If you were a sophisticated and up-to-the-minute science buff in 17th century Europe, you knew that there was only one properly scientific way to explain anything: "the direct contact-action of matter pushing on matter," (as Peter Dear puts it The Intelligibility of Nature). Superstitious hayseeds thought that one object could influence another without a chain of physical contact, but that was so last century by 1680. Medieval physics had been rife with such notions; modern thought had cast those demons out. To you, then, Newton's theory of gravity looked like a step backwards. It held that the sun influenced the Earth without touching it, even via other objects. At the time, that just sounded less "sciencey" than the theories it eventually replaced.
This came to mind the other day because, over at Edge.org, Richard H. Thaler asked people to nominate examples of "wrong scientific beliefs that were held for long periods." He also asked us to suggest a reason that our nominee held sway for too long.
My nominee was that notion that only physically touching objects were allowed in legitimate scientific explanations. It endured because in the context of recent history, it looked more serious and rigorous than what came before. A lot of theories have hung on like that—because their rivals look old-fashioned or naive or simply chaotic. More than once I've heard people defend the old rational-economic-man paradigm on the grounds that, flawed though it be, it's rigorous and formal and lets people think clearly. Another example of how the science-ish is the enemy of actual science? Time will tell.
Anyway, there are many interesting candidates over there (I liked, for example, the notion that the brain had nothing to do with thought or feeling, and the notion that we understand the interactions we call "nature and nurture." The list, with 62 entries, is worth checking out, in my of course unprejudiced opinion.