Dear Editor, 

We submitted our manuscript titled, "Efficient quasi-kinematic large-scale dynamo as the small-scale
dynamo saturates" for publication in Physical Review Letters. 

As the first round of referee reports were in disfavor of the manuscript for publication, we responded
to the concerns pointed out by the referees and resubmitted a revised manuscript for a second round of 
refereeing. Well, we again received a negative response. 

We have decided to submit our case for a formal appeal. We had two referees evaluating our manuscript. 
The following are our concerns with both of the referee's evaluations of our manuscript. 

In the case of referee A, we believe he/she appreciates the significance of the result but seems to be  
unconvinced that such a result is available in our data. The main issue for the referee is related to Fig.1
and Fig.2 in our manuscript (which show the main results). These figures show evolution of the magnetic spectral modes 
M_k with time in the top panel and specifically, the growth rate of M_1 (i.e. the k=1 mode shown in black in
the top panel) in the bottom panel. Our claim is that M_1 has three stages of evolution where the first
two stages are that of exponential growth. This can be seen clearly in the log-linear plot as two straight lines 
with different slopes one after the other in the time ranges of t=0-100 and t=100-270. The third stage 
is after t=300. The second stage is the novel result which the rest of our paper focuses on. 
It shows the operation of a large-scale dynamo after the small-scale dynamo has saturated.

The referee expects well-defined plateaus in the growth rate curves,
during the periods of exponential growth. However since we are working with 
turbulent systems, the resulting plateaus from taking the derivative of a turbulent quantity turns
out to be noisy. And thus the referee is unhappy that we do not get very neat looking plateaus. 
In order to clarify to the referee, we contrasted our result on the second stage
with a case where the second stage is not expected
and we provided this comparison in the inset in Figs.1 and 2. We have shown how in a system where the second stage is 
not expected the growth curve dips and oscillates around zero indicating the absence of a second stage. 
Well, the referee remained unconvinced with our explanatory revisions. However, we think our figures do 
amply demonstrate that there are three stages of growth of M_1 
including the second stage which we have subsequently shown to be a quasi-kinematic large-scale dynamo. 

After receiving the second report, we have now performed additional expensive runs with a varying parameter to show 
the robustness of the second stage. We have now added some brief new text related to this in our manuscript. 
This is highlighted in red colour. We briefly mention the results and refer the reader to the supplement for the details. 
(These details are given in a new subsection and an additional figure 4 of the supplemental material.) 
Note that the text in blue are the revisions that we had previously made to our manuscript 
in response to our referee's concerns.  

In the case of referee B, he/she did not appreciate the significance of the result in their first
report. In our response, we explained the main questions and issues that we were addressing. To
clarify this in the manuscript, we revised large portions of our introduction. However, the referee responded to our
revisions by merely saying that there is no convincing case for PRL, without any explanation. 

We believe our manuscript contains high-quality numerical results and the main result is an
important one that goes a long way in understanding the origin of large-scale magnetic fields in
turbulent astrophysical systems.  With these concerns, we find there is sufficient cause for an extra degree
of consideration by PRL and hence we are submitting this formal appeal. 

Yours sincerely, 
Pallavi Bhat 
Kandaswamy Subramanian 
Axel Brandenburg