Classification of the Architecture of Dicotyledonous Leaves

Author(s): Leo J. Hickey

Reviewed work(s):
Source: American Journal of Botany, Vol. 60, No. 1 (Jan., 1973), pp. 17-33
Published by: Botanical Society of America
Stable URL: http://www.jstor.org/stable/2441319 .
Accessed: 09/10/2012 09:28

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

.
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 support@jstor.org.

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

http://www.jstor.org
Amer. J. Bot. 60(1): 17-33. 1973.

CLASSIFICATION OF THE ARCHITECTURE OF DICOTYLEDONOUS

LEAVES'

LEO J. HICKEY
Divisionof Paleobotany,Smithsonian
Institution, D. C.
Washington,

A B S T R A C T
A classificationof the architectural featuresof dicot leaves-i.e., the placementand form
of those elementsconstituting the outwardexpressionof leaf structure,includingshape,
marginalconfiguration, venation,and glandposition-has been developedas the resultof an
extensivesurveyof bothlivingand fossilleaves. This systempartiallyincorporates modifica-
tionsof twoearlierclassifications:thatof Turrillforleaf shapeand thatof Von Ettingshausen
for venationpattern.Aftercategorization of such featuresas shape of the whole leaf and
of the apex and base, leaves are separatedintoa numberof classes dependingon the course
of theirprincipalvenation.Identification of orderof venation,whichis fundamental to the
is determined
applicationof the classification, by size of a vein at its pointof originand to
a lesserextentby its behaviorin relationto thatof otherorders.The classification concludes
by describing featuresof the areoles,i.e., the smallestareas of leaf tissuesurroundedby veins
whichforma contiguousfieldover most of the leaf. Because most taxa of dicots possess
consistentpatternsof leaf architecture, this rigorousmethodof describingthe featuresof
leaves is of immediateusefulnessin both modernand fossiltaxonomicstudies. In addition,
as a resultof this method,it is anticipatedthat leaves will play an increasingly important
partin phylogenetic and ecologicalstudies.

LEAVES are generallyneglectedin taxonomicand Tertiaryflora,which will be reportedelsewhere

comparative morphologicstudies due in large (Hickey, in press).
part to the lack of a detailed,standardized,un- In thisand in subsequentreportsI will use the
ambiguousclassification of theirfeatures.Students term"leaf architecture" to denote the placement
ofmodernplantshave been content,in mostcases, and formof those elementsconstituting the out-
withbriefdescriptions of leaf outline,margin,and ward expressionof leaf structure, includingvena-
major vein configuration.Most paleobotanists, tion pattern,marginalconfiguration, leaf shape,
forwhomleaves representthe majorityof angio- and gland position. This termis appropriatebe-
sperm remains,have also neglectedto go much cause the elementsof leaves are organizedinto
beyond the use of comparativelysuperficialleaf certaindefinitestructuralpatternscapable of de-
featuresin theiridentifications.This has resulted scription,thus conformingto the definitionof
in a prohibitivelyhigh percentageof incorrect architectureas "formationor construction whether
genericassignments,especiallywith leaves older the result of conscious act or of growthor of
than the mid-Tertiary(Pacltova', 1961; Wolfe, randomdispositionof parts" (Webster'sNew In-
1966, 1968, 1969; Doyle, 1969; Dilcher and ternationalDictionary,thirdedition).Architecture
Dolph, 1970). is the aspect of morphologywhichapplies to the
The purpose of this reportis to presentan spatial configurationand coordinationof those
inclusive,coherentclassificationof the architec- elementsmakingup partof a plantwithoutregard
ture of dicotyledonousleaves. Many of the fea- to histology,function,origin,or homology. The
tures to be enumeratedare also found in the termwas used in thissense by Devadas and Beck
monocots,but no specificattempthas been made (1971) and the specificterm"leaf architecture"
to extendtheclassificationto them. The proposed by Hickey (1971a) and Delevoryas and Gould
systemis based in part on portionsof previous (1971).
classificationsand has been modifiedand ex- Botanical termsdescribingspecificaspects of
panded as the result of four years of research leaves are in ample supply,largelybecause of
with the collectionsof the U. S. National Her- the influenceof early systematists such as Lin-
barium. The initiativefor developing such a naeus, Bentham,Lindley, and Asa Gray. Un-
scheme resultedfrommy research on an early fortunatelythesetermsfrequently have ambiguous,
1Receivedfor publication28 December1971. overlapping,or imprecisemeaningsand are not
Research conducted,in part, while an NAS-NRC- organizedinto any coherentsystem. The avail-
SmithsonianVisitingPostdoctoralResearch Associate. able descriptiveterminology is mainlyconcerned
Gratefulacknowledgment is made of the advice and
criticalcommentof Dr. Edward S. Ayensu,Dr. James withthe shape of the leaf and its extremitiesand
A. Doyle, Dr. WilliamL. Stern,and Dr. JackA. Wolfe. themostobviousaspectsofitsmajorveinpatterns.
17
18 AMERICAN JOURNAL OF BOTANY [Vol. 60

The methodof classifying leaf shape employed pallywiththedevelopmental morphology of vena-

here is based on a scheme developed by W. B. tion,but neitherof these had any importantin-
Turrillat theRoyal BotanicGardens,Kew, about fluenceon the terminology used here.
1925. First publicationof the method was by The ability to catalogue and describe leaf
Lee (1948), followedbymorecompletesummaries characterspreciselywill add considerablyto the
by Stearn (1956) and Krussmann(1960). The usefulnessof leaves as taxonomic featuresfor
systemrepresentsa considerableimprovement in botanistsand paleobotanistsalike. In addition,a
terminology because it establishespreciselydefined more sophisticatedknowledgeof leaf architecture
shape categoriesforleaves, based on the position has permitteda startto be made in discerning
of the axis of greatestwidthon the long axis of phylogenetic trendsfromleaves (Hickey, 1971b),
theleafand theratiooflengthto width(l/w ratio). whichmay have importantbearingon the study
Apical and basal shapes were also given more of modernand fossilgroups. Finally,theexistence
preciseformulation.The onlyothershape classi- of a bodyof carefullydefinedtermsforleaf archi-
ficationin use is that of the SystematicsAsso- tecturemay facilitatethe studyof environmental
ciation Committeefor DescriptiveTerminology effectson leaves.
(1962). This systemalso definesshape classes
using1/wratioand positionof theaxis of greatest MATERIALS AND METHODS-In developingthis
widthbut includesothercriteriasuch as theshape classificationthe leaf architecture of 1212 genera
of the margins,thus arrivingat a more elaborate in 135 families of dicotyledonswas surveyed,
and less generalizedclassificationthanthe Turrill using uncleared leaves in the U. S. National
system. Herbarium.A clearedleafcollection,now amount-
Constantinvon Ettingshausenmade the first ing to 473 species in 223 generaand 70 families,
comprehensiveeffortto systematizethe descrip- was made forthisstudyin orderto examinefiner
tion of leaf architecture withhis classificationof details such as areolation patternsand higher
venationpatternsin 1861 (Foster, 1952). His ordervenationwhichare not visiblein uncleared
categorizationproceeds in logical fashion from material. These clearings were made by the
the configuration of the primaryveins to that of methodof Foster (1952) with modifications by
the areolation. Althoughit is over-elaborateand Wolfe (personal communications)and myself.
mixesfeatureshavingtaxonomicutilitywithothers These modifications includethe clearingof some
having no significance,I have found it a useful leaves (freshor herbaceous) in commercialbleach
startingpointforthevenationalclassification pro- solutions (approximately5 % sodium hypochlo-
posed here. Unfortunately Von Ettingshausen's rite) instead of 5 % sodium hydroxide;the use
systemnevercame into widespreaduse, although of plastic screeningto weightthe leaves during
the paleobotanists Lesquereux (1878), Berry clearing; of toluene rather than xylene as the
(1916), and Hollick (1936) used a few of his solventforthe mountingmedium;and the use of
termsto describethe configuration of the lower a hard-rubberphotographicroller to flattenthe
ordersof venation. petiole and midveinof the leaf duringmounting.
Later, Kerner (1895) formulateda classifica- In constructing thisclassification,
an effortwas
tion,treatingonlytheprimaryand secondaryvein made to includea large numberof taxonomically
pattern. Althoughthis systemappears to be at significant lower orderfeatureswhichcan be ob-
least partly derived from that of Von Ettings- servedin unclearedleaves and whichare mostapt
hausen,the approach used is somewhatdifferent to be preservedin fossilleaves.
from the original and, regrettably, Kerner em- The architecture of fossilangiospermleaves of
ployedsome of the same or closelysimilarterms all ages, fromtheEarly Cretaceousto the Pleisto-
for different patternsof venation.A modernap- cene, was surveyedby examiningsubstantialpor-
plication of Kerner's terminologycan be seen tions of the U. S. National Museum collections2
in Krussmann's"Handbuch der Laubegeh6lze" and all cards in the Type Catalogue of North
(1960). Anotherderivativesystemwas outlined AmericanFossil Plants maintainedat Princeton
by Takhtajan (1963). Universityby ProfessorErlingDorf. (This cata-
One important exampleof a totallyindependent logue includes illustrationsof virtuallyall de-
classificationof some elementsof leaf architecture scribed fossil species of North American angio-
is found in Lam's (1925) monographon the sperms.)
Sapotaceae of the East Indies. Lam's recognition Fossil and modernleaf
of a numberof taxonomicallysignificant charac- architecture were com-
teristicsof secondaryand tertiaryvenation,such 2
Amongspecificcollectionsexamined,includingthose
as their angular relationships,curvature,paths of the author,are: the Potomac Flora (Lower Creta-
both insidethe leaf and near its margins,branch- ceous); the Raritan,Dakota, and Lance Floras (Upper
Cretaceous);the Fort Union,Raton,Denver,and lower
ing patterns,and vein junctions,was importantin Golden Valley Floras (Paleocene); the upper Golden
influencingmy choice of such featuresin the Valley,Wind River,YellowstonePark, KisingerLakes,
presentsystem. Both Goebel (1905) and Troll Wilcox, Green River, Claibourne,and JacksonFloras
(Eocene); the FlorissantFlora (Oligocene); and the
(1938) developed classificationsdealing princi- Latah Flora (Miocene).
January,1973] HICKEY-ARCHITECTURE OF DICOTYLEDONOUS LEAVES 19

pared by means of high contrastphotographs mean to which the familyis adapted. For ex-
(takenon Dupont OrthoA Litho or KodalithPan ample, most of the Theaceae live in paratropical
plate films) or camera lucida drawingsof the and subtropicalforestswhere the majorityof
fossilsand highresolutionphotographsof modern thempossess the characteristic glandularserrate
leaves (made fromWrattenM glass plate nega- marginsmentionedabove. Species livingin drier,
tives of the whole, cleared leaf taken with the more open areas (Laplacea and Ternstroemia),
anti-halationbacking toward the camera lens). in highmontanesettings(Eurya), or in the low-
A leaf architecturalclassificationcould include land tropics (some Gordonia and Shima sps.)
an almostinfinitenumberof features.The taxo- show a trendfirsttowardloss of the serrations
nomicutilityof thosefeatureschosenforthissys- and finallyto the loss of marginalglands. As a
tem was evaluated by assemblingdata sheets generalrule,thesmall,coriaceous,entire-margined
showing their distributionamong various taxa. leaves ofxeric,arctic,or alpineenvironments show
These featureswere then used in the identifica- relativelyfew of the charactersneeded for taxo-
tionof fossilleaves in the earlyTertiary,Golden nomic identification.
Valley Flora (Hickey, in press) and to separate The outlinebelow beginsby describingfeatures
varioustaxa of modernleaves fromthe ordinalto of the whole lamina,such as directionswithinit
the specific level, thus providinga continuous and its form. For the sake of simplicitysimple
testof thecomponentsof thisclassification during leaves and leafletsare treatedtogetheras laminae.
its formulation.An earlyTertiaryfloraprovided Arrangement of the elementsof compoundleaves
an importantimpetusfor this studybecause its is not discussedin thissummarysince theyhave
highcontentof extinctformsand thelargenumber been adequatelydescribedelsewhere,e.g., Law-
of generic misidentifications secured by super- rence (1951) and Federov, Kirpichnikov,and
ficial and traditionalmatchingtechniquesboth Artushenko(1956). Afterconsideringthe shape
necessitatedthe developmentof a more precise of the parts of the leaf and position of glands,
approach. Examples of the use of thisclassifica- the systemproceeds to an examinationof vein
tionin the identificationprocess as well as actual configuration, startingwiththe primaryand sec-
photographsof leaves displayingthe featuresde- ondary veins which determinethe major vein
scribedwill appear in the reporton the Golden classes, such as pinnate, acrodromous,or acti-
Valley Flora (Hickey,in press). nodromous(see below). Proceedingas in theVon
Ettingshausenclassification,the present system
SCOPEANDBASISOF THECLASSIFICATION-The describesthe traitsof progressively higherorders
essential justificationfor the classificationpre- of venation,terminating withthose of areolation.
sented here is the fact that the various taxa of In view of the rudimentary state of our present
dicotshave leaves possessingconsistentand recog- knowledgeof leaf architecture,the system set
nizable patterns of architecturalorganization. forthhere should not be regardedas a finalfor-
Most membersof the Theaceae, forexample,ex- mulationbut ratheras open-endedand subjectto
hibit serrationsof a characteristicshape, with modification as more information accumulates.
glandular setae; have pinnate, camptodromous All termslisted in the followingoutline are
venation with elongated intercostalareas and illustratedand defined,exceptwheretheirmean-
large, loosely organized irregular areolation. ing is obvious. AlthoughI have set limitsbased
Charactersets of familiesor genera sometimes on observedbreaks in morphologicalfeaturesin
overlap,as in the case of the lauraceous genera, describingsome characters,such as size of pri-
makinga positiveassignmentimpossible. Sets of maryveins and of areolation,in manycases it is
charactersprovingsignificant forrecognizingone not possible to avoid being somewhatarbitrary.
taxonmaybe completelydifferent fromthosedis- Thus, althoughmostacute leaf apices can be dis-
tinguishing another,althoughin generalareolation tinguishedfrommost acuminateones, therewill
and marginal features are rather reliable (cf. be a numberwhichfall in a transitional zone; the
Wolfe, 1968). same sortof transition is presentbetweenthe im-
Some familiesand generacontainseveralbasic perfect actinodromousvenation class and the
patternsof leaf architecture.This is particularly pinnateclass wherethe lowestpair of secondary
true of "artificial"or paraphyleticfamiliessuch veinsis set at an angledifferent fromthoseabove.
as theEuphorbiaceae. In otherfamiliesvariation Such difficultiesare inevitable in imposing a
fromthe basic patternappears in thosegeneraor classificationon naturalfeaturesand in no way
species in environments more extremethan the detractfromthe utilityof doingso.

OUTLINE OF LEAF ARCHITECTURAL CLASSIFICATION

Orientation-The basic axes of orientationin the leaf are indicatedin Fig. 1 (as modifiedafter
Wolfe,writtencommunication, 1968):
A. Apical-toward the apex (upward).
B. Basal-toward the base (downward).
20 AMERICAN JOURNAL OF BOTANY [Vol. 60

APICAL >900
O,XI\\8
9 10
ACUTE OBTUSE
EXMEDIAL \ / 1 1 13i
...... ..
ADMEDIA O
/
'
,' AOMIDIAL
RETUSE
pi l | MT
EMARGINATE tV

1 \ ~ .. ...
14,-t 15,--r- SETA APICAL
BASAL 2 0 MUCRONATE ROUNDED
ORIENTATION 16TNy2 Z

LEAF APEX ATTENUATE |l

111 V IJ-------- I 18 19 20

OBLONG ELLIPTIC OVATE

LEAF FORM
ACUTE OBTUSE CUNEATE
PETIOLAR { -
I=

(o)} WHOLE (b) BASE ONLY 21 22! 23 o

LAMIlNA /> ROUNDED DECURRENT TRUNCATE
28 4

( (0) | ~~~~~~24
I25 24~'2~'~~26(//}\\J
26 POSITION
GLAND
CORDATE LOBATESATAE
7 8 LEAF BASE
ASYMMETRICAL _._ _ _ _

MARGINS SINUS(ANGULAR)
APICALSIDE

APEX
30 ~~~~~~~~~~~~~~TOOTH
DENTATE 32 33 34
--
o C BASALSIDE
-7 CRENATE / REGULAR
IJR
31 REGULAR SPACING 35
LOBED 29
LOBED 29 SERRATE SPACING
PARTS OF A TOOTH

SERRATIONTYPES
APICAL SIDE
CONVEX STRAIGHT CONCAVE ACUMINATE

EROSE36 3|ROUNDED
SINUSES
/

J =t 38 REVOLUTE }~~~~-|--
TEETH
ORDERS3OF
..
A B C
i-W ...
D... 40

38

nORIDER OF TEETH n4
January,
1973] HICKEY-ARCHITECTURE OF DICOTYLEDONOUS LEAVES 21

C. Exmedial-awayfromtheleafaxis. Notsynonymous withabaxialwhichimpliesthepresence

of a physicalaxis suchas a stem,ratherthanan axisof symmetry.
D. Admedial-towardthe leaf axis. Not synonymous withadaxial whichtermrejectedfor
reasonsgivenabove.
The curvature ofleafelements (marginor partthereof,
sidesof a serration,
venation)is re-
ferredto as (Fig. 2):
E. Convex-curvedawayfromthecenterof theleaf,or its axis,or (forvenationin theacti-
nodromous class) thepointoforiginoftheprimaryveins.
F. Concave-curvedtowardthecenterof theleaf,etc.
II. Shape.
A. Lamina.
1. Balance.
a. Wholelamina.
1) symmetrical (Fig. 1).
2) asymmetrical (Fig.7).
b. Base only.
1) symmetrical (Fig. 1).
2) asymmetrical (Fig. 8).
2. Form(modified afterStearn,1956).
a. Oblong-widestportion constitutinga zonethrough themiddleofthelongaxisofthe
leaf,margins parallelor nearlyso within thiszone (Fig. 3). The length-width (l/w)
ratiosgivenforthesubclassesare lowerlimitsexceptforthelast.
1) linear l/w ratio 10:1 or more
2) lorate l/w ratio 6:1
3) narrowoblong l/w ratio 3:1
4) oblong 1/wratio 2:1
5) wide oblong I/w ratio 1.5:1
6) verywideoblong l/w ratio 1.2:1 or less
b. Elliptic-axisof greatest widthperpendicular to theapproximate oftheleaf
midpoint
axis (Fig. 4). Marginsin thisand succeeding classesconvexor somecombination of
convexand concave.
1) verynarrowellipticl/w ratio 6:1 or more
2) narrowelliptic 1/wratio 3:1
3) elliptic 1/wratio 2:1
4) wide elliptic 1/wratio 1.5:1
5) suborbiculate 1/wratio 1.2:1
6) orbiculate 1/wratio 1:1
7) oblate 1/wratio 0.75:1 or less
c. Ovate-axis of greatest widthintersecting theleafaxis basal to themidpoint of the
latteraxis (Fig. 5).
1) lanceolate 1/wratio 3:1 or more
2) narrowovate 1/wratio 2:1
3) ovate 1/wratio 1.5:1
4) wide ovate 1/wratio 1.2:1
5) verywide ovate 1/wratio 1: 1 or less
d. Obovate-axis of greatest widthintersecting thelongaxis of theleaf apical to the
midpoint ofthelatteraxis(Fig. 6).
1) narrowoblanceolate 1/wratio 6:1 or more
2) oblanceolate 1/wratio 3:1
3) narrowobovate 1/wratio 2:1
4) wide obovate 1/wratio 1.2:1
5) verywideobovate 1/wratio 1:1 or less

and formof wholeleaf,shape of apex and base, glandposi-

features-orientation
Fig. 1-40. Leaf architectural
tion,and marginalconfiguration.
22 AMERICAN JOURNAL OF BOTANY [Vol.60

e. Special shapes (includingneedleor awl shaped,reniform,

deltoid,spatulate,etc.).
B. Apex-that portionof the leaf boundedby approximatelytheupper25 % of theleaf margin.
1. Acute-straight to convex marginsformingan angle of less than90? (Fig. 9).
2. Acuminate-tip acute,marginsmarkedlyconcave,eitherlongor shortacuminate(Fig. 13).
3. Attenuate-marginsstraightor only slightlyconcave, graduallytaperingto a narrow
acute apex (Fig. 17).
4. Obtuse-straightto convex marginsformingan angle of more than900 (Fig. 10).
5. Rounded-margins forminga smootharc across the apex (Fig. 15).
6. Mucronate-apex terminating in a sharp point which is a continuationof the midvein
(Fig. 14).
7. Retuse-apex slightlynotchedas if by removalof thetermination of themidvein;internal
angle of the sinus generallyless than 250 (Fig. 11).
8. Emarginate-apex broadly notchedby the embaymentof the leaf tissue (Fig. 12).
9. Truncate-Apex terminating abruptlyas if cut, marginperpendicularto the midveinor
nearlyso (Fig. 16).
10. Other.
C. Base-that portionof the leaf bounded by approximately the lower 25 % of the margin.
1. Acute-margins formingan angle of less than90?.
a. normal-base with curved marginsterminatingat the petiole withoutappreciable
change in direction(Fig. 18).
b. cuneate-marginsstraight, or nearlyso, and forminga "wedge" of less than90? (Fig.
20).
c. decurrent-marginextendingdownwardalongthepetioleat a low angleto it (Fig. 22).
2. Obtuse-margins formingan angle of greaterthan 90? (Fig. 19).
a. normal-(as above).
b. cuneate-marginsstraightor nearlyso and forminga "wedge" of greaterthan 90?
(rare).
c. decurrent-(as above).
3. Rounded-margins forminga smootharc across the base (Fig. 21).
4. Truncate-terminating abruptlyas if cut, marginperpendicularto the midveinor nearly
so (Fig. 23).
5. Cordate-leaf base embayedin a sinuswhose sides are straightor convex (Fig. 24).
6. Auriculateto Lobate-small to large roundedprojectionswhose inner margins(those
towardthe petiole) are in partconcave (Fig. 25).
7. Saggitate-with two large pointedlobes whose apices are directeddownward,i.e., at an
angle of 450 or less fromtheleaf axis (Fig. 26).
8. Hastate-with two large pointedlobes whose apices are directedoutward,i.e., at an angle
of greaterthan450 fromthe leaf axis (not illustrated).
9. Peltate-petiole attachedinside the leaf margin(not illustrated).
10. Other.
III. Margin.
A. Entire-margin forminga smoothline or arc withoutnoticeableprojectionsor indentations.
B. Lobed-margin indented1/4 or more of the distanceto themidveinor (wherethisis lacking)
to thelong axis of the leaf (Fig. 29).
C. Toothed-margin havingprojectionswithpointedapices,indentedless than /4of thedistance
to the midveinor long axis of the leaf (Fig. 30, 31, 33-35, 37, 38).
1. Dentate (Fig. 30)-dentations are pointedwithaxes approximately perpendicularto the
trendof the margin;as withleaf apices these can be acute (Fig. 9), obtuse (Fig. 10),
acuminate(Fig. 13), attenuate(Fig. 17), or mucronate(Fig. 14) (definitionsas under
apices, above).
2. Serrate (Fig. 31 )-serrations are pointed with theiraxes inclined (i.e., at an oblique
angle) to the trend(tangent) of the margin.
January,
1973] HICKEY-ARCHITECTURE OF DICOTYLEDONOUS LEAVES 23

a. Apical angle.
1) Acute-angle formedby the two sides less than90? (typical).
2) Obtuse-angle formedby the two sides greaterthan90? (rare).
b. Serrationtype-determinedby the shape of the basal side (shown on the verticalor
letteredlines of Fig. 40) vs. the shape of the apical side (shown on the horizontalor
numberedlines of Fig. 40). Families and generaoftenexhibita highdegreeof con-
sistencyin theirpossessionof one or twotypesof serration.See Fig. 35 fororientation.
The configurations used for each side are given in Table 1.
D. Crenate (Fig. 32)-crenations are smoothlyrounded,withouta pointedapex.
E. Erose (Fig. 36)-irregular, as if chewed.
F. Revolute or enrolled (Fig. 39)-margin turnedunder or rolled upon itselflike a scroll-
applies to both entireand non-entiremargins.
G. Sinuses-incisions betweenmarginalprojectionsof any sort-lobes, dentations,serrations, or
crenations.
1. Rounded (Fig. 37)-margins of sinus meetingin a smoothcurve.
2. Angular (Fig. 35)-margins of sinusmeetingat a point.
H. Spacing-intervalbetweencorresponding pointson the teethor crenations.
1. Regular (Fig. 33)-interval varyingby no more than 25 %.
2. Irregular(Fig. 34)-interval varyingby morethan25 %.
I. Series-teeth separatedinto size groups.
1. Simple(Fig. 33)-teeth all of one size.
2. Compound (Fig. 38)-teeth in two or moredefinitesize groups-double serrations,etc.
IV. Texture.
A. Membranaceous-thinand semi-transparent, like a finemembrane.
B. Chartaceous-opaque and like writingpaper.
C. Coriaceous-leathery,thick,stiff.
D. Other.
V. Gland Position (includesnectaries,hydathodes,tanniniferous glands,etc.).
A. Petiolar(Fig. 28)-on thetissueof thepetiole;includesacropetiolar(Fig. 28)-at thetop of
the petiole.
B. Basilaminar(Fig. 28)-on thefoliartissueat the base of the blade.
C. Laminar (Fig. 28)-generally distributed on thefoliartissue.
D. Apical (Fig. 28)-on theleaf apex.
E. Marginal-distributedon the marginor marginalprocesses.
1. At themarginin entiremarginalleaves (Fig. 28).
2. On the teeth.
a. As a glandularthickening(Fig. 28).
b. As a glandularseta or bristle(Fig. 27).
3. In the sinus (Fig. 28).
VI. Petiole(ule).
A. Normal-withoutnoticeablethickenings or otherprocesses.
B. Inflated-thickened,includespulvini.
C. Winged-with a narrowstripof blade tissueon each side.
D. Absent-blade sessile, arisingdirectlyfromthe axis of attachment,
withoutan intervening
bladeless area of the leaf.
VII. Venation.
A. Type (for definitions of vein orderssee p. 25).
1. Pinnate-with a singleprimaryvein (midvein) servingas the originforthe higherorder
venation.
a. Craspedodromous3-secondaryveins terminating at the margin.
1 ) Simple-all of the secondaryveins and theirbranchesterminating
at the margin
(Fig. 41).
3 From the Greek kraspedon, edge, border; and dromos, a running,course.
24 AMERICAN JOURNAL OF BOTANY [Vol.60

2) Semicraspedodromous-secondary veins branchingjust withinthe margin,one of

the branchesterminating at the margin,the otherjoiningthe superadjacentsec-
ondary(Fig. 42).
3) Mixed-some of thesecondaryveins terminating at the marginand an approxi-
matelyequal numberof (usuallyintervening)secondariesotherwise(Fig. 43).
b. Camptodromous-secondaryveins not terminating at the margin.
1) Brochidodromous4-secondaries joined togetherin a series of prominentarches
(Fig. 46).
2) Eucamptodromous-secondariesupturnedand graduallydiminishing apically in-
side the margin,connectedto the superadjacentsecondariesby a series of cross
veins withoutformingprominentmarginalloops (Fig. 47).
3) Reticulodromous-secondarieslosingtheiridentitytowardthe leaf marginby re-
peated branchinginto a vein reticulum(Fig. 48).
4) Kladodromous5-secondariesfreelyramifiedtoward the margin (Fig. 49).
c. Hyphodromous-all but the primaryvein absent,rudimentary, or concealed withina
coriaceous or fleshymesophyll(Fig. 44).
2. Parallelodromous-two or more primaryveins originating beside each otherat the leaf
base and runningparallelto the apex wheretheyconverge(Fig. 45).
3. Campylodromous6-severalprimaryveins or theirbranches,originatingat, or close to,
a singlepointand runningin strongly developed,recurvedarchesbeforeconverging toward
the leaf apex. Vein patternconvergentabove and below (Fig. 58).
4. Acrodromous-two or more primaryor stronglydeveloped secondaryveins runningin
convergentarchestowardthe leaf apex. Archesnot recurvedat base (Fig. 59-62).
a. Position.
1) Basal-acrodromous veinsoriginating at thebase of the leaf (Fig. 59, 61).
2) Suprabasal-acrodromous veinsoriginating some distanceabove theleafbase (Fig.
60, 62).
b. Development.
1) Perfect-acrodromousveins well developed,runningat least 2/3 of the distance
to the leaf apex (Fig. 59, 60).
2) Imperfect-acrodromousveins runningless than 2/3 of the distanceto the leaf
apex (Fig. 61, 62).
5. Actinodromous-threeor more primaryveinsdiverging radiallyfroma singlepoint (Fig.
50-56) or
6. Palinactinodromous-primaries havingone or more subsidiarypointsof radiationabove
the lowest point,e.g., Platanus (Fig. 57).
(The followingcategoriesapply to 5 and 6 above.)
a. Positionof the firstpointof primaryvein radiation.
1 ) Basal (Fig. 50, 52, 54, 55) initialpointof radiationat the leaf base.
2) Suprabasal (Fig. 51, 53, 57) initialpointofradiationlocatedsome distanceabove
the leaf base.
b. Development.
1) Perfect-ramificationsof the lateralactinodromousveins coveringat least 2/3 of
the blade area (Fig. 50-53).
a) Marginal-actinodromousveins reachingthe margin (Fig. 50, 51).
b) Reticulate-actinodromousveins not reachingthe margin(Fig. 52, 53).
2) Imperfect-veinsoriginatingon the lateralactinodromousprimaryveinscovering
less than 2/3 of the blade area.
a) Marginal (Fig. 54).
b) Reticulate(Fig. 55).
3) Flabellate-several to manyequallyfinebasal veinsdivergeradiallyat low angles
and branchapically (Fig. 56).
4 brochosGr.,a noose.
5 kiados Gr.,abranch.
6 campylosGr.,a bendorcurve.
 1973]
January, HICKEY-ARCHITECTURE OF DICOTYLEDONOUS LEAVES 25

Orders of venation-In most leaves venation TABLE 1. SerrationTypes

is clearly differentiated into a number of size
classes. Veins of a particularsize class also dis- Marginal configuration basal apical
play some degree of uniformity in theircourses 1) Convex-Fig. 40 lines A & 1
and patternsof distributionin relationto other 2) Straight-Fig.40 lines B & 2
classes and to the marginalfeaturesof the leaf. 3) Concave-Fig. 40 lines C & 3
In practicethe objectivedesignationof vein order 4) Acuminate-Fig. 40 lines D & 4
is more complexthan easily observeddifferences
in thickness,course, and pattern among size
classes would seem to indicate. However, the Lateralprimary veinsfrequently occuras supra-
recognitionof vein ordersis essentialin describ- basal branchesof a primaryvein. In most ex-
ing leaf architecture. amples,the thicknessof theseveins at theirpoint
The primaryvenationof the leaf servesas the of branchingwill be the same, or verynearlythe
startingpointforthe identification of the various same, as the continuationof their source. In
vein orders. Veins of the primaryorder are the othercases the lateral primarybrancheswill be
thickestveins of the leaf and occur singlyor as somewhatthinnerthan theirprimarysource but
a medial vein accompanied by others (lateral still considerablythickerthan the next thinner
primaries) of roughly equal thickness. These set of veins,i.e., the secondaries. Here a number
emergefromthe petioleof theleaf,or the medial of featurespermitthe recognitionof theselateral
and lateralprimariesmay give rise to additional branchesas primaryveins. The primarybranches
lateralprimaryveinsabove thebase. Recognition will give rise to secondaryveins whose size and
of these branchesas primaryveins is based on behavioris no different fromsecondariesarising
the criterionthatthe branchbe roughlyequal in on primaryveins which originateat the base of
thicknessto its primarysource when both are the leaf. The primarybranchmay formthe mid-
measuredjust above the pointof branching.The vein of a leaf lobe as do the otherprimaries;and
nextset of branchesof markedlysmallersize than in all otherways its behaviorwill be comparable
theirprimarysourceare thesecondaryveins,while to that of the othermembersof its order (Fig.
the next finerset arisingfromboth primaryand 57). However,if thethicknessof a branchof the
secondaryveins are designatedthe tertiaryveins primaryvein is the same as that of the typical
and so on, untilreachingthe ultimatevein order secondaryveins, it is regarded as a secondary
presentin the leaf. despite some differences in its behavior (usually
angle of origin) in relation to them.
The primaryrule for the determinationof
Branches of the secondaryveins ("outer sec-
the order of a vein is its relative size at its ondary nerves" of Von Ettingshausen,a-2? of
point of origin. Where a lateral vein branch is Fig. 41) are frequently of the same thicknessat
approximatelyequal in width, measured just theirpoints of branchingas the continuationof
above the pointof branching,to the continuation the secondaryveins fromwhich they arise. In
of thesourcevein just above the pointof branch- cases where the branches are slightlythinner
ing, both branchesare of the same order;where thereis no difficulty in classifyingthemas sec-
thelateralvein branchis markedlyfinerthanthe ondaries because of their geometricrelationships.
continuationof its source, that branch is of a However,in serieswherethesize of thesebranches
higherorder. This rule is applied when tracing graduallydiminishestowardthatof thethirdorder
theirbehavioralso graduallyaltersuntilby both
a vein of any order distallyfrom its point of
criteriatheyhave become indistinguishable from
origin. the tertiaries.
Subsidiaryto size at the point of branchingin Intersecondary veins are intermediate in thick-
the recognitionof vein order is a consideration ness betweensecond and thirdorder veins and
of the behaviorof a vein in relationto veins of most often occur interspersedbetween the sec-
otherordersand to marginalfeaturesof the leaf ondariesof pinnateleaves (e, Fig. 63). Although
blade. It is necessaryto considerthis additional theirlengthis shorterthan that of the secondary
criterionwhen dealingwithordershavinga wide veinstheircourseis parallel,or nearlyso, to them,
range of sizes or where the boundariesbetween and they more closely resemblethe secondaries
size classes are not particularlysharp. Three of in theirrelationshipto otherordersof veins.
the mostcommonsituationswheredesignationof Otherdifficulties in designating vein orderarise
as the tendencytoward dichotomousbranching
vein orderis difficultinvolve: increasesin the highervein orders. However,the
a) suprabasal lateral primaryveins-i.e., pri- behavior of the resultantbranches frequently
mary veins originatingas lateral branches serves to distinguishseparate vein orders. In
of a primaryvein above thebase of theleaf other cases the higherorder venationof a leaf
b) branchesof secondaryveins mergesinto a reticulumin which designationor
c) intersecondary veins. vein orderis impossible(Fig. 79, 80).
26 AMERICAN JOURNAL OF BOTANY [Vol. 60

TYPES OF VENATION I HYPHODROMOUS

IPARALLElO-
IDROMOUS
CRASPEDODROMOUS D

I~~~~~~~~

e \\ / / / iSEM ICRASPEDODROROUSCRASPEDODROMOUS
w 44 1 45
220MXE CAMPTODROMOUSS I

2'

PERF CAMPTODROMOUSOU
a-?0
41
I T~46 47489
SIMPLE CRASPEDODROMOUS BROCHIDODROMOUS
EUCAMPTODROMOUSRETICULODROMOUSCLADODROMOUS

ACTINPDROMCU PEFC I ACRODROMOUS

I
MARGINAL RETICULATE
rw
A A
^ 2 8 PERFECT
A

10 1A < 1/3

51 BASAL SUPRABASAL 5 6
50 SUPRABASAL j BASAL SUPRABASAL
BASAL
IMPERFECT K % L o u
I IMESFET
IA
I CAMPYLODROMOUS
~~ ~ ~~I IMPERFECT
I ~~~A

54M55AI I I 1010
MARGINAL RETICULATE

PALINACTINODROMOUS I 6
FLABELLATE 57
ttUt4@%JFLABELLATE [ ~ 57 158
5 58 I BASAL SUPRABASAL

56

Fig. 41-62. Leaf architectural

features(continued)-typesof venation.
 1973]
January, HICKEY-ARCHlTECTURE OF DICOTYLEDONOUS LEAVES 27

COURSEOF 1?

b ~~~
~~~~~~~~~65
66
e ZIG-ZAG
~~~~~~~~SINUOUS
2 BRANCHES VARIATION IN 2cANGLE
VNOF DIVERGENCE

ge N IIETA NTRAMARG
INAL

73 {X ~ 3

K
72~~~~~2
RAMIFIOREDRRHOOA
DISTINCT 4 5"~RTICLAT
a
RAMIFIEDR
7H5 3 EXMIAL
8
RAMIFIED9 7
76~RNDML
OIENDTE7
73~ ~~ ~~~~~~~~1

WEAKLY COMPOSITE RETICULATE

PERCURRENT INTERSECONDARY
- 82_-?- 83 85~~~~7 8688 - 30 3? N3

RELATIONSHIP
OF TERTIARYVEINSTO MID-VEIN 80

RAMIFIHG ETCUAT H ORTHOGONAL

30AGEADICLAR CONSTANT
Fig.63-7.eafarciul RANDOMes
(contiued)-odersovRETICULATE

OUTWARD PWARD j HIGHERVI ORDERSVNTO

DISTNCT-4OM5ORETHOGONALE
ICL_ 20~~~~~~~~~~~~~~~~OLIU
Fig 6387.Lea arhitctual eatres(cotined)ordrs f vnaton nd ei cofgrain
28 AMERICAN JOURNAL OF BOTANY [Vol.60

1. Primary Veins(1?)-the thickestvein(s)of theleaf,occurring eithersinglyas themidvein, or as

a seriesofveinsofrelativelyequalthicknessemergingfromthepetiole(Fig.50, 52), or as branches
of themedialor lateralprimary veinsabovethebase of theblade (suprabasalprimaries).In the
lattercase, boththeprimary branchand thecontinuation of its primarysourceare of thesame
relativethicknessmeasured justabovethepointofbranching.
a. Size-determined midwaybetweentheleafapex and base as to theratioofveinwidth(vw)
to leafwidth(LW) vw/LWx 100% = Size.
1) Massive> 4% (Fig. 44).
2) Stout2-4% (Fig. 42).
3) Moderate1.25-2% (Fig. 43, 46, 47).
4) Weak < 1.25% (Fig. 59).
b. Course.
1) Straight(Fig. 46)-lacking noticeable curvatureor changein course.
a) Unbranched(Fig. 63)-lacking ramifications of primary rank.
b) Branched(Fig. 51)-with one or moreprimary ramifications.
2) Markedly curved(Fig. 42)-bent noticeably intoa smootharc.
3) Sinuous-repeated smoothchangesin direction of curvature (Fig. 65).
4) Zig-zag-repeated angularchangesin direction (Fig. 66).
2. Secondary Veins(20)-the nextsmallersize class of veins(whicharisefromtheprimaries)and
theirbranches(Fig. 63) of relatively
equal thickness
measuredjustabovethepointofbranching.
a. Angleofdivergence-measured betweenthebranchand thecontinuation of thesourcevein
abovethepointofbranching (Fig. 63).
1) acute-angleless than80?.
a) narrow< 45?.
b) moderate 45-65?.
c) wide65-80?.
2) rightangleor approximately so, 80-100?.
3) obtuse> 1000.
b. Variationsin angleof divergence.
1) divergence anglenearlyuniform (Fig. 63). The generalcase,departures fromwhichare
oftentaxonomically usefulfeatures.
2) uppersecondary veinsmoreobtusethanlower(Fig. 67).
3) uppersecondary veinsmoreacutethanlower(Fig. 68).
4) only lowest pairof secondary veinsmoreacutethanpairsaboveit (Fig. 69).
5) lower and upper secondary veins moreobtusethanmiddlesets(Fig. 70).
6) divergence anglemoreacuteon one sideoftheleafthanon theother(Fig. 47).
7) divergence anglevariesirregularly (Fig. 71).
c. Relativethicknessofsecondary veins-a measureof thewidthof themiddlesecondary veins
comparedto thoseof theprimary and tertiaryorders.Suchrelative estimatesofthicknessfor
thisand succeeding veinordersare essentially a measureof the proportional reductionin
widthfromone veinorderto thenext.Thisis a usefulcharacter onlyin cases ofmarkedde-
parturefromthewidthexpectedin theproportional reductionseries.
1) thick-proportionately widein relationto theprimary and tertiary ordersor to thesec-
ondariesin otherleavesof similarsize.
2) moderate-thegeneralcase.
3) finetohairlike-proportionately narrowin relationto theprimary and tertiary
veinorders
or to thesecondaries in otherleavesof similarsize.
d. Course-morethanone of thefollowing termsmayapply:
1) straight(Fig. 41)-withoutnoticeable deviationin course.
2) recurved (lowersecondaries in Fig.41 )-arching basallyfora portion ofitscourse.
3) curved-bending in an arc.
a) uniformly (Fig. 54)-arc smoothor gradually increasingin degreeofcurvature.
b) abruptly-sharp local increasein degreeof curvature (Fig. 46).
4) sinuous-repeated smoothchangesin direction ofcurvature (Fig. 71a).
January,1973] HICKEY-ARCHITECTURE OF DICOTYLEDONOUS LEAVES 29

TABLE 2. Analysisof tertiaryvein origin

Angle of 3? origin on the Exmedial (Lower) Side of the 2?'s

Acute Right Obtuse

Angle of 30 origin Acute AA (Fig. 63f) RA (Fig. 63i) OA (Fig. 631)

on the Admedial
Right AR (Fig. 63g) RR (Fig. 63j) OR (Fig. 63m)
the 20's Obtuse AO (Fig. 63h) RO (Fig. 63k) 00 (Fig. 63n)

5) zig-zag-repeated angularchangesin direction(Fig. 71b).

6) unbranched(Fig. 68)-without second orderramifications.
7) branched(Fig. 63)-with one or more secondaryramifications.
8) providedwithoutersecondaryveins (Fig. 41a)-a seriesof secondarybranchesarisingon
the exmedialside of thesecondaryvein.
e. Behaviorof loop-forming branches(if any).
1) joiningsuperadjacentsecondaryat acute angle (Fig. 63a).
2) joiningsuperadjacentsecondaryat rightangle (Fig. 63b).
3) joiningsuperadjacentsecondaryat obtuse angle (Fig. 63c).
4) enclosedby secondaryarches.30 or 4? arches (Fig. 63d).
5) formingan intramarginal vein (Fig. 64).
f. IntersecondaryVeins-thickness intermediate betweenthatof thesecond and thirdorderveins;
generallyoriginatingfromthe medial primaryvein, interspersed among the secondaryveins,
and havinga course parallel,or nearlyso, to them.
They are of two types:
1) simple-consistingof a singlevein segment(Fig. 63e).
2) composite-made up of coalesced tertiary vein segmentsforover 50 % of its length(Fig.
76).
g. Intramarginal Vein-a vein closelyparallelingthe leaf marginsand into whichthe secondary
veins are fused. Probablythe resultof thefusionand straightening of the exmedialbrochido-
dromoussecondaryarch segmentsintowhatappears to be an independentvein (Fig. 64).
h. IntercostalAreas-those portionsof the leaf blade lyingbetweenthesecondaryveins (Foster,
1950).
3. TertiaryVeins (30)-the next finestbranchesof thesecondaryveinsand thosebranchesof equal
thicknessfromthe primaries.
a. Angle of Origin-(defined above). When the predominantangle of tertiary originon the ex-
medial (lower) side of the secondaryveins is comparedwith that on the admedial (upper)
side of the secondaryveins,the combinationsshownon Table 2 are possible. This traitis of
diagnosticvalue. As a rule,in thosetertiaryveinswhichoriginateon the admedialside of the
secondaryveins and curveto join the primaryformingthe midvein,the angle of tertiary vein
originon the midveinequals the angle of tertiaryvein originon the exmedial side of the
secondaryveins of the leaf. Departurefromthisrule is a taxonomicallysignificantfeature.
b. Pattern.
1) Ramified-tertiaryveinsbranchingintohigherorderswithoutrejoiningthesecondaryveins
(althoughtheirhigherorderderivativesmay do so).
a) exmedial-branchingorientedtowardthe margin,rare (Fig. 78).
b) admedial-branchingorientedtowardtheleafaxis (Fig. 75). [Not obmedialas in Fig.]
c) transverse-branching orientedacross intercostalarea-the commonestcase (Fig. 73).
2) Reticulate-tertiaryveins anastomosingwith other tertiaryveins or with the secondary
veins.
a) randomreticulate-anglesof anastomosesvary (Fig. 77).
b) orthogonalreticulate-angles of anastomosespredominantly
rightangles (Fig. 79).
3) Percurrent-tertiaries
fromthe oppositesecondariesjoining.
a) course.
30 AMERICAN JOURNAL OF BOTANY [Vol.60

AREOLE
891)] | DEVELOPMENT
89 90l
LOOPED

LOOPED INCOMPLETE 104

FIMBRIAL VEIN
91 LOOPED \ I MM
93 LACKING

94

~~O MPLECOPLTE4W
92
105
~~~~~~~~~IMBRIATE
MARGINALULTIMATE VENATION TYPES
\
AREOLATIONSHAPE
4/ I~~~~~~~~~~~~~~~~~~MM
96 97 INCOMPLETE
5 + SIDES a
95 POLYGONAL 106 107
PENTAGONAL QUADRANGULAR TRIANGULAR

VEINLETS SIMPLE BRANCHED

B RA A
NOE
NONE 'LINEAR CURVED I I
'ONCE TWICE 3 TIMES

_ _ _ _ _ _ _ _ __
WELL
I IMPERFECT
98 99 100 101 102 103 DEVELOPED

features(continued)-ultimatevenationand areolation.
Fig. 88-107. Leaf architectural

(1) simple-unbranched(Fig. 63).

(2) forked-givingrise to thirdorderramifications (Fig. 76).
(3) straight-passingacross theintercostalarea withouta noticeablechangein course
(Fig. 63h, 1).
(4) convex-middle portionof thevein curvingaway fromthecenterof theleaf (Fig.
63f).
(5) concave-middle portionof the vein curvingtowardthe centerof the leaf (Fig.
63n).
(6) sinuous-repeatedlychangingdirectionof curvature(Fig. 74).
(7) retroflexed-forming a singleS shaped curveconcave apicallyand convexbasally
(Fig. 63o).
(8) recurved-curving inward frompoint of originon the adaxial side of a sec-
ondaryvein to terminateon the midveinof the leaf (Fig. 63).
b) relationshipto midvein(Fig. 82).
(1) approximately at rightangles (Fig. 83).
(2) longitudinal-approximatelyparallel (Fig. 84).
(3) oblique-trendingin an obtuseor, rarely,an acute angle to the midvein.
(a) tertiaryangle withmidveinremainingapproximately constant(Fig. 85).
(b) angle decreasingexmedially(Fig. 86).
(c) angle decreasingapically (Fig. 87).
January,
1973] HICKEY-ARCHITECTURE OF DICOTYLEDONOUS LEAVES 31

c) arrangement.
(1) predominantlyalternate-joiningeach otherwithan offseti.e., an abruptangular
discontinuitysuch as a branch (Fig. 76, 81).
(2) predominantly opposite-joining each othersmoothlyi.e., in a straightor curved
path (Fig. 63).
(3) alternateand oppositein about equal proportions.
(4) distant-intervalbetweenveins0.5 cm or greater.
(5) close-interval betweenveins less than 0.5 cm.
4. Higherordervenation-the nextfinerorderof veins originating fromthe tertiariesand those of
equal size fromlower orderveins is knownas the quaternary(40) venationand the veins origi-
natingfromtheseand thoseof equal size fromlowerordersare thequinternaries (50), etc.
a. Resolution.
1) Higherordervenationforminga reticulumin whichvein orderscannot be distinguished
(Fig. 80).
2) Vein ordersdistinct(Fig. 72, 81).
b. Quaternaryveins.
1) Size-a relativemeasure of the widthof the quaternaryveins comparedto those of the
thirdand fifthorders. Such relativeestimatesof thicknessforthisand thefifthorderveins
(below) are essentiallya measure of the proportionalreductionin widthfromone vein
orderto the next. Notable onlyis any markeddeparturefromthe widthexpectedfor the
fourthorder (fifthorder,below) veins as partof a proportionalreductionseries.
a) thick-wider thanexpected.
b) thin-narrowerthanexpected.
2) Course.
a) relativelyrandomlyoriented(Fig. 72).
b) orthogonal-arisingat rightangles (Fig. 81). Their subsequentcourses may or may
not be at rightangles.
c. Quinternary veins [analyzedas in b 1 above].
1) Size (as above).
a) thick.
b) thin.
2) Course [analyzedas in b 2 above].
a) random (Fig. 72) (as above).
b) orthogonal(Fig. 81) (as above).
d. Highestvein orderof leaf: 3?, 40, 50, 60, 70.
e. Highestvein ordershowingexcurrentbranching: 20, 30, 40, 50, 60.
f. Marginalultimatevenation.
1) Incomplete-freelyendingveinletsdirectlyadjacentto the margin(Fig. 90, 92).
2) Looped-the majorportionof themarginalultimatevenationrecurvedto formloops (Fig.
88, 89, 91).
3) Fimbriate-highervein ordersfusedintoa vein runningjust insideof themargin(fimbrial
vein) (Fig. 93, 94).
5. Veinlets-the freelyendingultimateveinsof theleafand veinsof thesame orderwhichoccasionally
cross areoles (see below) to become connecteddistally.
a. Veinletsnone (Fig. 98).
b. Simple-withoutbranches.
1) Linear (Fig. 99).
2) Curved (Fig. 100).
c. Branched-givingrise to ramifications by dichotomizing.
1) Once (Fig. 101).
2) Twice (Fig. 102).
3) Three times (Fig. 103), etc.
32 AMERICAN JOURNAL OF BOTANY [Vol.60

6. Areoles-the smallestareas of the leaf tissue surroundedby veins which taken togetherforma
contiguousfieldover most of the area of the leaf. Thus, smallerareas occasionallyformedwhen
veinletscrosstheirareoles are excluded. Any orderof venationin a leaf fromthe primaryto the
highestorderbelow that of the freelyendingveinletscan formone or more sides of an areole.
However onlythe orderrepresentedby the veinletswill intrude,or occasionallycross, the islets
formedby thenon-freely endingveins. The appearanceand characteristics of theareolesare termed
areolation.
a. Development.
1) Well developed-meshes of relativelyconsistentsize and shape (Fig. 107).
2) Imperfect-meshesof irregularshape, more or less variablein size (Fig. 106).
3) Incompletelyclosed meshes-one or moresides of themeshnot boundedby a vein,giving
rise to anomalouslylarge meshesof highlyirregularshape (Fig. 105).
4) Areolationlacking-as in hyphodromous or succulentleaves. Very rarely(Fig. 104) vena-
tion simplyramifying into the intercostal
spaces withno coherentshape,size or patternto
the areas surroundedby veins.
b. Arrangement.
1) Random-areoles showingno preferredorientation(Fig. 106).
2) Oriented-areoles havinga similaralignmentor patternof arrangement withinparticular
blocks or domains (Fig. 107).
c. Shape.
1) Triangular(Fig. 97).
2) Quadrangular(Fig. 96).
3) Pentagonal(Fig. 95).
4) Polygonal-with more than 5 sides.
5) Rounded.
6) Irregular.
d. Size-the size classes of areoles chosen appear to represent,
at least partially,naturallyoccur-
ringgroups.
1) Very large > 2 mm.
2) Large 2-1 mm.
3) Medium 1-0.3 mm.
4) Small < 0.3 mm.

LITERATURE CITED theleafin QuiinaacutangulaDucke. Amer.J. Bot.

37: 159-171.
BERRY, E. W. 1916. The lower Eocene floras of south-
. 1952. Foliar venationin angiospermsfrom
eastern North America. U.S. Geol. Surv. Prof.
an ontogenetic viewpoint. Amer.J. Bot. 39: 752-
Paper 91.
766.
DELEVORYAS,T., ANDR. E. GOULD. 1971. An unusual
GOEBEL,K. 1905. Organography of plants. Eng. ed.
fossil fructificationfrom the Jurassic of Oaxaca,
by I. B. Balfour. PartII. Oxford.
Mexico. Amer. J. Bot. 58: 616-620.
HICKEY,L. J. 1971a. Leaf architectural classification
DEVADAS, C., AND C. B. BECK. 1971. Development
of the Angiosperms(abstr.). Amer. J. Bot. 58:
and morphology of stelar components in the stems
450.
of some members of the Leguminosae and Rosaceae.
Amer. J. Bot. 58: 432-446. 1971b. Evolutionary significance
ofleafarchi-
DILCHER, D. L., ANDG. E. DOLPH. 1970. Fossil leaves tectural featuresin the woody dicots (abstr.).
of Dendropanax from Eocene sediments of south- Amer.J. Bot. 58: 469.
eastern North America. Amer. J. Bot. 57: 153- In press. Stratigraphy and Paleobotanyof the
162. Golden Valley Formation (Early Tertiary) of
DOYLE, J. A. 1969. Cretaceous angiosperm pollen of westernNorthDakota: Geol. Soc. Amer. Memoir.
the Atlantic Coastal Plain and its evolutionary sig- HOLLICK, A. 1936. The Tertiaryfloras of Alaska.
nificance. J. Arnold Arboretum 50: 1-35. U.S. Geol. Surv. Prof.Paper 182.
ETTINGSHAUSEN,C. VON. 1861. Die Blattskelete des KERNER VON MARILAUN, A. J. 1895. The naturalhis-
Dicotyledonen. K. K. Hof. Staatsdruckeri,Vienna. toryof plants. Tr. and ed. by F. W. Oliver. Vol. 1.
FEDEROV, A. A., M. E. KIRPICHNIKOV, AND Z. T. ARTU- Holt, New York.
SHENKO. 1956. Atlas po opisatelnoi morfologii KRUSSMANN, G. 1960. Handbuch der Laubgeholze.
visshich rastenii. Akad. Nauk, USSR. Moscow, Paul Parey,Berlinand Hamburg.
Leningrad. LAM, H. J. 1925. The Sapotaceae, Sarcospermaceae,
FOSTER,A. S. 1950. Morphology and leaf venation of and Boerlagellaceaeof the Dutch East Indies and
January, 1973] HICKEY-ARCHITECTURE OF DICOTYLEDONOUS LEAVES 33

surrounding countries. Bull. Jard.Bot. Buitenzorg TERMINOLOGY (SADT). 1962. Terminologyof

III, 8: 1-289. simplesymmetrical plane shapes (Chart 1). Taxon
LAWRENCE, G. H. 1951. Taxonomyof vascularplants. 11: 145-156; 245-247.
Macmillan,New York. TAKHTAJAN, A. L. 1963. In A. L. Takhtajan,V. A.
LEE, A. T. 1948. The genusSwainsona.Contrib.New Vakhremeev,and G. P. Radchenko[ed.], Osnovy
South Wales Herb. 1: 131-271. paleontologii Golosemennyei pokrytosemennye.
LESQUEREUX, L. 1878. Contributionsto the fossil Moskow.p. 395-399.
floras of the WesternTerritories.Part II. The TROLL, W. 1938. VergleichendeMorphologie der
TertiaryFlora. U.S. Geol. Geog. Surv.Terr.Rept. hohem Pflanzen. 1 (2). Borntraeger, Berlin. Re-
7. printed,1967, Koeltz, Koenigstein-Taunus.
WOLFE, J. A. 1966. Tertiaryplants fromthe Cook
PACLTOVA, B. 1961. Zur Frage der gattungEucalyptus
InletRegion,Alaska. U.S. Geol. Surv.Prof.Paper
in der bbhmischenDreideformation.Preslia 33: 398-B.
113-129. 1968. Paleogenebiostratigraphyof nonmarine
STEARN, W. T. 1956. In P. Synge [ed.], Supplement rocksin King Co., Washington.U.S. Geol. Surv.
to the Dictionaryof Gardening.Oxford. p. 318- Prof. Paper 571.
322. . 1969. Paleogene floras from the Gulf of
SYSTEMATICS ASSOCIATION COMMITTEE FOR DESCRIPTIVE Alaska Region. U.S. Geol. Surv. Open File Rept.