1: \begin{abstract}

2:  The most common spin foam models of gravity are widely believed to be

3:  discrete path integral quantizations of the Plebanski

4:  action. However, their derivation in present formulations is

5:  incomplete and lower dimensional simplex amplitudes are left open to

6:  choice. Since their large-spin behavior determines

7:  the convergence properties of the state-sum, this gap has to be

8:  closed before any reliable conclusion about finiteness can be

9:  reached. It is shown that these amplitudes are directly related to

10:  the path integral measure and can in principle be derived from

11:  it, requiring detailed knowledge of the constraint algebra and

12:  gauge fixing.

13: % of its first class part which in the

14: % case of gravity generates space-time diffeomorphisms. 

15: % It has been

16: % suggested that the discretization of space-time in a spin foam model

17: % breaks the diffeomorphism gauge without introducing an explicit gauge

18: % fixing. Here we show that 

19:  In a related manner, minimal requirements of background

20:  independence

21: %--which are reminiscent of cylindrical consistency in

22: % loop quantum gravity--

23:  provide non trivial restrictions on the form of

24:  an anomaly free measure. Many models in the literature do not satisfy

25:  these requirements.

26: % Moreover, we show that an anomaly free model will

27: % necessarily contain divergent amplitudes that could be interpreted as

28: % due to infinite contributions of gauge equivalent

29: % configurations. Exploring these issues we come across a 

30:  A simple model satisfying the above consistency requirements is

31:  presented which can be thought of as a spin foam quantization of the

32:  Husain--Kucha\v{r} model.

33: \end{abstract}

34: