The Type Concept in Systematic Botany

Author(s): A. S. Hitchcock
Source: American Journal of Botany, Vol. 8, No. 5 (May, 1921), pp. 251-255
Published by: Botanical Society of America, Inc.
Stable URL: http://www.jstor.org/stable/2434993
Accessed: 26-06-2016 08:17 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact [email protected].

Botanical Society of America, Inc. is collaborating with JSTOR to digitize, preserve and extend access to
American Journal of Botany

This content downloaded from 87.219.178.58 on Sun, 26 Jun 2016 08:17:17 UTC
All use subject to http://about.jstor.org/terms
 THE TYPE CONCEPT IN SYSTEMATIC BOTANY'

A. S. HITCHCOCK

(Received for publication January 8, I92I)

The binomial system of nomenclature has been an important factor

in the development of taxonomy. The increase in the number of known
species since the time of Linnaeus has been many fold; because of care-
lessness and ignorance the number of names applied to the species of plants
has been much greater than the number of species; the increase in our knowl-
edge of genetic relationships and the diversity of opinions among botanists
concerning generic limitations have still further increased the synonymy.
The confusion arising from these causes soon emphasized the need of a code
of nomenclature by which the naming of plants might be regulated. Many
codes have been proposed, but only two have received the support of inter-
national conferences: the Paris Code of I867, and the Vienna Code of I905.
I have pointed out in another place (Science n. ser. 30: 597. I909)
that absolute stability in nomenclature is unattainable so long as botany is a
growing science. The limits of genera will vary according to the knowledge
and the opinions of individual workers, and the names of the plants as
they are assigned to this or that genus will change in a corresponding degree.
A universal code cannot bring about a permanent nomenclature, but it
enables botanists to apply names according to definite rules, and this is
all that we may expect of any code.
The two codes mentioned have been a great help in stabilizing nomen-
clature. Experience has shown, however, that they lack definiteness in
directing the application of names, especially of generic names. In the
early days of taxonomy a name was applied to a concept rather than to an
entity. A generic name was based upon all the known species of the genus;
a specific name was based upon all the known specimens of the species.
When a genus was divided the original name was retained for one of the
parts, usually the larger part, or was sometimes discarded altogether.
The Vienna Code introduced many reforms, but the procedure for applying
names when a genus or species was divided was still vague and uncertain
in its application.
About 30 years ago a new system began to receive serious attention
among American botanists, the system of applying names by means of
types. It is not my purpose here to give a history of this idea, but rather
to point out some of the advantages of the system. The type concept lies
1 Read before the Systematic Section of the Botanical Society of America at Chicago,
December 29, 1920.
251

This content downloaded from 87.219.178.58 on Sun, 26 Jun 2016 08:17:17 UTC
All use subject to http://about.jstor.org/terms
252 AMERICAN JOURNAL OF BOTANY [Vol. 8

at the basis of modern botanical nomenclature. The type species of a

genus or the type specimen of a species is the species or the specimen respec-
tively that directs or controls the application of the generic or specific
name. A generic name shall always be so applied as to include its type
species; a specific name shall always be so applied as to include its type
specimen. The old concept was that a genus was a group of species having
a given combination of characters; a species, similarly, a group of specimens.
The new or type concept is that, from the nomenclatural standpoint, a
genus is a group of species allied to the type species, a species a group of
individuals similar to the type specimen.
If a genus or species is divided, that part which includes the type species
or specimen retains the generic or specific name, be this part relatively
large or small. The American Code2 recognized the type concept as a
fundamental principle. The Paris and Vienna codes do not refer to this prin-
ciple. But the idea had made such headway by I9IO that it was recognized
by the Brussels Congress in a recommendation as a guide for the future
(an addition to Recommendation XVIII). This reads:

[Botanists will do well, in publishing, to conform to the following recommendations:

XVIII . . . ] XVIII bis. When one publishes the name of a new group, to indicate
carefully the subdivision which is considered to be the nomenclatural type of the group;
the type genus of a family, the type species of a genus, the type variety or the type specimen
of a species. This precaution avoids the nomenclatural difficulties in the case where, in
the future, the group in question comes to be divided. (Act. Congr. Internat. Bot. BruLX.
19IO I: I05.)

It is to be regretted that this recommendation was not made retroactive.

I feel confident that the retroactive fixation of nomenclatural types is a
fundamental necessity in stabilizing nomenclature. I feel confident also
that this aspect of the type concept will appeal more and more strongly to
the followers of the Vienna Code as its advantages are recognized, especially
as there is nothing in the concept that is contrary to the principles of that
code. One must carefully distinguish between the concept itself and the
rules for its application. The American Code has recognized the principle
of types and has also formulated rules for type fixation. One may accept
the principle and reject these particular rules.
The congress which adopted the Vienna Code appears to have been
actuated by a desire to formulate rules that should, in a general way,
preserve the current usage of generic names. I wish to point out to the
followers of the Vienna Code that this laudable purpose can be accomplished
with greater definiteness by applying the type concept than by applying
the vague and uncertain rules adopted by the Vienna Congress.
The Vienna Code contains the following rule:

ART. 45. When a genus is divided into two or more genera, the name must be kept and
given to one of the principal divisions. If the genus contains a section or some other

2Formulated in 1907 by a Nomenclature Commission of the Botanical Club of the

American Association for the Advancement of Science.

This content downloaded from 87.219.178.58 on Sun, 26 Jun 2016 08:17:17 UTC
All use subject to http://about.jstor.org/terms
 May, I92I] HITCHCOCK - THE TYPE CONCEPT 253

division which, judging by its name or its species, is the type or the origin of the group, the
name is reserved for that part of it. If there is no such section or subdivision, but one of
the parts detached contains a great many more species than the others, the name is reserved
for that part of it.

Let us apply this rule to the Linnaean genus Panicum. There are 20
original Linnaean species. Several of them, including P. miliaceum and
its allies, belong to the genus Panicum as delimited by most modern botan-
ists. Among the 20 are also P. italicutm and its allies, now generally dis-
tinguished as Setaria or Chaetochloa. But Panicum italicumn is the historic
type of Panicum, that is, the species which was known as Panicum by pre-
Linnaean authors and the one which I should interpret as, "judging from
its name or its species, is the type or the origin of the group," anld therefore
the segregated genus containing it should have retained the name Panicum.
However, in the process of taxonomic and nomenclatural development of
the various species involved, this procedure was not followed. If botanists
wish to retain the name for the allies of Panicum miliaceum, the simplest
method to insure this result is to select Panicum miliaceum as the type of
Panicum.
The Linnaean genus Holcus, presenting certain complications, illustrates
the advantage of the type method. The name in pre-Linnaean literature was
applied to the sorghums, but in the Species Plantarum Linnaeus unites with
the three species of the sorghum group four other species of diverse rela-
tionships, one of which is Holcus lanatus, the only one of the species belonging
to Holcus as now recognized by European botanists. The Vienna Code
provides (Art. i9) that

It is agreed to associate genera, the names of which appear in this work [Species Plan-
tarum] with the descriptions of them in the Genera Plantarum ed. 5 (I754).

According to the Vienna Code (as well as to the American and Type-
basis codes) the name Holcus should be applied to the sorghums and this I
have done, since the author's concept is most accurately interpreted by his
own description. But when the aggregate included under Holcus by Lin-
naeus in 1753 was divided, a century or miore ago, the sorghunms and species
of other genera were taken out and the name Holcus was left for H. lanatus,
which until recently has generally borne that name. The followers of the
Vienna Code have accepted current usage regardless of the rules of that
code. Would it not be simpler and more definite to make an exception and
to crystallize current usage by fixing Holcus lanatus as the type of Holcus?
Examples could be multiplied indefinitely. Apparently the rules of
the Vienna Code were left indefinite in order that botanists should not be too
much restricted in the application of names and should have some freedom
to use personal judgment. It is impossible to foresee all contingencies and
to provide for them by definite rules. As shown above, when, in par-
ticular cases, the rules lead in the wrong direction they are likely to be
ignored. The desired results can be accomplished with much greater

This content downloaded from 87.219.178.58 on Sun, 26 Jun 2016 08:17:17 UTC
All use subject to http://about.jstor.org/terms
254 AAMERICAN JOURNAL OF BOTANY [Vol. 8

precision by using the type method. I commend to the followers of the

Vienna Code the proposal that the International Rules be modified by a
recommendation to the effect that the application of names be fixed by
means of nomenclatural types, this to apply retroactively.
The American Code provides for fixing the application of names by
means of types. It goes further and provides rules for determining the
type. It should be emphasized that the acceptance of the concept of types
does not involve the acceptance of a particular set of rules for selecting
types.
The code formulated by the Committee oIi Nomenclature of the Botan-
ical Society of America is called the Type-basis Code of Nomenclature.
Like the American and the Vienna codes, the rules of the Type-basis Code
are founded on the principle of priority. The rules for selecting types of
genera and of species are in conformity with this principle, while, as stated
previously, the Vienna Code omits altogether the rules for selecting types
(though type appears incidentally in Art. 45 as indicated above). It will be
seen then that the chief difference between the Vienna Code and the new
Type-basis Code is that the one ignores the subject and the other formulates
rules for selecting types. If the Vienna Code could be modified to include
a set of acceptable rules governing the selection of types, the most impor-
tant difference between the two codes would disappear.
Attention should here be called to the fact that selecting the type of a
group does not validate the name of that group. Types are selected for
both valid names and synonyms. It only means that if a certain name is
used it should be so applied as to include the type.
I will review briefly the proposed rules for selecting the types of genera.
I will pass over certain particular cases such as those in which there was
but one species in the genus as originally published, or in which the type was
designated originally, and refer to the troublesome cases where there were
several species included in the genus as originally published. This is true
of many Linnaean genera, and the typification of these is basic so far as
stability of nomenclature is concerned. There was an attempt at one time
to select arbitrarily the first species as the type. This would be definite,
but would often run counter to the historic development of the group and
would cause so nmany changes in names as to introduce serious and needless
confusion. The new code provides for selection by applying the rule of
reason, taking into consideration all the factors in each case. In preparing
a recent bulletin I found it necessary to typify over 300 grass genera. I will
select a few examples from these. If the genus was used in his earlier
works, Flora Lapponica or Flora Suecica, the type should be chosen from
among those in the Species Plantarum that are cited by Linnaeus as being
in one of the earlier works, since these are the species with which he was
more familiar. Under Andropogon in the Species Plantarum Linnaeus
describes 12 species. The name Andropogon was first used in the Flora

This content downloaded from 87.219.178.58 on Sun, 26 Jun 2016 08:17:17 UTC
All use subject to http://about.jstor.org/terms
May, 1921] HITCHCOCK - THE TYPE CONCEPT 255

Leidensis where two species are described, both being included in the Species
Plantarum. Frorn these two Andropogon virginicus was chosen as the type
because that species retained the name in its usual significance. The
other species, A. hirtus, is now by many botanists referred to a different
genus.
Poa L. Linnaeus describes 17 species. He first used the genus in his
Flora Lapponica. From among the species there described Poa pratensis is
selected as the type because that retains the name of this economic species
in its usual signification.
Uniola L. Two species are described. One is referred now to Distichlis.
The other is selected as the type, thus retaining the name in its cturrent
usage.
Hordeum L. Six species are described. The reference in the Genera
Plantarutm is to figure 295 in Tournefort's work, representing Hordeum
vaulgare, the common barley, which is therefore selected as the type.
Aira L. Of the 14 species described four are included in the Flora
Lapponica. To take the first of these as the type would transfer the name
Aira to what we now call Trisetum. Hence another one of the four, A.
caespitosa, is selected in order to retain the name in its usual signification.
Some botanists apply the name Aira to the last two of the 14 original species,
including A. caryophyllea, and refer Aira caespitosa and its allies to Des-
champsia. These two species are from southern Europe and were not
included by Linnaeus in his first use of the term Aira in the Flora Lapponica,
and hence did not represent Linnaeus's original idea of the genus.
In general, one should ascertain if possible what species or group of
species an author had chiefly in mind in establishing a new genus.
The application of the type concept to species is similar. If more than
one specimen is cited, one should find which one the author had chiefly in
mind. This may be shown by comparison with the description, by one
having been selected for an illustration, by notes on the original sheet, by
the specific name. Only when other methods fail should the first specimen
cited be arbitrarily selected.
The above illustrates what is meant by applying the rule of reason in
the selection of types. Let us hope that soon all taxonomic botanists will
accept the concept of types and that they may agree on the types to be
selected.
BUREAU OF PLANT INDUSTRY,
WASHINGTON, D. C.

This content downloaded from 87.219.178.58 on Sun, 26 Jun 2016 08:17:17 UTC
All use subject to http://about.jstor.org/terms