Evaluating clusters of variables produced by varclus & hclustvar

Background - I want to cluster analyze a mixed dataset, clustering the variables on the basis of correlational similarity. SPSS gives me this option, but doesn't allow me to evaluate the clustering solutions by providing statistical measures of heterogeneity change (e.g. pseudo F statistic) or direct measures of heterogeneity (e.g. CCC).

As such, I'm learning R. I've managed to cluster my variables using both the varclus and hclustvar procedures, both of which generate nice dendrograms that I can interpret visually. However I'm struggling to get R to provide me with some actual numbers, such as the statistics mentioned above, which might indicate what constitutes the "best" clustering solution. How can I do this? I've been through the documentation for both the Hmisc and ClustOfVar packages and can't find any way to do this.

I've read elsewhere that clustering criteria can be applied using the NBClust package, but as far as I can see the NBClust function only works on dissimilarity matrices / distance measures, which aren't an option for me as I'm interested in correlational similarity.

Any suggestions?