One of the most successful groups of Mesozoic mammals are called multituberculates. The name refers to the many cusps (tubercles) on their molar teeth. These multis actually survived way beyond the Mesozoic. The most recent representatives died out about forty million years ago. As the earliest fossils come from Middle Jurassic rocks, the order had an impressive history lasting about 130 million years. Remains have been found all across the northern hemisphere and a few specimens have also been recovered from South America and North Africa.

Although far from close relatives, multis seem to have lived much like rodents do today. They were typically mouse-sized plant-eaters, but probably ate animal protein too, if available.

The oldest fossils are a few isolated teeth from England. Better remains have come from Upper Jurassic Guimarota in Portugal. These include jaws and some skull material from animals called paulchoffatiids, (named after a geologist, Paul Choffat). Over fifteen different species have been described from the one location, but the actual diversity was probably lower. It's generally not been possible to pair up lower jaws with skulls. The paulchoffs and other primitive multis have collectively been placed in the suborder of Plagiaulacida. Later plagis have been found mainly in Europe, Asia and North America, but material other than jaws or teeth is very rare. The best preserved fossil is Sinobaatar from China.

Trends in the story of multis are a reduction in the number of teeth and an increase in specialisation. Old fashioned forms had the following maximum numbers of teeth on each side of the mouth: upper jaw -three incisors, one canine, five premolars, two molars; lower jaw -one incisor, no canine, four premolars and two molars. There doesn't seem to have been much difference in the work performed by premolars and molars, especially when worn down. The most spectacular feature was the long, deep-rooted, lower incisor. This was angled forwards from the mouth and looks quite vicious, although it was probably a useful tool rather than a weapon.

During the Lower Cretaceous more advanced multis emerged. These are called cimolodontans. The number of premolars per side had been reduced to two. On the lower jaw number four had expanded and changed shape and function. The third was either small or (later still) absent. The top of the fourth formed a long, narrow and serrated blade. It was an effective salad knife. Food could then have been pushed back onto the molars for grinding.

The best preserved multi specimens have been found in Upper Cretaceous rocks of Mongolia. Rather than scraps of jaw and isolated teeth, hundreds of skulls have been found, and even a few complete skeletons. At least in these cases, the hips are surprisingly narrow. This suggests they either laid extremely tiny eggs or reproduced like maruspials; 'premature' live births. If so this habit must've developed independently, seeing as the two groups aren't closely related. At present the relationship of multis to other mammals is unclear. Several features of the anatomy are similar to monotremes such as the platypus, whilst others suggest possible affinities with an extinct group called triconodonts. More fossils will need to be found for a clearer understanding. In any case, they're not close relatives of mine.

Whilst most multis were mouse-sized there were exceptions, especially in the Paleocene after the Cretaceous. An animal called Taeniolabis was widespread in North America, and it was one of the largest mammals then to have evolved. Taenio was beaver-big and presumably weighed 20 kilos and more. Not many rodents have since reached such a size, though there was a buffalo-big guinea-pig relative around about eight million years ago, (should they actually be proper rodents).

After the Paleocene, multis became rarer. As rodents are more commonly found and the two groups seem to have led similar lives, this might well explain their eventual disappearance. During a long career multis had coped with immense changes; new kinds of plants and a mass extinction. It may well have been that they were nibbled out of existence.

Fact File

Plagiaulacida
This suborder contains the most primitive multis. As it doesn't consist of one ancestor and all of its descendants, it isn't seen as a natural group. Nevertheless, it's a convenient one.

Cimolodonta
This suborder contains the more advanced multis. It is a natural group of related animals. Cimolodontans are known from the Lower Cretaceous to the Eocene, (about forty million years ago).

Paulchoffatia
Meaning: 'for Paul Choffat'
Place: Guimarota, Portugal and Galve, Spain
Age: Upper Jurassic to Lower Cretaceous
Remarks: A number of lower jaws are known. When alive, the animal had a skull-length of about 2.5cm. That's about the same as for a mouse.

Sinobaatar
Meaning: 'Chinese hero'
Place: Liaoning, China
Age: Lower Cretaceous
Remarks: This is known from an exceptionally well-preserved specimen. Its teeth have some characteristics more like the advanced multis of Cimolodonta. The feet may indicate the animal spent at least some time living in trees. 'Baatar' is a Mongolian word meaning 'hero', and it often gets used for multis.

Taeniolabis
Place: North America
Age: Paleocene
Remarks: This is the largest known multi, and remains have come from many places in both the USA and Canada. A reasonably complete skull has a length of about 16cm.