I first met Persi in the Berkeley Statistics coffee room in 1979.
I think we had previously corresponded (handwritten letters in envelopes with stamps attached)
briefly, and we had two interests in common, exchangeable arrays and card-shuffling, so 
we hit it off and started a regular interaction.
<p>
I'm always surprised to remember that we have only <xxx> 5 joint papers -- it feels like more.
In all of these, he was the prime mover.
I confess that a typical start would be for him to mention some simple-looking problem and for 
me to interrupt with "that's just a calculation" or "the answer must be ----- because -----".
It's a miracle that he put up with me at all, so he's a saint!
But the point is that he has the marvelous gift of seeing below the surface -- that a 
superficially simple problem will lead to deep and beautiful mathematics.
<p>
Continuing with the sainthood theme ....  at the Friday dinner, in addition to assigning us a 
homework involving the Mexican hat dance, he promised everyone an hour of his time.
And he really will -- but I encourage you to come prepared with questions.
<p>
When I'm in sarcastic mode -- permissable by virtue of my ethnic heritage as an Englishman -- I tend to say stuff like the 
following. Suppose there are 40 people working on some math topic.
Someone has to be "the best", and one could devise an algorithm -- citations and awards etc --
which would output some name as "the best".  But if you <i>need</i> an algorithm then the 
winner is <i>ipso facto</i> not outstanding but merely <i>primus inter pares</i>, to continue
with high school Latin.  
The pinnacle of achievement is to be recognised as <i>sui generis</i> -- truly unique -- 
and we can all agree that Persi is the epitome of <i>sui generis</i>. 
We are all priviliged to have been associated with him.