The current disconnect between the frontiers of theoretical physics and traditional practice of validating theory with experiment and predicted observation makes exciting reading hard to come by for physics addicts; myself, philosophers, science historians and experimentalists alike. To loosen up I recommend reading Helge Kragh’s recent well-written and scholarly tale of physics follies, both ancient and modern. Kragh discusses follies in which the theoretical envelope has often been pushed too far without any predictive support, as may also be the case today. His book is called Higher Speculations, subtitled Grand Theories and Failed Revolutions in Physics and Cosmology. (Oxford University Press, 2011). Anyone who has read and enjoyed Martin Gardner’s older and humbler Fads and Fallacies in the name of Science , which describes more ordinary follies, will, I think, appreciate Kragh’s irreverent volume. It shows how far off the rails of reality even admirably intelligent and competent folk can ride --- famous folk like Gerald FitzGerald (he of the eponymous length-contraction); J.J.Thomson (of electron fame); Hilbert (who invented now-discarded world equations), Eddington (who helped to put general relativity on the map but was also a bit of a crackpot) --- and many others, some still going strong today. The follies of the great and famous may encourage lesser folk (like myself) to speculate, but Kragh’s book also makes one wonder if some of the old, now unfashionable, ideas put forward by clever physicists should not be revisited now and then, however crackpot they now seem. For instance the idea that circular motions in fluids (a.k.a. vortices) lead to structure formation has long since been discarded on a microscopic scale (like the vortex theory of atoms). But it is now becoming fashionable, I believe, to consider the role vortices play in structure formation in the early solar system among planetisimals and other dusty stuff. Then there is the fixation in general relativity (because it works so superbly on large scales?) about using the real number continuum on which to base its model geometry. However secondary geometries in structures based on discrete substrates show various kinds of discontinuities – I’m thinking here of glissile defects based on a discrete lattice substrate and also of defects in moiré fringes, seen when patterns that are repetitive but not entirely regular are superimposed. Geometry includes distortions other than smooth curvature, a fact which might be worth exploring just in case nature is discontinuous on some scales. Or one might take a very obvious feature of light, namely; that it travels very fast, and suppose that it is a wave travelling through something very rigid indeed. We now know that in a solid body freely glissile line defects are unable to move faster than sound. For fancy’s sake imagine that such defects were sentient and communicated with sound waves. They mightthen devise protocols like those of special relativity to measure distance and time with sound, while they treating the speed of sound as an invariant limiting speed at which inertial mass becomes infinite. They might even invent a name for their rigid solid like the forbidden E-word. Of course this is just a faulty (spot the flaw!) fancy (because light is a particle, right?) stimulated by my observing how surface waves in water can spread very much faster than normal over water when caused by tapping a thin, partially formed surface ice-layer with a stick. Rigidity is a great wave-speed accelerator. Kragh's analysis hints that wisdom and folly are often two sides of the same coin.