Elementary abelian subgroups in p-groups of class 2

All the results in this work concern (finite) p-groups. Chapter 1 is concerned with classifications of some classes of p-groups of class 2 and there are no particularly new results in this chapter, which serves more as an introductory chapter. The "geometric" method we use for these classifications differs however from the standard approach, especially for p-groups of class 2 with cyclic center, and can be of some interest in this situation. This "geometry" will for instance, prove to be particularly useful for the description of the automorphism groups performed in Chapter 3. Our main results can be found in chapters 2 and Chapter 3. The results of Chapter 2 have a geometric flavour and concern the study of upper intervals in the poset Ap(P) for p-groups P. We already know from work of Bouc and Thévenaz [8], that Ap(P)≥2 is always homotopy equivalent to a wedge of spheres. The first main result in Section 2.4, is a sharp upper bound, depending only on the order of the group, to the dimension of the spheres occurring in Ap(P)≥2. More precisely, we show that if P has order pn, then H~k(Ap(P)≥2) = 0 if k ≥ ⎣n-1/2⎦. The second main result in this section is a characterization of the p-groups for which this bound is reached. The main results in Section 2.3 are numerical values for the number of the spheres occurring in Ap(P)≥2 and their dimension, when P is a p-group with a cyclic derived subgroup. Using these calculations, we determine precisely in Section 2.5, for which p-groups with a cyclic center, the poset Ap(P) is homotopically Cohen-Macaulay. Section 2.7 is an attempt to generalize the work of Bouc and Thévenaz [8]. The main result of this section is a spectral sequence E1rs converging to H~r+s(Ap(P)>Z), for any Z ∈ Ap(P). We show also that this spectral sequence can be used to recover Bouc and Thévenaz's results [8]. In Section 2.8, we give counterexamples to results of Fumagalli [12]. As an important consequence, Fumagalli's claim that Ap(G) is homotopy equivalent to a wedge of spheres, for solvable groups G, seems to remain an open question. The results of Chapter 3 are more algebraic and concern automorphism groups of p-groups. The main result is a description of Aut(P), when P is any group in one of the following two classes: p-groups with a cyclic Frattini subgroup. odd order p-groups of class 2 such that the quotient by the center is homocyclic.