Science: If there’s one thing you can be sure about with particle accelerators, it’s that they’re expensive to build. The €3 billion Large Hadron Collider at CERN is the most extreme example. But even at the other end of the scale, a hospital that wants an accelerator for proton beam therapy for cancer patients will likely have to fork out more than $100 million, and neither of the two most common existing technologies—cyclotrons and synchrotrons—is well-suited to the task. Now a handful of accelerator physicists are experimenting with a new type of machine—a cross between a cyclotron and a synchrotron—that avoids many of the shortcomings of both and is simpler and cheaper to build.