously classified as a bovid, a giraffe-relative, or as "none of the above". This remained the case until 2010, when a complete enough skeleton was discovered to confirm its moschid status.
So what do we know of this animal? As its name indicates, the first fossil was discovered in Spain, and that's also where most subsequent ones were found. However, in 2011, one was found as far away as Mongolia, indicating that they must have ranged over a considerable area. Fragmentary remains have also been found in Turkey, Georgia, and China, and, collectively, we know of four different species. All of these lived in the Middle to Late Miocene, and range from 13 to 8 million years old.
According to a recent re-analysis of the known fossils, the Mongolian species most closely resembles the original form of the animal, lacking some of the subtle features that distinguish the Spanish species. Since it isn't really any older than most of those, it must represent a form that became isolated and did not evolve as its relatives did - presumably due to differences in the local climate and habitat. The little we know of it suggests that it may only have lived for a million years or so, while those in Spain survived rather longer, although this may be an illusion created by the limited number of fossils available.
Back in Spain, however, the species there became gradually larger with time, although they were still only around 30cm (1 foot) tall at the shoulder, considerably smaller than living musk deer. Around 9 to 10 million years ago, the teeth of the lower jaw began to more closely resemble those of bovids such as antelope, and it's probably this that led to the earlier confusion as to whether the animals really were musk deer at all.
However, the fact that some Spanish fossils are as old as the Mongolian ones - and the few from other Asian localities - suggests that Hispanomeryx was already widespread at this time, and that its ultimate origins lie further back. This makes sense, given that we know of another fossil musk deer, Micromeryx, that dates back at least 15 million years, and is seemingly much closer to the modern sort.
In fact, Micromeryx also lived in Spain, alongside its cousin. Fossils of this genus have been found across a number of different sites that we believe had different habitats from one another back in the Miocene, so that Micromeryx must have been a reasonably adaptable animal. In contrast, Hispanomeryx was thought to be more restricted, preferring open country and drier weather, living in environments something like the savannas of East Africa today.
But that's no longer looking so likely. The same analysis that revised the relationship between the four known species also identified a fifth one and, more significantly, identified a fossil of one of the known species from a site just north of Barcelona. The reason this matters is that the same site has also provided numerous fossils of dormice, flying squirrels, beavers, frogs, and apes. It's hard to imagine that such a place wasn't heavily forested at the time the fossils were deposited, and, moreover, probably had substantial bodies of fresh water.
Today, musk deer live only in Asia, but the oldest fossil representatives of their family are European. They may have spread further east fairly rapidly, helped by an ability to survive in a range of habitats, from dense subtropical forest to open savanna. Perhaps never hugely successful (or we'd have found more fossils than we have), over time they became more restricted in their habitat. By the time of the Ice Ages, when the first "modern" musk deer appear, they lived only in forests, perhaps retreating to the higher slopes they now inhabit as the world warmed.
[Photo by "Ra'ike", from Wikimedia Commons.]