Section 2.1: why `only' 50 galaxies? Don't we already have data for
all 75?  Granted, things will not change much, but this would make
this paper the `final SINGS word' on the issue.

--All of the nuclear spectroscopy has been taken by now, but not all
  of the data has passed through the processing pipeline and been made
  available to us (the last nuclear spectra were taken Feb 3, 2006).
  In addition, getting the spectroscopy data to a usable state still
  requires a lot of effort--so far we have only post-pipeline
  processed the 50 nuclear sources in DR3.

Section 3.3: although not explicitly stated, I assume you are implying
that, unlike the HighRes, background spectra are obtained for the
LowRes (?). I am worrying about the 6.2 um PAH EW derivation....

--Correct.  We subtract the sky for LowRes spectra.

Section 4.1, end. Would it be worth to list somewhere (maybe as a
footnote in Table 1) the 8 undetected galaxies?  At least there
will be a record of this characteristic/problem somewhere.

--Done.

Section 5.1: any merit in considering to add a diagnostic plot showing
the actual nuclear spectra we have (I recall they are 2.5"x2.5" or the
like, centered on the galaxy's nucleus).  It can help gauge what we
have or miss when we use line ratios with much larger apertures.

--I originally had the plot in the paper, but the consensus now is to
  focus on the plot that utilizes 20"x20" apertures.

Second paragraph of section 5.1: Any wisdom on mis-classifications (or
insufficient classifications) coming from Ho et al. 1997, ApJS, 112,
315?  There is an overlap of 22 galaxies between that paper and the
Table 1 list, and most of the galaxies listed at the end of the second
paragraph are contained in Ho et al.'s survey.  We could reference this
paper as well, for information about nuclear types.

--You raise a good point.  But I would prefer to focus on the spectra
  and classifications taken and derived by Moustakas and Kennicutt.  We
  know that John and Rob used the same coordinates as our SINGS 
  pointings, and the they obtained optical spectra for far more than
  just 22 SINGS targets.
    
Third paragraph of section 5.1: technically, a die-hard-IR-guru will
tell you that Ha/Hb is not enough to probe the entire dust depth, thus
Ha/Hb will only give you an underestimate on the extinction in a region.
You may want to corroborate your statements in this paragraph by adding
a few sentences about the fact that except for very few cases, we haven't
seen heavily buried objects when *comparing Halpha observations with the
Spitzer 24 micron images* (e.g., Calzetti et al. 2005, and also the work
Rob has been doing with Janice???).

--Helene also pointed out something along these lines.  I have added
  a line pointing out the consistency with prior SINGS work involving
  comparisons between H-alpha and 24um data.

Page 9, 4th para: do you worry (or need to worry) about possible
effects on the PAH emission from metallicity variations in galaxies?

--Good question.  However, all of our low metallicity objects do not 
  appear in Figure 3 due to lack of data/detections.

Figure 3: A short explanation in the figure caption of the roman
numbers in the two panel will make the figure clearer.  Or just provide
a reference to Table 4.

--Done.