Most male mammals carry precious cargo in an awfully precarious package. External testicles—which swing delicately outside the abdominal cavity in an exposed, thin-skinned sack—are sensitive, finicky, and make a glaringly obvious target for any enemies of men (as well as footballs, stray knees and waist-level furniture). So it’s no surprise that the evolution of descended testicles has both baffled and bemused biologists for decades.

A study published today in PLOS Biology offers an answer to one of the mystery’s biggest debates: did our earliest ancestors wear them up, or down? The researchers conclude that the first mammals already had this bewildering trait, with millennia of descendants thereafter inheriting the family jewels on full display. Yet strangely, it appears that since then, internal testes have evolved at least four separate times.

During embryonic development in human males, reproductive structures morph into proto-testes deep in the abdomen (in females, these same structures go on to form ovaries). Prior to birth, testes descend down through the abdomen and into the scrotum in a two-step process. This downward migration is set into motion through the action of two key genes , INSL3 and RXFP2 . Deleting either of these “scrotal genes” in mice completely derails the testes’ southbound trajectory.

The few human males who are born with undescended testicles (between 2 and 4 percent) can be in for trouble: if the condition persists into adulthood, it can contribute to hernias, infertility and testicular cancer. But for a whole different group of mammals, having no scrotum at all is the norm. Mammalian species of the Afrotheria clade—which includes elephants, manatees, cape golden moles and rock hyraxes—instead retain their testicles inside the abdomen in a condition called “testicondy.”

These diverging traits have posed a puzzle for evolutionary biologists: Did the common ancestor of all living mammals, like Afrotherians and female mammals, kept its valuable reproductive organs inside its abdomen? Or, like humans and most other mammalian lineages, did it carry them outside its body for all to see? Because soft tissues like testicles preserve poorly in the fossil record, no physical evidence of ancestral testes remain, and the location of ancestral mammalian testes has remained elusive.

Lead author Virag Sharma and senior author Michael Hiller , genomicists at the Max Planck Institute of Molecular Cell Biology and Genetics in Dresden, Germany, took a genetic approach to the debate. Knowing how critical the scrotal genes were for testicular descent, they reasoned that comparing these genes in a variety of mammalian lineages would provide the most direct route to pinpointing the ancestral state. (This method bypasses the limitations of the fossil record, which can sometimes produce conflicting or vague information about relatedness between species.)

“Being able to use molecular data to answer a question like this is something we weren’t able to do 10 years ago,” says Smithsonian National Zoo genomicist Natalia Prado-Oviedo , who was not affiliated with the study. Importantly, Sharma and Hiller’s method “works [with any interpretation of the fossil record].”

When Sharma compared scrotal genes in 71 mammals, he found that four Afrotherian species lacking descended testicles— manatees , cape golden moles , cape elephant shrews and tenrecs (tiny insect-chomping mammals that resemble hedgehogs)—all carried defunct copies of the scrotal genes . Sharma then used this genetic information to approximate when one of the genes had lost functionality in each species. When genes become nonfunctional, there is no longer pressure to maintain coherence, and they begin to decay and accumulate mutations out of neglect. The more errors a gene sequence carries, the longer ago it is likely to have been lost.

By working backwards, Sharma also tracked the loss of testicular descent in all four species to 23-83 million years ago—instances all more recent than the estimated divergence of the Afrotherian lineage 100 million years ago. Unlike other mammals, as Afrotherians split off from the main pack, their testes failed to do the same.

Sharma also found that the types of genetic errors found in these four species all differed from each other, and apparently appeared at separate points in time. Had they been identical mutations, Sharma would have inferred that a single ascrotal ancestor had passed the same broken genes onto all four species at once. But the variation showed that scrota disappeared on four separate occasions over the course of evolutionary history. In other words, evolution “independently invented” undescended testes four times.

Scientists have known for years that one of the most important benefits of scrota is ventilation : mammalian sperm matures and stores better at temperatures 2.5 to 3 degrees Celsius lower than the rest of the body, and jettisoning these organs keeps them cool. But we are far less sure if this is the reason that scrota evolved. It’s a classic rooster-and-egg dilemma: testes may have fled the abdomen because temperatures got too toasty, or sperm may have adapted to love the chill because they had already been ousted for some other reason.

(Other theories abound , including the idea that testes are ornaments that boast male virility . Or maybe, as pediatric urologist John Hutson believes, testes were expelled as a byproduct—or mistake—of another anatomical rearrangement.)

But if temperature is the main factor, then there’s still one puzzle researchers have to answer. Elephants and elephant cape shrews—which both keep their testes locked away in the abdomen—have internal body temperatures similar to those of humans. How do they cope?

In Afrotherians, the costs (exposure, vulnerability) could simply outweigh the benefits (slightly cooler temperature), says Hiller. Or perhaps these mammals use a yet-undiscovered method of maintaining their chill. To tether these phenomena to testicular retention, geneticists will likely have to join forces with physiologists.

“We can’t rely on just genetics or just the fossil record alone,” says computational biologist Melissa Wilson Sayres of Arizona State University. “Genomics is powerful, but we need to understand it in concert with natural history and anatomy.”

For now, the rest of us are left hanging.