If you try to identify cultivars by their seedlings, you will probably be unsuccessful because the seedlings will most likely not resemble the parent plants. Preserving the traits or genetic makeup of canna cultivars is the same as with cultivars of other plants. They must be propagated asexually. If the parent plants are species cannas, that won't be such a problem as long as they were not cross polinated with another species or cultivar.

In any case, it is extremely easy to propagate cannas asexually because their rhizomes can simply be dug up and divided, If it's not too late to get them, dig up some rhizomes or divisions of the cannas you wish to save and replant them. They'll over-winter in the ground just fine in Houston and will come up as good as new their new location.

As far as identifying a canna goes, I have found that researching on the internet is the best way to go. A book devoted to or going into any depth on cannas is very hard to find.

If the seeds came from a hybrid, it will not come true. In other words, it may resemble the parent, but wouldn't BE the parent. Then it again, it may be something completely different.

Not all cannas make seeds. The seeds are big black round things. Some people collect sterile seed pods and think the little tiny balls of stuff inside are seeds. But it's not. The seeds are big and hard and need to be nicked before sowing.

Many cannas look the same--same leaves, same blooms. So once a name has been lost, it's hard to "re-name" them, unless they have distinct characteristics.

If you want to preserve that canna variety, it would be best to dig the rhizomes. That would be the only sure way of continuing the historical canna, as the seeds could very well be slightly different or significantly different from the parent.

BC