The first mammals appeared around 225 million years ago, not long after the first dinosaurs did. However, they were not descended from reptiles, but from an entirely separate evolutionary line - the Synapsida - that had lived alongside the reptiles for a very long time. Yes, if you go back far enough, reptiles and mammals do have a common ancestor, and that animal would have looked, to modern eyes, more like a reptile than it does a mammal. But it wasn't a reptile, not in the modern scientific sense, and it belonged to a group of animals that just aren't around any more.
But between the point when that animal lived, and the first mammals appeared, there were a whole host of other, non-mammalian, synapsids. We often divide these creatures into two broad "grades" of evolution: the early, reptile-like, pelycosaurs, and the later, more mammal-like, therapsids. (In fact, of course, the two lived together for quite some time - it's not that one simply turned into the other. Evolution is more complicated than that). A greatly simplified tree of the "true" therapsids is included below, and shows that early mammals had rather a lot of relatives.
The most successful of these, so far as we can tell from the fossil record, were the dicynodonts.These creatures flourished during the late Permian, long before the first dinosaurs made their appearance, let alone the true mammals. Along with other early synapsids, it was these animals that ruled the world in the period before the Triassic, and the dawn of the Age of Dinosaurs.
Perhaps the largest and best-known area in which we find Permian fossils is the Karoo Supergroup, centered, as its name implies, on the Great Karoo, a desolate region of semi-desert in the heart of South Africa. However, there are others, and one that's particularly close is a similarly aged series of escarpments running all the way from Kenya down to Zimbabwe. Being the same age, and in roughly the same location, while less fossils have been discovered here than in the Karoo proper, those that have been found are generally somewhat similar
One particularly large formation within this region lies in southern Tanzania, with strata covering almost the whole of the Permian, and even back into the preceding Carboniferous period. Excavations have been going on here since the 1930s, and, so far as air-breathing vertebrates are concerned, most of the focus has been on the highest, and therefore youngest, of the layers. A number of fossils have been recovered from this layer, including a number of dicynodonts, as well as other synapsids. The layer immediately underneath it, however, is a different matter.
This lower level consists of rocks dating back to the mid Permian, about 270 to 260 million years ago. (For comparison, the very oldest dinosaurs we know about are around 230 million years old... so, even geologically speaking, this is a fair way back). The rocks in question seem primarily to be mudstone, deposited on an ancient lakebed, so it's perhaps not surprising that fossils of land animals, rather than, say, fish, are far from plentiful here. The first fossils of air-breathing vertebrates were recovered from these rocks in 1963 on an expedition led by the British Museum of Natural History, but there hasn't been much follow-up since.
The most common synapsid fossil found on this expedition belonged to Endothiodon, an early dicynodont. It's not the oldest dicynodont known, but, dating from the mid, rather than the late Permian, it's not too far off. It, or relatives close enough that we can't easily distinguish them at this remote distance in time, was quite widespread, being found across the southern parts of what was then the single world-spanning continent of Pangaea. As a result, since the breakup of that continent, fossils of the same animal are found today, not just in southern Africa, but also places such as Brazil - which, at the time, was not so far away.
We can now add another dicynodont to the fossils found in this particular, Tanzanian, formation, however. Abajudon was first discovered during an expedition in 2008, but was only formally described at the end of last year. It's a relatively small dicynodont, although, since we only have half of the skull, it's difficult to say quite how large it was - I'd take a guess at maybe the size of a small dog, but it's hard to know for sure. Despite the relatively small and fragmented nature of the specimen, it's enough to show that the teeth were highly distinctive both in shape and arrangement, which is enough to qualify it as a previously unknown genus of animal.
Indeed, one of the features that distinguishes these early dicynodonts from the sort that were around later, when the group found wider success in the late Permian is that they had a fairly full set of teeth. The word "dicynodont" literally means "two dog-teeth", and refers to a pair of tusk-like canines in the upper jaw. At first sight, these look like the sort of thing that carnivores might have, but the fact that they're often the only teeth the animal has argues against that. Indeed, the rest of the mouth is usually a toothless beak, rather like that of a turtle, and the chewing teeth, when they are present, look like those of herbivores.
Perhaps then, the tusks were used in fighting, whether between rivals searching for a mate, or against predators. Or both, of course, since the one doesn't preclude the other. They probably had quite powerful jaws, too, since the openings in the skull that anchor the jaw muscles were unusually large in the group.
But still, they were probably herbivores, quite unlike the earliest true mammals, which seem to have been insectivores similar to modern shrews. Like many other synapsids of the time, they were also larger than early mammals, implying that, at some point in our earliest history, there must have been a reversal of the usual trend of animals getting larger over the course of their evolutionary history.
We know little about how these creatures lived their lives. That, in this case, they were found in lake deposits, probably means that at least some of them wandered along the shore - they certainly don't seem to have any aquatic adaptations - and that, sometimes, they drowned, perhaps in a sudden flood, leaving their bones to sink to the bottom. Would they have had fur, as true mammals do? Probably not, although we can't tell that much from a skeleton, and it's at least possible that they had some sort of bristles. Certainly, we can assume that they laid eggs, as some of the most "primitive" mammals do today.
These odd, not-quite mammals, were once widespread, and it does not seem that there was much variation in vertebrate fauna across at least the southern parts of Pangaea at the time. At the end of the Permian, a catastrophe of epic proportions, far more severe than that which would later kill off the dinosaurs, would spell the doom for many dicynodonts, as it did for many, many, other animals besides. They survived, but never in great numbers, and managed to make it at least to the end of the Triassic, living alongside both the first mammals and the first dinosaurs before they likely met their end in the next great extinction event at the end of that period.
But, if it hadn't been for the end-Permian extinction, dinosaurs would never have appeared, and the synapsids, which had ruled the world before they did, might well have dominated vertebrate life without ever giving the reptiles much of a chance.
[Photo by "Ghedoghedo", from Wikimedia Commons]