This poor onion was suffering from mold around the base. I spilled too much sugar into the soil, and kept it too wet in the dark.

I removed the feeding apparatus, and gave it a few days in the light, so the mold would die. I'm noticing a general problem here, that fungus is everywhere, and unlike plants, have actually evolved to gobble up external sugar. I will have to learn more about how plants protect themselves from having fungus grow into them.

I'm trying a new method, of simply dipping the cut tip in sugar water (there's a chip out of the sugar-water cup, so the plant is not being squished):

This time, I'm making no effort to isolate it from light. I won't know for sure from this kind of experiment whether it's actually working, but I will be able to see if the sugar-water is causing obvious harm. Not killing the plant seems like an important thing to learn.

This is fairly thin sugar-water. Syrup might work better. I know you can store syrup at room temperature for a long time without having fungus grow in it, while sugar water is eventually going to ferment. I'll have to give that a try, too.

I've come across an interesting page on DIY microfluidics, which I think is the right kind of thinking, if not specifically applicable methods, for making a grafting interface:

http://science-practice.com/blog/2015/01/29/low-tech-microfluidics/