1: \begin{abstract} We suspect that the ``genus part'' of the class 

2: number of a number field $K$ may be an obstruction for an ``easy 

3: proof'' of the classical $p$-rank $\varepsilon$-conjecture for $p$-class 

4: groups and, a fortiori, for a proof of the ``strong $\varepsilon$-conjecture'': 

5: $\order (\Cl_K \otimes \Z_p) \ll_{d,p,\varepsilon} 

6: (\sqrt{D_K}\,)^\varepsilon$ for all $K$ of degree $d$. 

7: We analyze the weight of genus theory in this inequality by means of an

8: infinite family of degree $p$ cyclic fields with many ramified primes, then 

9: we prove the $p$-rank $\varepsilon$-conjecture: 

10: $\order (\Cl_K \otimes \F_p) \ll_{d,p,\varepsilon} 

11: (\sqrt{D_K}\,)^\varepsilon$, for $d=p$ and the family of degree $p$ cyclic  

12: extensions (Theorem \ref{thmf}) then sketch the case of arbitrary 

13: base fields. The possible obstruction for the strong form, in the degree $p$ 

14: cyclic case, is the order of magnitude of the set of ``exceptional'' $p$-classes 

15: given by a well-known non-predictible algorithm, but controled thanks to 

16: recent density results due to Koymans--Pagano. Then we compare the 

17: $\varepsilon$-conjectures with some $p$-adic conjectures, 

18: of Brauer--Siegel type, about the torsion group ${\mathcal T}_K$ 

19: of the Galois group of the maximal abelian $p$-ramified 

20: pro-$p$-extension of totally real number fields $K$. We give 

21: numerical computations with the corresponding PARI/GP programs.

22: \end{abstract}

23: