It has taken us some time to accept that Edmundoa is a special sort of the old Canistrum. I try to think of them as having hairy flowers! I had always thought that Edmundoa ambigua had red primary bracts even though this is not mentioned in the description but recent photos on Florapix has shown these can be green. This started me pondering why we had Edmundoa lindenii var rosea.
In the 1980’s we had -
Key to Varieties and Forms of Canistrum lindenii

1.Primary and outer bracts yellowish white to nearly white, sometimes faintly green at apex; inflorescence 100-500-flowered. -> var lindenii
- 2. Inflorescence sunk in the center of the rosette or raised only slightly => var lindenii forma lindenii
- 2. Inflorescence raised 20 cm or more above the center of the rosette. => var lindenii forma elatum
1. Primary and outer bracts colored green or rose; inflorescence 50-90-flowered.
- 3. Primary and outer bracts green. => var viride
- - 4. Inflorescence raised 20 cm or more above the center of the rosette. => var viride forma magnum
- - 4. Inflorescence sunk in the center of the rosette or raised only slightly. => var viride forma parvum
- 3. Primary and outer bracts rose to bright red. => var roseum
- - 5. Inflorescence raised 15 cm or more above the center of the rosette. => var roseum forma procerum
- - 5. Inflorescence sunk in the center of the rosette or raised only slightly. => var roseum forma humile

Then in 1997 we had Edmundoa lindenii (Regel) Leme, var. lindenii. "Canistrum – Brom Atl. Forest" 46-51, 1997 where all the varieties disappeared except for var. rosea.
This is what Leme had to say at the time.
"After examining numerous E. lindenii plants in the wild, Reitz (1950, 1952) arranged this material in varieties and forms, separating the type variety with its yellowish, whitish or greenish tipped bracts from the variety viride with its entirely green bracts. He subdivided these varieties into forms based on the length of the floral scape (inflorescence sunken or raised). He used the same criterion to establish forms for the variety rosea, and also mentioned the smaller number of flowers in this variety when compared to the type variety.
Reitz's criteria were discarded here because, though very logical from an horticultural point of view, they are decidedly artificial. The color gradation of the involucral and primary bracts, from yellowish to whitish to green, falls within a very narrow range of chromatic variation, so much so that some specimens even have a combination of these colors (yellowish or whitish with a greenish apex). Furthermore, in the post flowering stage, the yellowish-whitish bracts may become greenish toward the apex (pers. obs.). The continued use of this criterion would encourage the establishment of numerous, biologically inconsistent varieties. For this reason, the variety viride with its entirely green bracts was considered to be a mere color variation of the type variety, and was therefore placed in synonymy.
The creation of forms based on scape length was also seen as artificial. I observed that specimens with a well-developed scape, that raised the inflorescence well above the rosette, became more compact and produced much shorter scapes when grown in cultivation, under a uniform, more intense light regime. The inflorescence was no longer perched above the rosette in these plants. Obviously, the variability that so strongly influenced Reitz is seen in the wild. But given the overall variation pattern of the species, this criterion becomes inconsistent and artificial, and segregates plants nomenclaturally that are practically identical. The taxonomic forms based on this criterion are placed in synonymy. The number of flowers also varies according to the stoutness of the plant and is discarded here."

Despite his reasoning Elton Leme still accepted var. rosea for its red primary bracts.
We now see Edmundoa lindenii var. rosea (E.Morren) Leme: Considered a synonym of the type variety - Reflora (cont.upd.) Lista de Espécies da Flora do Brasil. Jardim Botânico do Rio de Janeiro. http://floradobrasil.jbrj.gov.br/ (Retrieved 28.3. 2018).

In 1997 Leme also reported, “In Rio Grande do Sul, var. lindenii and var. rosea are sympatric and may be found in the same area (J. C. da Silva, pers. Comm). The reason I am saying this is that Peter Tristram of New South Wales, Australia received seed called Edmundoa lindenii from Rio de Janeiro Bot. Garden which had red primary bracts on flowering. Has this instability in colour of primary bracts been noted by other seed raisers? Or has the Rio de Janeiro Bot. Gardens dropped the use of ‘var. rosea’?

What has happened to all those varieties/forms mentioned by Reitz. Are they still being grown? What names are on the labels? There seems to be no record in the Bromeliad Cultivar Register other than the variegated E. ‘Alvim Seidel’ and ‘Brazil’.
If var. rosea is treated with Edmundoa lindenii what will growers call the one with the red primary bracts. The ICNCP rules frowns on the use of colour as a single word and we could go back to the Lectotype where Comte de Germiny is involved and call it Edmundoa ‘Germiny’ but somehow I cannot see this being noted by horticulturists. I can see Edmundoa ‘Rosea’ being accepted and acted upon, and that will be my course of action.

We have Flora do Brasil 2020 ignoring the existence of sub-species of Edmundoa lindenii and The World Checklist of selected Plant Families by Kew Gardens preferring the genus name Canistrum to Edmundoa which makes you wonder where we go next. While the botanists dither, at least having ‘Rosea’ in the BCR will give you a reference point.