Hi Koji:
      I've changed my mind about one part of the reply 
as noted by ???? below. 
I think we should not discuss responding to Berger et al.
I started that part with ``Frankly'', but actually I meant
``Unfrankly.'' We do know what Berger et al. meant:  i.e.,
the jet model is completely wrong by orders of magnitude
somehow.   The real reasons I don't want to respond to
Berger et al. are (i) if we are still trying to get into
ApJ letters, then we don't want to add more discussion;
(ii) I don't want to get into a discussion about things
on which I'm not an specialist:  in a future work all
the issues should be thrashed out, of course.

     Of course, if a new referee asks for such a response,
then we'll have to do it.

                                Regards
				David
---------------------------

Major Concern:

         The referee believes that the jet model is ruled out
already.   However, the referee has not given a demonstration
of or a reference for this conclusion.  He/she in his/her first
report have merely referenced an observational report in
IAUC circular 7817 by Berger et al. whose most relevant remark is
that ``SN 2002ap is thus > 3 orders of magnitude less luminous in
the radio band than SN 1998bw."  We don't think that this report
without theoretical interpretation by itself rules out a jet.

         Shortly after our letter was submitted and posted at
astro-ph, a paper by Berger et al. was posted at astro-ph
(Berger et al., 2002, astro-ph/0206183) reporting their radio
results and modeling of SN 2002ap.
They address our jet model briefly in their conclusion and
implicitly rule it out on the grounds that ``Such a jet,
regardless of geometry, would have produced copious radio
emission.'' In the paper, Berger et al. do not give either
a demonstration as to why the model should be ruled out nor
a reference for such a demonstration.
We need to point out that Berger et al.'s remarks regarding
the jet model were based on the suggested jet parameters in
the first version of our letter.
Although we explicitly said those parameters were not
reliable values, Berger et al. seem to have read them as
such:  it is hard to say what Berger et al. understood
since they are vague and brief.
We have since eliminated most of the suggested parameters
and Berger et al. would now say about our jet model
we know no longer.

        Two further remarks are needed about Berger et al.
First, all of their detailed modeling assumes spherical
symmetry for the supernova ejecta and circumstellar environment.
Thus, their assertion that geometry does not matter does
not follow self-evidently from their analysis.
Second, we believe that geometry does matter.  A simple
argument shows this. Say the progenitor and circumstellar
environment had a special axis along which there was no
significant circumstellar matter and along which the
jets propagated. Then there would be no significant radio
emission from the jet. This argument is speculative, but
not wildly so. The one supernova (i.e., SN 1987A) for which
we have a detailed picture of the circumstellar environment
and the shape of the ejecta, shows that both have a
complicated structure and an axis of symmetry (e.g.,
Wang et al., 2002, astro-ph/0205337; Crotts & Heathcote
2000, ApJ, 528, 426). There is also a speckle interferometry
analysis (Nisenson & Papaliolios 1999, ApJ, 518, L29) that
suggests that relativistic jets emerged from SN 1987A
although apparently not along the axis of symmetry shown
by the other analyses. That speckle interferometry
analysis is subject to doubts, but we are not aware that
the analysis on its own grounds has ever been refuted.

         We conclude that radio observations and modeling
have not hitherto absolutely ruled out a jet in SN 2002ap
of any kind.
Given that our letter is not about radio observation and
modeling, we do not believe we need to discuss the radio
studies and what they may or may not rule out under different
assumptions. 
(?????? Thinking it over, I don't think we should even mention
Berger et al. again.  If a new referee wants us to discuss them,
then we will.????)


Minor Concern:

        The referee does not understand how a redshift
can give rise to a wavelength dependent polarization
from electron scattering.
We do not understand why he/she does not understand.
We have shown clearly both in our first response and
in our revised paper how a redshift will cause
wavelength polarization:  it is not a subtle effect.

        The referee is correct that there will be
some wavelength dependence due to the different epochs
of the direct spectrum from the bulk ejecta and
scattered spectrum from a jet.
However, this is a lesser effect.
Say the jet is moving at speed beta perpendicular to
the line of sight.
The time delay between the two spectra (not counting
relativistic effects) will be of order

      Delta t = (beta c t)/c = beta t   ,

where t is the time since explosion.
If beta=0.2 say (which of order the number
we suggest for the jet) and t is of order 10 days,
then Delta t=2 days.
Near maximum light, the spectrum and light curve are
changing relatively slowly over time scales of 2 days.
Our observations relevant to the jet model are from
near maximum light (i.e., of order 10 to 11 days after
explosion). Thus, we think that differences in the
spectral epoch will not greatly affect the polarization.
However, in detailed modeling the time delay should be
incorporated. But that is beyond the scope of our letter.


Conclusion:

         We think that the referee's judgment has failed
in regard to our letter and that the revised version we
submitted in response to the first referee's report is
acceptable.
We ask for a second referee to consider that version.