I would appreciate if you consider the following two questions on $1$ dimensional foliations whose leaves are geodesic.

1)Assume that $M$ is a Riemannian manifold which is either an open manifold or is a compact manifold with zero Euler characteristic. Does $M$ admit a foliation by geodesics?

2)Assume that $M$ is a Riemannian surface which admit at least one foliation by geodesics. Does there necessarily exist a foliation of $M$ by geodesics which satisfy the "Isocline Locale property"?

The Isocline local property is defined as follows:

For every $x\in M$ there is locally a geodesic $\alpha $ which is transverse to the foliation and it intersect all leaves with the same angle.


No to the first question. Let $M$ be a Riemannian $2$-manifold whose universal cover is the hyperbolic plane $H$, and whose fundamental group is not cyclic. For any foliation of $H$ by geodesics (lines), it seems to me that the endpoints of these lines on the boundary circle of $H$ will fill up the whole circle except for two points. These two points will have to be preserved by deck transformations, so any discrete group of isometries of $H$ that preserves the foliation must be cyclic.

  • 1
    $\begingroup$ But for these surfaces the Euler characteristic is not zero (which is indeed a necessary condition for a codimension 1 foliation, be it geodesic or not). $\endgroup$ – ThiKu Jun 16 '18 at 16:02
  • 1
    $\begingroup$ @Thiku: The $M$ in this answer is an open manifold. $\endgroup$ – Lee Mosher Jun 16 '18 at 17:59
  • 1
    $\begingroup$ Yes, I meant to say choose it to be open, for example the sphere minus three points (with constant curvature metric). $\endgroup$ – Tom Goodwillie Jun 16 '18 at 18:58
  • $\begingroup$ To be honest I can not understand this answer. For example consider $M=H\setminus \{p,q\}$. Then is not $H$ the universal covering space of $M$? But M admit a foliation by geodesic. The vertical foliation. Am I mistaken? $\endgroup$ – Ali Taghavi Jun 17 '18 at 10:54
  • 2
    $\begingroup$ This $M$ is not complete, as it would have to be if its universal covering space were to be metrically isomorphic to $H$. $\endgroup$ – Tom Goodwillie Jun 17 '18 at 11:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.