Botanical Code question

[Mark Egger]

Hello,

In my research on types within the genus Castilleja, I've come across
a puzzling situation that I hope someone out there with more
experience in the fine points of the I.C.B.N. may be able to clarify
for me.

The question involves the typification of Castilleja oregonensis
Gand., traditionally treated as a synonym of C. rhexifolia Rydb. The
protologue reference is Bull. Soc. Bot. France 66:119, 1919. The mode
Gandoger employes to describe this taxon is very odd. It starts with
the species name, followed immediately by a Latin diagnosis,
distinguishing it ("Differet a"...) from C. oreopola Greenm.,
presently treated as C. parviflora Bong. var. oreopola (Greenm.)
Ownbey. The text of the diagnosis describes well the characters one
might now use to distinguish C. rhexifolia from C. parviflora var.
oreopola.

Next comes the "Hab." description, which consists of a somewhat
mangled Latinization of the collection label for a Cusick specimen
from the Wallowa Mts. of NE Oregon, which is cited in parentheses at
the end of the paragraph (Cusick 2443). This fairly well distributed
collection (F, LY-GAN, NY, OSC-ORE, WS) appears in catalogs of
several herbaria as the type collection for C. oregonensis, and it is
listed as such, in synonymy under C. rhexifolia, in the Vascular
Plants of the Pacific Northwest and in the Intermountain Flora.

But now comes the complication. The final paragraph of the protologue
reads in its entirety, "Typus (Suksdorf n. 6047! e monte Padds [=Mt.
Adams], Washington) minor est, foliis trifidis floribusque majoribus
condensatis."  There are THREE peculiar aspects to this, as I read it.

1. Gandoger seems to be distinguishing the "Typus" for this name from
the Latin diagnosis he just wrote in the first paragraph of the
protologue.

2. He uses a collection from Washington state to apparently typify
the name "C. oregonensis", immediately after citing an apparent
paratype that was collected in Oregon (of course, this fact, alone,
is of no nomenclatural significance...it's just odd!).

3. The Suksdorf collection (LY-GAN, WS, WTU and probably other
herbaria), apparently cited as the type, is of typical C. parviflora
var. oreopola, the same species from which the first part of the
Latin diagnosis attempts to separate the new taxon! In effect,
Gandoger first distinguishes his new species from C. parviflora var.
oreopola with a good description of and specimen citation for C.
rhexifolia, then immediately indicates as type a collection of C.
parviflora var. oreopola!

After puzzling over this for some time, I wonder if the word "Typus"
starting the final paragraph of the protologue, was SUPPOSED to
follow the citation of the Cusick collection, which ends the
immediately preceding paragraph? Of course, this is totally
conjectural, but it would  explain several odd features of the
protologue, especially by accounting for the strange manner in which
the Suksdorf collection is distinguished from what came before (in
other words, listing it as an atypical paratype of the new species.

So, my question has two parts:

1. From the wording of the protologue, must one interpret it
literally and treat the Suksdorf collection as the type? Or can one
use the fact that the first and main part of the diagnosis much
better describes the Cusick collection to justify a
lectotypification(?), using that specimen? The St. Louis Code (9.17)
says that a lectotype or neotype choice can be "superseded...if one
can show that...it is in serious conflict with the protologue....",
but this option does not appear to be available in the case of an
original holotype citation, at least I could not find such a
statement for holotypes.

2. The answer for the first question should basically answer the
second: should C. oregonensis be treated as a synonym of C.
rhexifolia (based on the Cusick collection and listed in several
regional floras as the type), or must it instead become a synonym of
C. parviflora var. oreopola (the Suksdorf collection)?

Whew, a lot of writing to explicate what is, in the end, a fairly
trivial matter, but I love this stuff,  and I would greatly
appreciate any clarifying comments from the nomenclatural gurus of
the TAXACOM community!

Thanks in advance,

Mark Egger
Research Associate, WTU