According to Richard Ashton's book, pomegranates can be propagated by grafting, but they usually aren't since they sucker so badly. Unless the graft union is well marked, it can be very difficult to tell what growth is coming from the scion, and what is coming from the rootstock. I figured I would try to capitalize on the healthy root system that's already in place, and give it a try anyway.
I was lucky enough to get hold of some Kandahar Early and Desertyni budwood for this experiment. Richard Ashton describes Kandahar Early as "a very old variety from Afghanistan that has red skin and aril color", and Desertyni as a choice pomegranate cultivar from the Turkmenistan collection of Dr. Gregory Levin, with a unique orange flavor. I cut down the Garnet Sash to approximately two feet, and used the same cleft graft method used to graft apples and pears. I used plumber's putty to fill the gaps, wrapped it tight with a rubber band, and then wrapped the whole thing with parafilm. I wound up grafting two sticks of Kandahar Early, two of Desertyni, and leaving a small branch of Garnet Sash just in case the grafts fail and I want to try again. I'll have to keep vigilantly pruning the buds that try to grow below the grafts in order to focus the tree's energy into the budsticks, and keep the suckers cut down. If all the grafts do fail I'll probably just dig up the tree and replace it, but I really hope it works. I don't know of anyone else with a multi-variety pomegranate tree. If you have one (or know of one), send me an email or a picture. I'd love to see it.