Assigned ID: 923 ----- Attempting Submission 1 (Fri Feb 24 23:40:45 GMT 2012) ----- There exist planetary nebulae (PNe) whose heavy element C, N, O, and Ne abundances as derived from optical recombination lines (ORLs) are a factor more than ~5 higher than those derived from the traditional method based on collisionally excited lines (CELs). This ratio is called the abundance discrepancy factor (adf). A promising proposition to explain this long-standing nebular abundance problem posits that these nebulae contain (at least) two distinct regions - one of "normal" electron temperature, Te (~10000 K) and chemical composition (~solar) and another of very low Te (< 1000) that is H-deficient, thus having high metal abundances relative to H. The latter component emits strong heavy element ORLs and IR fine-structure (FS) CELs, but essentially no optical/UV CELs. Efforts to directly detect these inclusions in PNe have been unsuccessful to date. However, there is mounting circumstantial evidence for their existence, such as presented in our recent paper that modeled the high-adf PN NGC 6153 using a 3-D photoionization code. The models that included the low Te, H-deficient knots fit most observations far better than did those models without the clumps. It has been shown that the adf varies with position in a PN and is highest close to the central star. The very low Te inclusions must be cooled predominantly by FS mid-IR lines. We propose to use HST archival images to derive [O III] 5007 A flux maps to compare with the [O III] 88 micron fluxes from our Herschel observations of four PNe - NGC 2392, NGC 2440, NGC 6720 and NGC 7009 - all on the largest adf list, to find if the IR line flux relative to the cospatial optical forbidden line flux peaks where the adf peaks. ISM AND CIRCUMSTELLAR MATTER 12 Small false 0.0 0.0 false false false false SUPPORTED