"It may be helpful to a good understanding of the conception of the physical universe implied by the general theory of relativity, to consider the different definitions of a straight line. ...In the old mechanics, there are four of these, viz.: (1) ray of light, (2) the track of a material particle not subject to any forces, (3) a stretched cord, (4) an axis of rotation. The fourth definition is the one favored by the great mathematician Henri Poincaré. ...Are they still identical in the theory of relativity? The definitions 1 and 2 define the straight line as a projection on the three-dimensional space x, y, z of a geodesic in the four-dimensional space-time continuum. This projection will be a geodesic in three-dimensional space only under very special conditions. In the general case the two projections will differ from each other, and neither of them will be a geodesic. Also the projection may be a geodesic in one system of coordinates but not in another. The stretched cord is by definition a geodesic in the three-dimensional space. As a rule, this will not be a geodesic in the four-dimensional continuum. The rotation axis is also by definition a line in three-dimensional space. The definition, however, presupposes the possibility of the rotation of a rigid body, which would be possible only in a homogeneous, isotropic, and statical field, i.e., in a world without any material bodies... in it, which by their gravitational field would upset the isotropy. The definition is thus meaningless in the general theory of relativity."