Edward Drinker Cope was the kind of hotheaded savant that science just doesn’t make room for anymore: he was wealthy, self-taught, and a product of the Victorian era’s craze for science, natural history, and fossils—when paleontologists might as well have been prospectors, and when making a reputation as a bone hound often meant beating the other guy to the dig. He was as stormy and wide open as the western plains where he made his name.

Cope never had much formal education, but he dug science from the get-go. At age eight, on a visit to Philadelphia’s Museum of the Academy of Natural Sciences, he carefully measured, sketched, and described several beasts, including a fossil Ichthyosaurus: “two of the sclerotic plates,” he wrote, referring to a bone in the animal’s eye socket. “Look at the eye, thee will see these in it.”

Years later, Cope plunged into an epic, career-long feud with a rival paleontologist: theft, bribery, espionage, sabotage, and dynamite all figured in what came to be known as the Bone Wars. But that’s a story for a different day.

By the time he died in 1897, Cope had funneled his intensity—and a good deal of his family fortune—into becoming one of the most prolific scientists in America. He named at least one thousand new species and published more than twelve hundred scientific papers.

And somehow, between all the digging and writing and caterwauling, Cope found time to leave us with a seemingly innocent little theory about how organisms evolve.

Cope’s rule, as it’s now called, states that animal and plant lineages tend to increase in body size over geological time. Cope based his theory on what he saw in the fossil record: for a given vertebrate group, it looked as though the oldest fossils were also the smallest. Over eons, members of that group looked as though they got bigger. The earliest ancestor of the modern-day horse, for example, was a dog-sized critter called Hyracotherium. Between Hyracotherium and, say, Mr. Ed, there have been several “horse” species, each larger than the one before.

Cope’s rule went over fairly well at the time. But by the 1970s, biologists had a few bones to pick with it. Some of them said Cope was just seeing things. After all, it’s easier to find fossils that are bigger, and large fossils might preserve better than small fossils. Stephen Jay Gould, who was something of a titan among paleontologists, called Cope’s rule a “psychological artifact,” saying that we humans tend to “focus on extremes that intrigue us”—that we singled out lineages that did get bigger instead of studying all lineages for which we had good data over long periods.

Even scientists who concur with Cope’s rule haven’t been sure what could actually cause it.

Many of them—including Carolina biology professors Joel Kingsolver and David Pfennig—suspect that maybe Cope’s rule works because things that are larger tend to have higher fitness. In other words, larger individuals survive better, mate more successfully, and produce more offspring than do smaller individuals in the same population. “To be successful, you have to survive, you have to mate, and you have to produce offspring,” Joel Kingsolver says. “The basic idea is that, from one generation to the next, individuals that are larger have higher fitness, and are favored, so the next generation gets a little larger, and the next a little larger, and so forth.”

Kingsolver is big on bugs—he studies natural selection in butterflies. “I was doing my own work within my own system,” he says, “and I kept getting this really simple result that, in the butterflies I study, faster growth and larger size were consistently associated with higher fitness.”

David Pfennig, on the other hand, leans more toward spadefoot toads and cannibalistic tiger salamanders. He studies the factors that produce new traits and new species and examines how evolutionary interactions at the genetic and cellular level might influence how organisms develop.

Both Kingsolver and Pfennig work on whether and how microevolution—small-scale changes from one generation to the next—might influence macroevolution, the large-scale changes that Cope and others have seen in the fossil record.

“A long-standing issue in evolutionary biology is whether the evolutionary forces that generate microevolution are sufficient to account for macroevolutionary patterns,” Pfennig says. “Evolutionary biologists have been especially interested in explaining the grand trends in the history of life, such as Cope’s rule.”

“I think of what we do as evolutionary mechanics,” Kingsolver explains. “We can study the nuts and bolts of evolution, from one generation to the next, experimentally and in lots of detail, and learn how it works and what it does. But you like to understand how that might generalize to larger-scale patterns.”

As part of a seminar a few years ago, Kingsolver and his graduate students created a database. They rounded up every published scientific study that examined, in any kind of organism, the strength of natural selection—the survival and perpetuation of one kind of organism over others that die or fail to produce offspring. They ended up with sixty-three microevolutionary studies on all kinds of living organisms—everything from plants to birds to lizards to insects.

Then, last year, Kingsolver and Pfennig decided to put Edward Drinker Cope to the test. They figured that by digging through all the studies in Kingsolver’s database, they might be able to uncover a mechanism for Cope’s rule. Many of the studies had to do with morphology—the form and structure of organisms—and, in particular, the strength of natural selection on traits such as shape, color, and size.

Generally, when biologists look at morphology, they find either positive selection for a particular trait, meaning increases in that trait are favored, or negative selection, meaning that decreases in the trait are favored. Darwin’s finches sometimes show positive selection for beak width: birds with wider beaks have higher rates of survival and reproduction. But sometimes, Darwin’s finches show negative selection: birds with narrower beaks have higher survival and reproduction rates.

If you look at lots of different traits, Kingsolver says, selection usually has no consistent direction, and the numbers tend to center nicely around zero: sometimes selection is positive; sometimes it’s negative. Just what you’d expect.

But not when it comes to size. When Kingsolver and Pfennig dug into Kingsolver’s database, they found that, in 80 percent of the studies, there was consistent selection favoring larger size. In other words, eight times out of ten, larger individuals tend to have higher fitness and are selectively favored.

They dug a little deeper. “We said, ‘Well, do you see this pattern if you look at plants versus animals, or vertebrates versus insects?’” Kingsolver says. “And the answer is that it holds for all those different groups.”

Then Kingsolver and Pfennig started thinking about the three components of fitness. They wanted to know if being larger was associated with being able to survive better, or with being more successful in mating, or with having more offspring. “It turns out that being bigger is better for all of those different aspects of fitness,” Kingsolver says. “So, on average, bigger individuals survive better; they produce more offspring; they’re more successful at mating.”

That was big news to Kingsolver and Pfennig, but it didn’t tell them enough to figure out whether Edward Drinker Cope was really on to something.

First, they had to make sure that positive selection for size was strong enough to account for Cope’s rule—could it, over hundreds of thousands of years, create the size increases that Cope saw in his fossils? It turned out to be more than strong enough. “It could account for the kinds of patterns that you see of size increases in dinosaurs and mammals and many other groups,” Kingsolver says. “In fact, it’s much stronger than you would need.”

Second, they had to consider whether any other phenomena might be muddying the waters, or flat-out working against Cope’s rule. Even though positive selection for size is strong, could other factors still be preventing evolution for size?

Kingsolver and Pfennig delved into the requirements for evolution. There are three: first, you need selection. Second, you have to have variation—a difference or deviation from an organism’s normal or recognized form, function, or structure. Third, the variation has to be inherited so it can be passed on to the next generation. “We knew there was selection and variation,” Kingsolver says. Turns out there was plenty of inherited variability, too. “If you look at studies that examine whether there’s inherited variation for body size, nearly every population has some inherited variability,” he says.

Then Kingsolver and Pfennig tested whether being big might come with its own disadvantages that would counterbalance Cope’s rule. For example, the bigger you are, the longer it takes you to reach fertility. If you’re smaller and can develop faster, you can have more generations per unit of time. So maybe the advantages of being large are, in the big picture, cancelled out by the advantages of developing rapidly. “We were able to look at that,” Kingsolver says, “and we discovered that yes, there is selection favoring developing more rapidly—but it still doesn’t cancel out the net effect of bigger being better.”

So why aren’t we all gigantic? And why have studies shown that, in some organisms such as fish, there have been no long-term increases in size?

For one, certain kinds of organisms come with their own physical limitations. Being an insect means having an exoskeleton. “You can’t just scale up an insect to the size of a dinosaur, or even a human, for that matter,” Kingsolver says. “An insect’s exoskeleton simply won’t support that kind of weight.” There are also upper limits to the sizes that other kinds of organisms can reach. Terrestrial mammals have an internal skeleton, but again, that skeleton can only support so much weight. Scientists think that certain of the largest species of dinosaur spent almost all their time in the water to help support their body mass. And water’s buoyancy allows whales to become much larger than any terrestrial mammal.

“Those kinds of things set absolute upper size limits,” Kingsolver says. But most species are nowhere near as large as these limits, so the limits are really not a very good explanation for body-size evolution in general.

“One thing that’s interesting,” Kingsolver says, “is that this can’t continue forever. If our result is true, and Cope’s rule is true, then everything ought to be the size of the universe—or at least pretty big. Everything ought to be bigger than dinosaurs. And, obviously, they’re not. So there need to be other things that are resetting the system.”

Kingsolver and Pfennig couldn’t find anything at the short-term, microevolutionary level that might be able to do that. But they suspect that several long-term evolutionary forces could cause an overall trend toward smaller species. One is extinction. In addition to their longer generation times, large species have smaller populations—fifty square miles is only going to support so many grizzly bears—so those species are more likely to go extinct.

Fossils tend to bear this out. The canid family includes several groups: dogs, wolves, jackals, and a few more that are now extinct. Each of those groups has shown evolutionary size increases within the group, but the larger species within the groups have been more likely to become extinct. “That resets the group as a whole to having smaller body sizes,” Kingsolver says. “That’s happening at ten- to forty-million-year time scales.”

But Kingsolver says that mass extinction events, such as the one that may have wiped out most of the dinosaurs, probably play an even bigger role. “That event reset everything: it wiped out all the dinosaurs, and left all the tiny mammals around—and we took over the world,” he says. Historically, there have been several cycles of large mammals evolving only to go extinct—think of North America’s woolly mammoth and saber-toothed cat. Those cycles of extinction may have repeatedly counteracted Cope’s rule.

So where does that leave Edward Drinker Cope? Was the old bone hound on the right trail? Or was he barking up the wrong tree?

“As with almost any question dealing with history, you can’t say with certainty what has caused a historical pattern,” Pfennig says. Proving Cope’s rule experimentally would be no picnic. In theory, Pfennig says, you might be able to do it using an organism that had really short generation times—some kind of microbe, for instance.

Kingsolver and Pfennig can say that their data show that bigger is generally fitter. They suggest that selection could explain the pattern described by Cope’s rule.

And how strong is that “could”?

“I would say that it’s a pretty strong ‘could’ until someone provides data refuting our hypothesis,” Pfennig says, “or until someone provides an alternative hypothesis that is better supported.”

And if Cope’s rule still makes for a bone of contention among biologists, so be it. Edward Drinker Cope would be proud.