1: \begin{abstract}

2: We explore the hypothesis that the set of symmetries enjoyed by the

3: theory that describes gravity is not the full group of diffeomorphisms

4: (\emph{Diff(M)}), as in General Relativity, but a maximal subgroup

5: of it (\emph{TransverseDiff(M)}), with its elements having a jacobian

6: equal to unity; at the infinitesimal level, the parameter describing

7: the coordinate change $x^{\mu}\rightarrow x^{\mu}+\xi^{\mu}(x)$ is

8: transverse, i.e., $\partial_{\mu}\xi^{\mu}=0$. Incidentally, this

9: is the smaller symmetry one needs to propagate consistently a graviton,

10: which is a great theoretical motivation for considering these theories.

11: Also, the determinant of the metric, $g$, behaves as a {}``transverse

12: scalar'', so that these theories can be seen as a generalization

13: of the better-known unimodular gravity. We present our results on

14: the observational constraints on transverse gravity, in close relation

15: with the claim of equivalence with general scalar-tensor theory. We

16: also comment on the structure of the divergences of the quantum theory

17: to the one-loop order. 

18: \end{abstract}

19: