Dear John,

We are relieved to finally resubmit our manuscript ApJL #21952 in 
slightly modified form with this response.  We have addressed these 
final comments and sincerely hope that our manuscript can be accepted 
in its current form. 

Best wishes,
Michael Meyer

*****************Response to the Referee*************************************

We thank the referee for her/his most recent comments.  Our responses 
follow below.  In addition to addressing these comments, we have
removed the phrase "...broadly consistent with models of 
terrestrial planet formation and the inferred evolution of
our solar system" from the abstract given the complications
in comparison our results to currently available models. 
In addition, we have added one sentence to the last paragraph in section 4:

"These results can be compared to the observed
frequency of gas giant planets of at least 6.6 \% within 5 AU
(and 12 \% when extrapolated out to 20 AU \citet{Marcy05}; see
also \citet{Lafreniere07})
and the suggestion that extant Spitzer data are consistent with
most stars having debris disks comparable to our
solar system \citet{Bryden06}."

REFEREE:  The weakness in the argument is primarily in the fact that the
Kenyon & Bromley terrestrial planet formation models which are being invoked
to justify the excess being detectable for a factor ~3 in age actually
produce detectable 24micron excesses for much longer than a factor of 3 in
age. The reply about the model results only being pertinent to 10 and 20micron
excesses is not valid. While 10 and 20micron evolution was plotted in
Kenyon & Bromley (2004), Fig. 4 of Kenyon & Bromley (2005) shows the
24micron evolution in which detectable emission is seen for a factor of
>100 in age for a range in radii of x2. The long duration of detectable excess
is due in part to the scaling of sigma as a^-1.5, but mostly because emission
in the Kenyon & Bromley models is not just detected at the "era of maximum
dust production", but for a long period afterwards as the small bodies are
depleted.

RESPONSE:  We have attempted to clarify in the text that our 
estimate for the duration of 24 micron excess emission does 
not follow solely from the detailed models of Kenyon and
Bromley, but more generally from consideration of the 
evolution of planetesimal swarm.  Though we do acknowledge
that our basic scaling does assume that we are detecting
dust near the era of maximum dust production.  Because the
models of Kenyon and Bromley predict maximum excesses x5 larger
than observed, it is difficult to scale their predicted duration
of the 24 micron excess phase.  We believe this may be due to 
the fact that their published models are tracing hotter dust
at smaller radii.  We have verified with the authors that
the results in Figure 4 of Kenyon and Bromley (2005) are
comparable to those shown in Figure 3 of Kenyon and Bromley 
(2004) for disks that range from 0.68-1.32 AU with excess
emission indicated at different wavelengths.  The model that 
shows 24 micron excess in Fig 4 (KB05) likely also shows 
excess at shorter wavelengths at early times due to dust
generated by planetesimals at the smaller radii (< 1 AU). 
As a result the duration of 24 micron excess shown in 
Fig 4 (KB05) is longer than expected if the models only
considered planetesimal belts outside of 1 AU.  Such models
are not currently available.  Because our sources lack excess 
emission at 8 microns, we conclude that they lack dust producing 
planetesimals inside of 1 AU.  We now state this explicitly in 
the second paragraph of section 4, and reference Silverstone 
et al. (2006) in support of this claim.  We have also added a 
footnote explaining the discrepency in timescales:

"While the published Kenyon and Bromley models
predict excess emission much longer than a factor of $\times$ 3--6,
we note that those models also predict hot dust at smaller radii
covering a wider range of radii than our observations imply."

A fuller discussion of these points awaits more detailed 
analysis of the dust temperatures as described in paragraph
4 of section 4. 

REFEREE:  I don't think this should prevent the paper being published, since 
the Kenyon & Bromley models include many assumptions and may not provide an
accurate picture of the dust production during terrestrial planet formation.
However, this does invalidate the current argument, the wording of which would
have to be changed for this to be acceptable. I think the problem is in the
sentence "This simple calculation...", since the authors have made a further 
assumption which needs to be mentioned, which is that 24micron emission is
only detectable at the epoch of maximum dust production at each radius. Also,
whatever the "simple calculation" is, it does not suggest that the "emission 
we observe SHOULD not persist" rather that it "MIGHT not persist". And 
there should be a caveat recognising that 24micron emission actually persists 
on timescales longer than x3 in age in the Kenyon & Bromley (2005) models. 
In this form the argument would be still be very weak, but would be sufficient 
to justify the "Perhaps many stars..." and "Nevertheless, one might consider..."
elements of the rest of the argument which leads to summing the
excess fractions over all ages.

RESPONSE:  We have changed "SHOULD" to "MIGHT" as the referee suggests
and explicitly mention the discrepency in timescales to which the
referee alludes (see response above).  We have also included an 
interpretation that would be robust if the dust production phase
lasts up to a factor of x10 in age:

"Averaging the results over factors of ten in age results in excess
fractions of 18 \% (3--30 Myr) and 12 \% (30--300 Myr) implying that
at least 30 \% of sun--like stars exhibit evidence for terrestrial
planet formation, provided that the epoch of 24 $\mu$m excess
emission lasts $< \times 10$ in age."

REFEREE:  The same paragraph also equates the "era of maximum dust production" 
with the transition from orderly to runaway growth. However, in the Kenyon & 
Bromley models, this era is associated with the growth of planetesimals to 
2000km in size, which is not the same thing.

RESPONSE:  We thank the referee for pointing out the confusion in our
description of the Kenyon and Bromley results.  Following Kenyon and Bromley 
(2004, section 3, paragraph 6; see also Kenyon and Bromley 2006, 
section 3.5 paragraph 2) We now state in paragraph 2 of section 4 that:

"Sometime after gas disk dissipation, orderly growth proceeds to runaway 
growth, transitioning to chaotic growth.  Maximum dust production is 
thought to occur between the latter two phases when the largest planetesimals
reach $\sim$ 2000 km at a given radius \citep{Kenyon04,Kenyon06}."