diff --git a/browse/templates/stats/monthly_downloads.html b/browse/templates/stats/monthly_downloads.html
index 1c8530a98..80bed358c 100644
--- a/browse/templates/stats/monthly_downloads.html
+++ b/browse/templates/stats/monthly_downloads.html
@@ -20,8 +20,8 @@
     <p><strong>Caveats:</strong> While we have attempted to extract download data representing unique full-text downloads by real users, there are many factors which affect accuracy.
     These factors include:</p>
     <ol>
-      <li>the data is from the main arXiv site and the arXiv mirrors, though some mirror data is incomplete.</li>
-      <li>only web downloads are included (and not FTP or email &quot;downloads&quot; that were formerly supported).</li>
+      <li>the data is from the main arXiv site.</li>
+      <li>only web downloads are included.</li>
       <li>we have counted downloads according to the COUNTER algorithm which excludes rapid repeat downloads.</li>
       <li>we have attempted to identify and remove robot or automated downloads from the count (false positives lead to undercounting, failing to identify robots leads to overcounting).</li>
       <lidata prior to 2009 has not been cleaned with as much care as later data, but it shows trends nonetheless.</li>