where 2L1 is the level spacing of two-state spin described by the Pauli matrix 6z and 6±=6x±i6y, w; is the excitation energy of boson with quantum number j, and v = 'LJ;(J..;/w;) · (bjb;). This is the unitary transformed version offamiliar H =-Ll· 6x +LJ;w;b;+b;+1/2·6z·'LJ;J..;(b;++b;)-LJ;J..]/4w; by unitary operator U=exp[6z·v/2]. The IS means that fi and H are invariant under inversion operation P( 6x, 6y, 6z, b;, bt)P-=(6x, -(Jy, -6z, -b;, -b;+), where P=exp[i7r'LJ;b/b;+i(.n/2)(6x-1)]. The spectral density of interaction is assumed to be 'LJ;J../8(w;-w)=2aw·(w/wcY-r Xexp[ -w/wc] where s>O and a is a dimensionless strength. Boson system fiB = 'LJ;w;b; + b; in (1) consists of bosons which displace themselves according to flips of spin. We assume that the boson system is thermodynamically large and white in the sense that cutoff wc')>Ll. We choose the eigenstates of 6z for those of spin. In previous worksl,ZJ the author proposed the concept of dynamic compensation, and predicted that the ground state (GS) in the white limit exhibits the transition at ac=1/2 from the delocalized (tunneling) one in a ac for s = 1, the localized one over the whole a except zero for s 1. The first two results of GS are due to the dynamic compensation of infrared divergence. There is no infrared divergence in s > 1 from the outset. This prediction is different from the paradigmatic description based on the static renormalization of infrared divergence.l,l In this paper we present the rigorous results of the two-branch structure of GS, excited state and their energies, thermodynamics and dynamics, restricting ourselves to the case s=1 in order to save space. The full details and other cases of s will be reported elsewhere.