1: \begin{abstract}

2: We study the structure of the no-core shell model wave functions for $^6$Li and

3: $^{12}$C  by investigating the ground state and first excited state electron scattering

4: charge form factors.

5: In both nuclei, large particle-hole ($ph$) amplitudes in the wave functions appear with

6: the opposite sign to that needed to reproduce the shape of the $(e,e')$ form factors,

7: the charge radii, and the B(E2) values for the lowest two states.

8: The difference in sign appears to arise mainly from the monopole 

9: $\Delta\hbar\omega=2$  matrix elements of the kinetic and potential energy (T+V)

10: that transform under the harmonic oscillator SU(3) symmetries as $(\lambda,\mu)=(2,0)$.

11: These are difficult to determine self-consistently, but

12: they have a strong effect on the structure of the low-lying states

13: and on the giant monopole and quadrupole resonances.

14: The Lee-Suzuki transformation, used to account for the restricted nature of the space in terms of an effective interaction, introduces large higher-order $\Delta\hbar\omega=n, n>$2, $ph$

15: amplitudes in the wave functions.  The latter $ph$ excitations

16:  aggravate the disagreement between the experimental and predicted $(e,e')$ form factors

17: with increasing  model spaces, especially at high momentum transfers.

18: For sufficiently large model spaces the situation begins to 

19:  resolve itself for $^6$Li, but the

20: convergence is  slow. A prescription to constrain the $ph$ excitations would likely

21: accelerate convergence of the calculations.

22: \end{abstract}

23: