I guess that if you had a look at the file everything would make much more sense, but anyway, here's the flow I ended up with:
it's actually a very basic flow, nothing fancy going on there, here's how it works.
the first event it's pretty straightforward, I birth particles, give them a position on the pf source's icon (quite small in this case), speed to make them move, and there are a couple of obvious forces: gravity and drag to make the particles behave as if they actually had weight (gravity pull them down and drag slow them down increasingly over time).
then comes a speed test, it checks when particles slow down under a given speed (in this case half their original speed), and when they test true it send them out to the next event.
in the second event there's a split amount test which immediately sends out a given percentage of particles to the third event (which is the one that gives the particles that radially sprayed look to them). the remaining particles in this event will keep going up their way until their killed by the delete operator (which is set by particles age, which sets a life span to them with random variation). the scale operator randomly scales down particles to almost half their original size.
in the third event there's a find target test which basically tells particles to look and go for a specific target. in this case the target is a geometric invisible object positioned just where the source is. having a circular shape all particles will start falling down in a radial fashion, otherwise the would just fall back pretty much linearly (which would look correct as well, I guess with actual water it would depend on the nozzle shape).
with this kind of target they basically follow this path (a bit exaggerated, but you get what I mean):
once again scale scales particles down randomly, and delete prevent them from going on forever and kills them after a short time they enter this event.
the last event doesn't really matter, it's there just to instantly kill any particle which actually reach the target. but this won't happen anyway, because they're all deleted much earlier in event 3.
I guess that's all, playing with settings for any operator and test will of course give different results from the one I got. in the rendering I posted I enabled mblur at pf source level, this is very important to make them particles look a bit less as single particles and a wee bit more like flowing water.
![[img=http://img514.imageshack.us/img514/438/scanrb0.th.png]](http://img514.imageshack.us/my.php?image=scanrb0.png){kind=link}