The 20:th century has taught us that physics not always is trajectories. Nevertheless, the description of natural phenomena in terms of trajectories has persisted very widely to this day. A Cartesian coordinate system equipped with trajectories provides no explanation of the geometry of atoms, other microscopic structures, or of the universe. The geometry of the universe is as little understood today as that of the Earth was before Galilei, particularly the problem of its extension and why its mass is scattered into particles. Bending the Cartesian coordinates by using special and general relativity is ignorant of the origins of space and time which, in standard models, together with the entropy and energy are referred to a "singularity" in which the "laws of Nature break down". Nor does relativity explain the difference of quality between progressing time and static space except by resorting to the additional hypotheses and experimental observation of the apparent cosmological expansion. Much attention has been paid to the causality principle embedded in relativity that information along trajectories about elsewhere events not may be transmitted before they occur. However, the stronger causality principle that observations in the laboratory frame only can be made in present time has been comparatively unnoticed in physics. In quantum mechanics, maintaining the concept of trajectories has lead to a situation in which non-local phenomena are ascribed to some probability of a wave having spatial extension while leaving the true nature of non-locality largely unresolved. The successful description of these phenomena in terms of existing mathematical language does not preclude that a deeper, yet quantitative, understanding of them remains possible.