One of the complaints in the gardening press in the late 19th century was that many of these new cannas were very similar and it was difficult to distinguish between some of them. If it was difficult for gardeners involved in all the excitement of these new releases, it has been even more difficult for researchers trying to unravel them one hundred years later. Slowly scraps of information have been pieced together. Old illustrations have been found, old gardening press articles and many catalogue descriptions sifted through for any crumb that might shed light on which Italian-type was which. The picture is slowly emerging as more and more of these old cultivars turn up and/or we notice that some of them are actually different from each other.
At least one can be confident that any of these big, blowsy, “floppies” with thinly textured staminodes that are constantly turning up in both hemispheres, are not chance seedlings or more recent creations, although Bernard Yorke of Australia has produced F2 and F3 generations in the past year or so, many were not sufficiently different from their pollen parent to be released and those that are being released exhibit few of the original Italian-type ancestor’s characteristics. The window for the orchid-flowered canna’s creation was brief and well documented. It can be placed between 1895 and 1905 in Italy and the USA. We also know that only two hybridizers worked with the difficult C. flaccida. Here in Australia we have quite a few of these old beauties still thriving. Most have long lost their names or acquired new ones.
The yellow coloured Italian-types have proven particularly difficult to differentiate, but at least three of them have been identified. These are ‘Austria’, ‘Asia’ and the USA bred ‘Burbank’. Their patterning of spots never varies, unlike the chameleon-like yellow and red cannas that boast loud splashes and splodges of red patterning on a yellow ground such as ‘Bavaria’, Britannia’, ‘Allemania’ etc., and even ‘Italia’ which is the most stable in flower patterning of that colour type yet still displays occasional variations. 
To be found growing absolutely everywhere in Australia is Canna X ‘Austria’. (Sprenger, Italy). (Erroneous synonyms: R. Wallace, Richard Wallace, Lemon Gem etc.) This Canna has tall, bright, apple-green, broadly lanceolate, upward pointing foliage. 
The acid yellow flowers open their faces upwards and are crowded on their stems. The throat, modified anther and base half of the lip is lightly lineally spotted with pale red which instantly fades to a pale brownish orange on exposure to sunlight. Each flower only lasts a day or two and is bleached white by strong sunlight. The lip is sometimes split and often entire. You usually have to grasp the stem and tip it forward to actually view the centre patterning. A real star gazer. For this reason the central spotting is often entirely ignored in some old nursery catalogue descriptions.
The plant is a profuse increaser with very tight tillering and quickly forms tall, massive, crowded clumps. Very easy to identify even when not in bloom by the distinctive foliage and habit.
Not so easy to identify is the very similar Canna X ‘Asia’. (Sprenger, Italy). The spotting in the centre is almost identical to ‘Austria’, being closer to orange than red, but the blooms are better distributed on the head and some face outwards. The shape is distinctly like a Cattleya orchid with the lip cocked up a little at the outer end where ‘Austria’ is flatter and reflexes slightly. ‘Asia’ has a split lip (seen in the photo at right), as do many of these Italian-type cannas but may occasionally be found with an entire lip free of the end notch.
The greatest difference is in the foliage. ‘Asia’ lacks the bright apple-green leaf colour, being what might be termed light green. The leaves are broader, held at about 45 degrees and their outer half is lax and droops a little. It does not increase as rapidly as ‘Austria’ yet increase is satisfactory.
Tillering is longer so the clumps do not appear as crowded. Overall, the appearance of the clump is more pleasing than ‘Austria’ due to the better spacing between the stems and at least the lower flowers in the head of blooms try (not very successfully) to look at the viewer. Flowers do not bleach out in sunlight as badly as ‘Austria’. ‘Asia’ is quite rare here probably due to it being not quite as drought hardy. The broader, openly held foliage would transpire moisture more quickly than the narrower, upright foliage of ‘Austria’.
I found my clone growing in an old municipal planting in shallow water retained on the top of that ghastly solid black plastic weed suppressor. It enjoyed permanent water provided by the sprinkler system, deep shade from nearby deciduous trees, and grew at the base of an old water tower. It’s only food source being the decomposing spent leaves from the previous season and any decomposing mosquito lavae that didn’t make it to adulthood. It had colonised the entire plastic ground cover making it a cinch to liberate a small piece. It is now thriving in good soil but I suspect it might be even happier back in its old swamp.
Head and shoulders ahead of the previous two as a garden star is Canna X ‘Burbank‘. (Burbank, USA). Strangely, Luther Burbank in America was experimenting with the same Canna cross breeding at the same time as Sprenger was flicking C. flaccida and ‘Madame Crozy’ pollen around in Italy. I suspect, from Burbank’s notes that he may have taken his experiments a generation or two further than Sprenger before he was satisfied with the outcome. He mentions that it was almost impossible to breed F2 and subsequent generations and so the process of improvement was very slow.
Canna X ‘Burbank’ is extremely beautiful. The yellow colouration of the blooms is indistinguishable from Sprenger’s two, but the red spotting in the centre is darker and heavier and there is a hint of red blushed on the stigma. It produces a lot of pollen and is well loved by the bees. Flower shape is more even and rounded than either ‘Austria’ or ‘Asia’, due to broader staminodes . The flowers are pleasingly arranged, gazing outwards all around the head and free from crowding. 
They last longer than those of ‘Austria’ and fade to white before browning off or shrivelling of the edges of the blooms. One more often sees multi-coloured heads of yellow and white. The red spotting shows little evidence of fading, remaining distinct and bright for the life of the bloom. Another tall grower, yet one does not need to grasp the stem and bend it downwards to see the true beauty of the blooms and the highly visible red patterning is invariably mentioned in the old catalogues.
The foliage is mid-green, very broadly spoon-shaped and large, One could almost describe it as massive for this type. All but the terminal leaves on a stem arch pleasingly outwards and have a lax tip. Tillering is good with well spaced stems and the whole quickly bulks up to make a large clump without galloping off into the neighbour’s backyard. It is not as drought tolerant as ‘Austria’ and is harder to find. The foliage badly pictured on the left was taken in furnace-like 42.8 degrees C. (109 F). Humidity was 30%. It shows remarkable hardiness for a “conservatory” plant. Perhaps Burbank employed a criteria for thriving both indoors and out in a hot climate (California). Rhizomes are stouter than the previous two yellow cannas. All have white rhizomes.
References:
Burbank, Luther. The Canna and the Calla: and some interesting work with striking results. Paperback ISBN 978-1414702001
Burbank, Luther. The Training of the Human Plant. Century Magazine, May 1907.
Links:
http://en.wikipedia.org/wiki/Luther_Burbank
Where I started to laugh was when the article asserted that Messrs J.C. Vaughan, the distinguished Chicago seeds men then renamed C. “Tarrytown” to C. “Florence Vaughan”, and at the same time the article transformed the cultivar into being “lemon yellow with creamy-orange splotches and dots”. All the contemporary reports and Vaughan’s own adverts from the 1890’s show C. “Florence Vaughan” to be a Crozy Group flower, yellow with red spots, a fully attested specimen is still grown in Australia, see below right. This total made-up version of history is quite remarkable and amusing, but there is the danger that the uninformed will take it seriously.
Canna News 