deletions | additions       diff --git a/MAYBE_REMOVE_NEXT_PARAGRAPH_AND__.tex b/MAYBE_REMOVE_NEXT_PARAGRAPH_AND__.tex  index 338b163..0c836b4 100644  --- a/MAYBE_REMOVE_NEXT_PARAGRAPH_AND__.tex  +++ b/MAYBE_REMOVE_NEXT_PARAGRAPH_AND__.tex 
...
%[MAYBE REMOVE NEXT PARAGRAPH AND PUT INFO IN FIG2 CAPTION OR MAYBE NOT GIVE REASON]  %We restrict our analysis to stars with \numax\ larger than 50\muhz\   %and masses below 2.1\msun, to ensure it includes only red giants that do  %not burn helium in their cores [ref Stello2013Mosser2014Montalban201?].   %A cross matching of our sample with samples of known classification [REF  %Stello2013,Mosser2014] we found only XX helium-core burning stars in our  %sample (all with \numax below 70\muhz), which were removed.  %Because the proposed theory for the suppression has only been developed for  %hydrogen-shell burning stars, we remove helium-core burning stars from our  %sample. We ensure that by looking only at stars with \numax\ larger than 50\muhz\  %and masses below 2.1\msol[ref Stello2013Mosser2014Montalban201?].   %The proposed theory for the suppression has only been developed for  %hydrogen-shell burning stars, and we restrict our analysis to those by  %looking only at stars with \numax\ between 50 and 240\muhz\ and masses  %below 2.1\msol[ref Stello2013Mosser2014Montalban201?] In Fig. 2 we show the%power ratio between dipole and radial modes (the  dipole mode visibility for about 3600 3,600  red giants observed over the first 37 months of the {\it Kepler} \kepler\  mission. %space saver: Put next bit in Methods  Our analysis is restricted to a sample of stars with \numax\ larger than 50\muhz\ and masses below 2.1\msol, which in absence of 2.1\msol\ which, assuming no  observational uncertainties uncertainties,  is expected to include only red giants that have not started burning helium in their cores \citep{Stello_2013}. %MC: Changed %space saver: Put  next sentence.   %DS: You cant make direct refs in Nature (references cant be part of the text).  %We cross matched our sample with the data of \citet{Stello_2013} and \citet{Mosser_2014}, which allowed us to identify and remove a few evolved stars burning helium bit  in their cores ($2\%$ of our sample, almost all with \numax < 70 \muhz). Methods  We cross matched cross-matched  our sample with those of known helium burning helium-burning  stars \citep{Stello_2013,Mosser_2014}, which allowed us to identify and remove a small fraction of evolved stars burning helium that that,  due to measurement uncertainty uncertainty,  had entered our sample ($2\%$ of our sample, almost all with \numax\ $< 70$\muhz).%to %remove helium burning stars that due to measurement uncertainties in \numax\ and mass had remained in our %sample. The two percent of helium burning stars that we found, and of which almost all %(70 out of 80)  %had \numax\ below 70\muhz, were subsequently removed from our sample.   %The missing dipole modes in a significant fraction of stars revented us using   %the dipole mode period spacings to select stars in this particular evolution stage  %\citep{Bedding_2011}.   %IF A REFEREEE ASKS IF WE CHECKED OUR SELECTION WORKED  %A comparison with lists of helium core burning stars identified through measured period   %spacings of non-suppressed stars \citep{Stello_2013,Mosser_2014} revealed only X stars in common with our   %sample. They all had \numax\$ < 70\,$\muhz.   %It is striking how   The stars form two distinct branches that gradually merge as the stars  evolve towards lower \numax, with % This trend is also evident in Fig.1.   most stars falling on the ``normal'' upper branch of $V^2\approx 1.5$, in  agreement with previous results \citep{Mosser_2011}.   The lower branch with suppressed dipole modes agrees remarkably well with theoretical predictions (black curve, Fuller et al. 2015). This reduction assumes that all the wave energy leaking into the stellar core is trapped.   The suppression occurs because wave energy leaking into the core is trapped due to a magnetic greenhouse effect caused by strong internal magnetic fields (Fuller et al. 2015).  %According to theoretical work (Fuller et al. 2015), this occurs in stars %with strong internal magnetic fields.   For this branch, the decrease of the suppression towards lower \numax\ is a consequence of the increasingly weaker coupling between acoustic waves in the envelope and gravity waves in the core as stars evolve (Fuller et al. 2015).   With this large stellar sample we can separate the stars into five different mass intervals, represented in Fig.2 from 0.9 to 2.1\msol. It is striking how %, which clearly shows that   the relative  population on the lower branch (stars with suppressed dipole modes) is strongly mass dependent.