Killed by its own obituaries: Explaining the demise of the ether

In this paper I follow the demise of the ether in the first half of the twentieth century to show how the first obituaries of the ether were instrumental in creating an object with specific and largely simplified properties related to, but different from, nineteenth-century ethers. I suggest that writing the history of dead objects (or objects an author wants to be dead) is not epistemologically neutral but, on the contrary, it involves a reformulation of the object itself. I show that this was indeed the case with the ether: those arguing for its demise in the early twentieth century tended to overlook as irrelevant one of the ether's most important properties, namely being the seat for the transmission of electromagnetic waves. Instead, they emphasized the contradictions between other properties of previous ether(s), so as to advocate for its disappearance.