Biology is the most complex of all the sciences and the one that touches us most closely; and natural history is half of it, simply observing what is, what's out there, and what it all does. Good naturalists are humble  properly awed by nature and always aware that life in the end is beyond our understanding. But the other, more practical half of biology  biotechnology, genetic engineering and all that  is flashier and more lucrative and so attracts all the funding. The humility is gone. The biotechnologists tend to assume that we already know all that's worth knowing; that we can take nature by the scruff and reshape our fellow creatures at whim, just by fiddling with their DNA. The Greeks were right about hubris. We need to restore the balance. Bring back natural history.

This is Richard Fortey's great strength. He is known as a palaeontologist, a scholar of fossils, but he is, he tells us, "a naturalist first". His latest book is in the great tradition of natural history  that of the nature ramble, led by the authoritative and other-worldly sage. He leads us on a ramble that is not only global but takes us through aeons, to look at creatures that haven't changed much to look at for hundreds of millions and in some cases for billions of years. Worldwide, there's a surprising number of these ancient types: it's a mistake to assume that the creatures that evolved later necessarily supplanted the ones that came before. Nature often arrived at very good solutions to life's problems a very long time ago  and why change a winning formula? Natural selection may work just as efficiently to keep things the same as it so obviously does to change them.

Take limulus  the horseshoe crab of North America, with an armoured carapace and jointed, arthropod legs, that scuttles on the sea bed like a clockwork toy: it's not in truth a crab at all but a very primitive arachnid, relative of the spiders and scorpions. It is very ancient indeed  fossils very similar are known from the Ordovician, 450m years ago; they clearly arose in the Cambrian, at least 500m years ago. For hundreds of millions of years they shared the seas with the trilobites, which were superficially similar and perhaps were distantly related. The trilobites were far more numerous than the early horseshoes, and far more varied: some of them predatory, some of them squeezing the nutrients from mud, and some of them free-swimming. But, as the fossils clearly show, the trilobites went extinct as most animals did in the greatest of all mass extinctions at the end of the Permian, around 255m years ago.

But the humbler horseshoes came through. Conditions deep down stay constant when turbulence reigns overhead. So life may favour the ultra-conservative. But horseshoes have other tricks, too. They can survive tremendous damage, largely because their blood clots so efficiently, sealing all wounds, and because they have a supremely effective immune system.

On the east coast of North America an estimated 17m horseshoes still come ashore each year to lay their eggs in the sand, like turtles. But whereas female turtles mate at sea and come ashore to lay eggs that are already fertilised, horseshoes lay infertile eggs. So the males come ashore too and compete to fertilise the eggs as they are laid. Fortey describes all this first-hand, in Delaware Bay, the greatest of all horseshoe breeding grounds; and if he had lived 400m years ago the spectacle would have been much the same. Seabirds come too. For them the horseshoe eggs are an essential feast, a staging post on their vast migrations from the deepest south to the Arctic.

The ramble continues: to New Zealand in search of the velvet worm, onychophora. On the west coast of Australia he encounters the stromatolites, built from layers of minerals and photosynthetic bacteria (bacteria invented photosynthesis, and plants nicked it from them). They're still going strong though they date from around 3,000m years ago. In North America's Grand Teton national park Fortey gets to grips with microbes that live in the super-heated water of the hot springs; possibly the conditions in which earthly life began. He likes plants, too, and meets clubmosses, whose ancient relatives were huge and abundant forest trees; tree ferns; and ginkgo biloba, sole survivor of the once huge and various tribe of the ginkgoales. And many more.

All in all it's a great story, and no one is better equipped to tell it than Fortey. Just one cavil: though evolution is well established as a fact of life (as much as any historical science can be), there are still some metaphysical loose ends that can never be put to rest but could certainly be addressed through a survey such as this. Is there progress in evolution? Can we really suggest that sophisticated creatures such as orb-web spiders  modern arachnids  have "progressed" since horseshoe days? What does progress amount to, if the ancient types are still thriving? What of the old concept of orthogenesis  the idea that evolution proceeds in straight(ish) lines from a primitive state to modern types? What do we make of "convergence"  while nature is endlessly inventive, why does it endlessly reinvent the same general forms and life-solutions in remarkably similar forms, such as the predilection for flat, armoured creatures, including trilobites and horseshoes? Where does this apparent sense of direction come from?

But Survivors is excellent natural history, even so. And, as those who finance science need reminding, natural history matters.