Let L <img src="http:/img/fbpe/aabc/v73n1/0059c.gif"> <img src="http:/img/fbpe/aabc/v73n1/0059c2.gif"> be a real 3 dimensional analytic variety. For each regular point p <img src="http:/img/fbpe/aabc/v73n1/0059e.gif"> L there exists a unique complex line l p on the space tangent to L at p. When the field of complex line p <img ALIGN="MIDDLE" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img4.gif" ALT="$\displaystyle \mapsto$"> l p is completely integrable, we say that L is Levi variety. More generally; let L <img src="http:/img/fbpe/aabc/v73n1/0059c.gif"> M be a real subvariety in an holomorphic complex variety M. If there exists a real 2 dimensional integrable distribution on L which is invariant by the holomorphic structure J induced by M, we say that L is a Levi variety. We shall prove: Theorem. Let <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img5.gif" ALT="$ \cal {L}$"> be a Levi foliation and let <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img6.gif" ALT="$ \cal {F}$"> be the induced holomorphic foliation. Then, <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img6.gif" ALT="$ \cal {F}$"> admits a Liouvillian first integral. In other words, if <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img5.gif" ALT="$ \cal {L}$"> is a 3 dimensional analytic foliation such that the induced complex distribution defines an holomorphic foliation <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img6.gif" ALT="$ \cal {F}$">; that is, if <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img5.gif" ALT="$ \cal {L}$"> is a Levi foliation; then <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img6.gif" ALT="$ \cal {F}$"> admits a Liouvillian first integral--a function which can be constructed by the composition of rational functions, exponentiation, integration, and algebraic functions (Singer 1992). For example, if f is an holomorphic function and if theta is real a 1-form on <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img8.gif" ALT="$ \mathbb {R}$">; then the pull-back of theta by f defines a Levi foliation <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img5.gif" ALT="$ \cal {L}$"> : f*theta = 0 which is tangent to the holomorphic foliation <img ALIGN="BOTTOM" BORDER="0" src="http:/img/fbpe/aabc/v73n1/0059img6.gif" ALT="$ \cal {F}$"> : df = 0. This problem was proposed by D. Cerveau in a meeting (see Fernandez 1997).
Determinacy of Analytic Foliation Germs
