diff --git a/fastlib/u/rriegel/nips07/gnp_nips07.tex b/fastlib/u/rriegel/nips07/gnp_nips07.tex
index 6d990525a2..598111f7cd 100644
--- a/fastlib/u/rriegel/nips07/gnp_nips07.tex
+++ b/fastlib/u/rriegel/nips07/gnp_nips07.tex
@@ -919,7 +919,9 @@ involves trees formed on the inputs sets.
   trees to obtain $\kdleft{X} \cup \kdright{X} = X$ and $\kdleft{Y}
   \cup \kdright{Y} = Y$.
 \end{definition}
-\noindent It is clear that partitioning data in this manner will be fast, but 
+\noindent It is clear that partitioning data in this manner will be
+fast, and for problems involving points in a metric space, spatially
+informed trees formed on the input sets tend to yield useful bounds.
 
 Trees built for the inputs sets may also serve other useful purposes,
 such as facilitating rapid, bottom-up precomputation of statistics and