This was in one of Edward Laskers Chess Books.

 

White to move and mate in 3. Much harder than it seems.

 

P.S - check out the Taverner 1881 problem i posted yesterday - it's even more impressive (im amazed it hasnt had more comments)

yes thats beatiful

Not too hard. Solved it on first try.

i'm a little more sadistic. i like na2 more. it's a little longer, but black gets to see his death coming

Once again we get a puzzle that depending on your point of view is either elegant, or unrealistic.  The notion of under promotion in this position is not going to occur to most players, nor should it when Nd3+ leads to mate in four after promoting to a queen.  Black clearly has no counter play, nor is there any realistic chance of stalemate.  In a game, Nd3+ is THE obvious move.  But the solution is lovely and appeals to our aesthetic sense.  Perhaps this is the difference between game play and puzzles?

What a spectacular puzzle.  It's true that puzzles that include pawns on the threshhold of advancement frequently hinge on promotion to a knight rather than a queen, but that doesn't detract from the cleverness of this puzzle.

I think there is, or at least should be, a clear distinction between chess problems and chess gameplay. It seems that we see very few problems for problems sake here, let alone puzzle-specific situations such as helpmates/reflexmates/selfmates.

 Although I agree that some problems are useful if not enjoyable when they mirror real-game situations, I don't see anything wrong with a problem purely for aesthetic/logical satisfaction

I would definitely promote to Queen

 

Very nice