Here is a shot of the pollination of the tiny Cephalotus flowers. It is done on an almost daily basis, until the ovules begin to visibly swell and/or the flowers close.
This is what I did my 1st year. Every day I was out there with my brush pretending to be a bee. The end result: some flowers were pollinated & some weren't (& some had a few seeds develop - not a full set). Some of the time my brush just got gunked up with flower nectar. The next several years I allowed the native bees to do all the work and I got just as many or more seeds.
This year I 'tried' to pay attention and notice what the flowers were doing. When the flowers open & the pollen 1st becomes visible, the stamen is not open & receptive (it's hardly visible). My 1st two stalks were almost 100% synchronized so that 1st wave of pollen had no receptive stigma.
Later when the stigma's opened & became receptive, there were no flowers with nicely-colored fresh pollen.
Over here on Matt's page (about 2/3 of the way down), there are two flower pics. The pic on the left clearly shows ripe pollen (click on pic to see larger view). The pic on the right shows a receptive stigma as well as the now dull pollen (and several developing seeds from older flowers - as the pollinated seeds develop, they stick out the end of the capsule). I don't know if this older, brown/gray pollen is still effective.
Although not visible in any of these pics, when the stigma on my flowers 1st 'matured', the pool of nectar at the base of the stigma & stamens almost overflowed. When I tried to use a paintbrush to pollinate, it regularly got covered in sticky nectar.
However, if I didn't get the brush below the surface, I didn't get any pollen since the pollen is on the lower inside of each stamen head. This year, when I wanted to help the bees, I would pluck a stamen that had colorful pollen with fine tweezers and then dab that on flowers with receptive stigmas. It seemed to work very well - as long as I had flowers in different stages of development...