A century before Darwin, the Swedish naturalist Carl Linnaeus created a biological taxonomic system that, with some revisions, is still used. Linnaeus, a creationist who presumably harbored no suspicions about common descent, classified everything basically according to what it looked like or by other easily identifiable characteristics, such as feeding habits or gross anatomical similarities.
He built his system approximately like this. He started with a group whose members, at first glance, all looked essentially alike, and he called that group a species. Then he would take species that looked alike and group those species into a larger category called genus. Each organism could was then assigned a unique name reflecting its genus and species.
The genus for animals that most closely resemble dogs was called Canis. Wolves were called Canis lupus, coyotes were called Canis latrans, and domestic dogs were called Canis familiaris by Linnaeus. Biologists nowadays have reclassified domestic dogs as a subspecies of wolf, called Canis lupus familiaris.
Linnaeus then worked his way up the resemblance hierarchy. Several other genera, including the various kinds of foxes, also looked sort of dog-like, and so he grouped them into a larger category called a family. That family of animals that all look dog-like was named canidae. Besides Canis, some genera comprising the family canidae include Vulpes (typical foxes), Alopez (arctic foxes), Dusicyon (Falkland Island wolf), and Lycaon (African hunting dog).
Other families comprising other sets of genera bear certain superficial resemblances to the canidae, including clawed digits and a set of teeth especially suitable for eating meat. Linnaeus put those families into his next larger group, which he called carnivora. Among the constituent families of carnivora besides canidae are ursidae (bears), mustelidae (weasels, skunks, badgers, and otters) and felidae (cats). Such a group of similar families is called an order.
Among the other orders of animals are hoofed animals with an even number of toes on each foot, called artiodactyla. A certain group of flying animals, called bats, comprises the order chiroptera. Certain air-breathing aquatic animals make up an order called cetacea. Moles, shrews, and hedgehogs, which all eat nothing but insects, form an order called insectivora.
Orders have their own groupings, but at this point we will switch perspective and work from the other direction, starting with the largest divisions and working down.
Linnaeus divided all living things into two groups that he called kingdoms: plantae (plants) and animalia (animals). Each kingdom included certain unicellular organisms that biologists have since reclassified into two kingdoms of their own, monera and protista. Another group, comprising various multicellular plant-like organisms, was assigned to a fifth kingdom called fungi. All organisms now classified among the animalia are multicellular.
Linnaeus divided each kingdom into categories he called phyla (singular phylum). All animals now living belong to one of nine phyla. One phylum is defined as follows:
Chordates are defined as organisms that possess a structure called a notochord, at least during some part of their development. The notochord is a rod that extends most of the length of the body when it is fully developed. Lying dorsal to the gut but ventral to the central nervous system, it stiffens the body and acts as support during locomotion. Other characteristics shared by chordates include the following (from Hickman and Roberts, 1994):
http://animaldiversity.ummz.umich.edu/site/accounts/information/Chordata.html
The chordate phylum is further divided into three subphyla, one of which is vertebrata, animals with a skull and a segmented backbone. The vertebrata subphylum comprises the seven subgroups called classes. They are: mammalia (mammals), aves (birds), reptilia (reptiles), amphibia (amphibians), osteichthyes (bony fish), chondrichthyes (cartilaginous fish, e.g. sharks), and agnatha (jawless fish).
Each class is a group of orders. Animals in the orders that comprise mammalia all have certain characteristics in common. They all have a bony skeleton that includes a backbone, they all have at least some hair, and they all suckle their young after giving birth. Linnaeus made the suckling of young the defining characteristic of the class mammalia.
And now: What does any of this have to do with common descent?
Well, the defining characteristics of each classification group (taxon, plural taxa) fall into a pattern. Every taxon looks like a modified version of some other taxon. This is most apparent at the species level. Linnaeus's taxonomy would not even have been possible if not for the fact that not only do species have common characteristics, but so genuses, and so do orders, and so do classes, and so on all up and down the taxonomic lines. Mammals look like modified reptiles. Reptiles look like modified amphibians. Amphibians look like modified bony fish. Bony fish look like modified sharks. And sharks look like modified jawless fish.
A possible explanation for this fact would be that some jawless fish became (over many generations) sharks, some sharks became bony fish, some bony fish became amphibians, and so on down over millions of generations until some reptiles became mammals.
Of course it could have been just by chance that Linnaeus came up with a nested heirarchy of characteristics that lent itself to an interpretation of common descent -- an interpretation that would surely have been foreign to his own thinking. But it is rather remarkable that essentially the same nested hierarchy showed up again when the DNA code came to be deciphered. Variations in DNA from species to species, genus to genus, family to family, and so on across the whole of biology show the same pattern. Organisms that look most closely related related have the most similar DNA while those that look least closely related have the most dissimilar DNA, notwithstanding visual similarities. Porpoise DNA looks much more like hippopotamus DNA than it looks like shark DNA.
It is sometimes objected: If B is descended from A, then how can A still be around? It must be remembered that a descendant form does not have to replace its ancestral form. Consider Chihuahuas and wolves. It is physically (though not biologically) impossible for them to interbreed. No wolf therefore is going to give birth to a Chihuahua pup, and no Chihuahua is going to have wolf puppies. But, every Chihuahua alive today did have a wolf ancestor, and so did every wolf alive today. Wolves had to exist before Chihuahuas could exist, but wolves did not have to become extinct in order to have Chihuahua descendants. Thus, although Chihuahuas came from wolves, we observe that wolves are still very much with us.
In other words, if we believe that B is a descendant of A, it is not a problem if A is still around.
It would be a big problem, though, if the defining characteristics of various groups were randomly distributed instead of clustered the way they are. Fish do not differ from whales only in lacking lungs. Their skeletons and many other anatomical features are also very different. A bird is not just a bat with feathers. Lizards and salamanders look superficially a lot alike, but a careful examination of all their characteristics will show that a salamander is much more like a frog than like a lizard while the lizard is much more like a snake than like a salamander.
From a naturalistic perspective, this did not have to be. There is no natural reason why there could be no lizards with fur or feathers instead of scales. There is no natural reason why some birds could not have four legs or, like some insects, two pairs of wings. There is no natural reason why a warm-blooded aquatic animal could not breathe with gills instead of lungs and thereby not have to surface frequently for air.
There is no natural reason why these combinations cannot exist, except for one thing. If birds are descended from bipedal reptiles, it is to be expected that all of them would have only one pair of wings and one pair of legs. If whales are descended from terrestrial mammals, it is to be expected that they would breathe with lungs, not gills, and that their skeletons would be much more similar to pig skeletons than to trout skeletons.
It would also be problematic if a species were now discovered that combined the definitive characteristics of distant taxonomic groups. Some examples of discoveries that would threaten the theory of descent with modification would be:
Any of those animals could have existed. No law of nature says it cannot happen. That nothing like it, as far as we know, ever has happened is one datum in support of descent with modification.
Next: The facts: Characteristics common to all life