In section 3.2, the beam areas that you are using for the SPIRE data are not the latest values. I'm told what we should be using is 423, 751, 1587" from ftp://ftp.sciops.esa.int/pub/hsc-calibration/SPIRE/PHOT/Beams/beam_release_note_v1-1.pdf The biggest difference is at 500 microns and is about 2.5%. These areas are what we are using in the papers from my project that we are currently working on. These are beam areas calculated from 1" pixel maps. I'm told that this is the beam area to use for conversion to MJy/sr or Jy/pixel, even if your pixel size in your map is not 1". *** Fixed start of Section 3.1 - were all the PACS images done with 15' long scan maps? None of the galaxies were bigger than this? *** No, they were between 10' and 34' long. Don't know where I got that misinformation about 15' leg lengths. Thanks! end of Section 3.2 - I already sent you a correction to the beam areas for SPIRE *** Fixed end of Section 4.1 - It wasn't obvious to me how Figure 4 shows the galaxies are not a one parameter family. What should I be looking at. *** Just that the data points don't follow a single trend (it appears to fork toward the upper and lower right corners). start of section 4.2 - in looking at the SEDs, I didn't always agree with your comments about how the SCUBA data didn't work well. Particularly, I thought the SCUBA data looked good for NGC 4579 and NGC 1097, and worse for NGC 4536 than either of these. Is there some way to quantify when the SCUBA data seem to over or underestimate the SED prediction significantly? *** I agree that N4536 does a poorer job than N4579 and N1097 of matching the extrapolation to 850um, but my discussion here was to point out known shortcomings to the SCUBA data for some of the galaxies. Section 4.6 - I had a similar reaction to the discussion of the 500 micron excess as to the SCUBA data - this needs to be quantified better. For example, NGC 3049 seems to have more of an excess to my eye than Holmberg II, for which the 500 micron point agreed pretty well. Basically my reaction was if Ho II has an excess, then many more galaxies beyond the 4 you single out do as well. *** I agree that more galaxies than just the dwarf/irregulars show an apparent 850um excess, but these galaxies consistently show the apparent excess. However, I've quantified things now and restructured the text quite a bit. I also now mention N3049 and N5474 as two spirals that show similar excesses. Thanks for pushing me to make this better! Section 4.4 - I thought it was interesting that in Figure 7 the cooler galaxies often show more of a systematic offset than the warmer galaxies i.e. the dust mass is lower, and Umin, q(Pah) and maybe even gamma are all higher when Herschel data are included. Looking at Figure 8, I wanted to know which galaxies had which 70/160 micron colors; could the points be color-coded? Section 4.5 - I found this really interesting because we have also been doing a variety of dust mass fits. In a couple of the galaxies from my sample, we are getting beta more like 2 (or even 2.2) when we allow it to vary (M83 and Cen A). *** I look forward to seeing your results! I found it very interesting that when you fixed beta=2 and included 70 microns that the dust temperatures got *lower*. My impression was that if 70 is included it can drive the T higher. Have you checked to see if the reason the fits give a lower T is because the fits that don't include the 70 micron data are over-predicting the 70 micron flux? I sometimes see that in my galaxies, in which case we include the 70 micron data point in the fit. *** My interpretation: beta=2 is overly steep (that's higher than any of the fitted values when beta is allowed to vary), and thus the BB Tdust must drop in order to better fit the submm data. I've made a note of this in the text. I couldn't see how Figure 11 was supposed to show me that the temperature was overestimated by the modified BB fit, as there was no temperature given for DL07 fit - what am I supposed to be looking at here? *** There is hot dust appearing shortward of the BB fit, and thus the fit overestimates the dust temp. DL07 doesn't yield a Tdust, so the comparison isn't direct. In general I'm rather unsure how to compare dust mass estimates from the Draine & Li model with the modified black body. My postdocs have not been able to figure out exactly how the mass is calculated in DL07 and I haven't had time to look myself. All I'm saying is I'm not 100% sure that a difference of a factor of 2 in dust mass proves that one or the other model is wrong. Figure 10 - Is this plot for 100-500 micron fits or 70-500 microns or both? *** 100-500um. The caption has been updated. Thanks.