You are on page 1of 22


Early  Journal  Content  on  JSTOR,  Free  to  Anyone  in  the  World  
This  article  is  one  of  nearly  500,000  scholarly  works  digitized  and  made  freely  available  to  everyone  in  
the  world  by  JSTOR.    

Known  as  the  Early  Journal  Content,  this  set  of  works  include  research  articles,  news,  letters,  and  other  
writings  published  in  more  than  200  of  the  oldest  leading  academic  journals.  The  works  date  from  the  
mid-­‐seventeenth  to  the  early  twentieth  centuries.    

 We  encourage  people  to  read  and  share  the  Early  Journal  Content  openly  and  to  tell  others  that  this  
resource  exists.    People  may  post  this  content  online  or  redistribute  in  any  way  for  non-­‐commercial  

Read  more  about  Early  Journal  Content  at­‐jstor/individuals/early-­‐


JSTOR  is  a  digital  library  of  academic  journals,  books,  and  primary  source  objects.  JSTOR  helps  people  
discover,  use,  and  build  upon  a  wide  range  of  content  through  a  powerful  research  and  teaching  
platform,  and  preserves  this  content  for  future  generations.  JSTOR  is  part  of  ITHAKA,  a  not-­‐for-­‐profit  
organization  that  also  includes  Ithaka  S+R  and  Portico.  For  more  information  about  JSTOR,  please  


ESPECIAL interestis attachedto the studyof the cactus family

because it is peculiarlywell adapted by habit and by structure
to withstandthe tryingconditionsof the desert. The greatest
developmentof the group occurs in the arid portionsof tropical
and subtropicalAmerica,but the conceptionwhichthisstatement
is likelyto give thatthe cacti thrivebest witha modicumof water
and live in localitiesthat are too severeforall otherdesertplants
to endure,is erroneous. The cacti like otherplants of the desert
are most vigorouswhen the water supply is adequate, and it is
by no means certainthat such a formas the giant cactus (Cereus
giganlteus)or the barrel cactus (Echinocactuswislizeni), as well
as the larger Opuntias, do not requirea larger amount of water
thanmanyof thelargeplantsof otherfamilies.
However, it is also likely that no desert plants can live and
perpetuatetheirkind under more arid conditionsthan some of
the cacti. The generalmeans by whichtheyaccomplishthis are
too well knownto requirerepetitionhere,but certainadaptations
to desertconditions,not so well knownperhaps,may be pointed
out. The most importantfactorin the life relationsof the des-
ert plants is unquestionablythe available water supply,and the
most strikingadaptations accordinglyare associated with the
absorption,the storage,or the conservationof water. This rela-
tion to the water supply, eitherin apparent independenceof it
or in intimateassociationwith it, is met at each stage of devel-
opment. For instance,the seeds of the giant cactus will ger-
minate in summerwhile lying on the top of air-drysand and
withoutprevious wetting. Seedlings of Oppuntiaversicolorare
provided with water-storageorgans (Fig. 1) although such are
absent in the adult plants. The reaction of the mature plants
to a variable water supply is also noteworthy.Specimens of
1 Papers fromthe Desert Botanical Laboratory of the Carnegie Institution,
No. 11.

Opuntia engelmanni,wrinkledfromthe loss of water during a

long drought(Fig. 2), absorbed sufficientwater withintwo days
followinga storm to make their joints plump and smooth (Fig.
3). The giant cactus is especially adapted by the peculiar for-


FIG. 1.- Young plants of Opuntia versicolorshowing the water-storageorgans -

the swollen roots - which are not present as such in the mature form.

mationof the rindto undergowithoutinjuryconsiderablechanges

in volume which are induced by a variationin the watercontent
(E. S. Spalding, : 05). Other adaptations,such as certainchar-
acteristicsof the root systems,certain peculiaritiesof structure
and their relation to transpiration,and the transpirationunder
differentconditions may be presentedsomewhat more fully.
b i

FIG. 2.-Opun~tiaengelmanni,
May 11. May 11 markedthecloseofa severedry
periodand thewrinkled
surfaceofthecactusshowsthatit was suffering
the drought.


A special studyof the structureand extensionof the root sys-

tems of desertplants cannot fail to be of great interestand im-
portancein contributing to a rightunderstandingof the biology
of these plants. Many characteristicsof the desert vegetation
are withoutmuch question directlytraceable to peculiaritiesof
the various root systems. It has been observed (Coville, '93,
p. 43) that the fairlyequal spacing of desert shrubs is one of
the characteristicsof theirdistribution. The primarycause for
this is presumablythe strugglefor water and their distribution
is, therefore,an expressionof the mutual relationshipof the root
systems. Again, frequentlythe formof the root is incompatible
with certainhabitats,- for example a subirrigatedplant would
finddifficultyin growingwherethe subsoilis therock-likecalliche,
- and plants with such deeply penetratingroots, for instance,
are forthisreasonlimitedin theirdistribution. It is theoretically
possible, and so far as I have observedactuallytrue,that those
plants that have a root systemwhich is at once superficialand
whichpenetratesthe grounddeeply,all otherthingsbeing equal,
may also have the widest choice of habitats. Certain it is that
the creosotebush, for example, which has a root systemof this
character(V. M. Spalding, 04) is perhaps the most widelydis-
tributedof our desertshrubs. Althoughthis view of the relation
of the characterof the root systemsto the distributionof these
plants is advanced tentativelyonly, the importanceof it as a
factorwhich must be taken into considerationin this connection
and sometimecarefullystudied,is veryapparent.
The root systemof a specimenof Echinocacts wislizeni which
was 60 cm. highand 35 cm. in diameter,growingabout 75 meters
northof the laboratory,was carefullyexposed and the course of
its roots mapped (Fig. 4). The roots, as the figureindicates,
were branchedveryfreely. There were threemain roots which
arose fromthe base of the plant not farfrom10 cm. fromthe sur-
face of the ground and which so directedtheirgrowth,and that
of the branches,that the area compassed by them was about
equally apportionedand well covered. As a rule the rootswere
FIG. 3.- Opurdiaengelmanni, May 14. This is the plant shown in Fig. 2. Rains
came May 10-12; the plump condition of the joints of the plant on May 14
indicates that water was absorbed promptlyafter the rains, and in consider-
able quantity.

slender. At a distanceof 15 cm. fromthe plant one of thelargest

of themwas 7.6 mm. in diameter,and one meterfromthe plant
it was 4.6 mm. in diameter. The roots ran about 6 cm. below
the surface,in places whichwerefreeof stones,but when a stone
was encounteredthe rootdipped beneath it and availed itselfof

0+E1 ~~~~f KX
_~~~~~~~~~~~~~~~~~~~~ _--1___,-A
> h_T-

Adt~~~~~~~~~~~~ A~t

_, ,,

- 1 T _ 11

FIG. 4.- Root system.ofEchinocactus wislizeni. Scale: 1 unit = 30 cm.

the better water supply to be found there. The most deeply

placed root,however,was not morethan 10 cm. below the surface
of the ground. There are thereforetwo noticeablecharacteristics
of the root systemof Echinocactus wislizeni,namely, the roots

FIG. 5.- Root system of Cereus giganttus. Scale: 1 unit = 10 cm.


are slender throughouttheir entirecourse and they are super-

The rootsof Cereusgiganteus,on the otherhand, in formand
position,and perhaps in extentand branchingalso, are verydif-
ferentfromthose of Echinocactus. Fig. 5 representsthe root
systemof a Cereus giganteus,about one meterhigh, which was
growing200 meterswestoftheEchinocactusjust described. Four
main rootswere observedto arise fromthe base of the plant. At
firsttheywere relativelyheavy, from2 to 4 cm. in diameterat
the proximalends, but they became smaller very rapidlyas the
distance from the plant increased-in a manner much as is
indicatedby A of the figure. Very soon afterleavingthe plant
the rootsbranched. One branch,whoselaterhistorycould notbe
traced,struckdirectlydownwards,and the othertook a more or
less horizontalcourse. The latterbranchedat intervals,although
perhapsnot so frequentlyas those of Echinocactus,and extended,
in one instanceat least, over one meter fromthe plant's base.
How much fartherthe root reached could not be learned because
and the small size of the distal branches. The
of its fragility
superficialportion of the root systemof Cereus giganteuswas
more deeply placed than were the roots of Echinocactus, and
owing to the fact that these parts were not so richlybranched,
the ground included by them was not so thoroughlycovered.
However,in one characteristic,which is of interestto note but
whose significance I have not investigated,the superficialrootsof
the two formsare alike,namely,the longerrootsand the greatest
numberof rootsare situatedon the uphill side of the respective
plants. This peculiarityis shown in the two figures. In Fig. 4
the uphillside is to the right,and in Fig. 5 it is at the top of the


Perhaps the correlationof structureand functionis nowhere

more patent than in the peculiar transpiration-controlling and
transpiration-promoting tissues,and the rate of transpirationin
certaincacti. As is well known,the cacti are well adapted struc-
turally,laying aside forthe momentthe matterof water storage,

to retainwaterforlong periods. As an illustrationof how long

a specimen of Ccrcu.
giganteuwmustordinary I(
lv retain water in this
locality, the following
may be suggested. Cal-
culationsbaseduponthe 'M..Y
tentofthese cacti,on the (
estimatedspread of the 4

rootsystem,and on the (
average rainfall at this
place indicate that ap- /

proximately two years'

rainfall, assuming the
rainfall to be normal
each .vear, 11.74 in., are
* X X FIG. 6.- Echinocactus wislizeni. A portion of the
required to supply a cortex showing chliorenchymawith large inter-
cellular spaces. The section was made parallel
cactus 15 feethigh with to the surface.
an amount of water
equal to what it usuallycontains. The necessityof husbanding
thewater so hard-gainedis even
<\WS >; moreapparentwhenthe relation
zj-I~K<I Ad(XXof thenormalrainfall
\ 9/1b evaporationis takenintoaccount.
</tv-M /J I I Ashas been pointedout (Coville
and MacDougal, :03, p. 27) the
annual evaporationis 7.7 times
the rainfall. This is thenormal
s T > (K ratio,butwhentheprecipitation
I 1I j is below average, as in 1904
(when it was 75 percentnormal)
thedisparityis evengreater(Can-
56tJ? non,:05b).
FIG. 7. - Echinocactus wiszeni. The
Although,as mentionedabove,
heavy-walled supporting tissue lying thestructural
immediately under the epidermis. retention
This section was parallel to the surface.
referenceshould here be made
to the special adaptationsfound in Echinocacduw

ticularlyon account of the small rate of transpirationwhich

was demonstratedin this plant. Echinocactushas a heavyouter
membranewhichis cuticularized
K<g As (Figs. 9, 10). The stomataare
peculiarlyfittedto guard well a
too rapid loss of water,and they
Jt 0 \have a somewhat complicated
structure. The stomaas charac-
teristicof manyxerophytic plants,
is sunkenbelow the general sur-
face of the stem,and fromit in
Echinocactusa tube,divisiblein-
to two p
to portionsofsseparate
te fune-
FIG. 8.- Echinocactus wislizeni. Surface led eens io outer
view ofof
view stoma.
stoma.Xi Intotheouter
tion,leads dleeply
portionofthe cortexandbecomes
the peripheralportion of the extensive intercellularaerating sys-
tem of the plant. The substomaltube, reallytrachea,is shown
schematicallyin Fig. 11. The outer portion,which is heavily
shaded in the sketch,is cuticu-
larized throughoutits course in
the sclerenchymatous tissue(Fig.
10) of thecortex,and in thispart
it acts merelyas a tube for the
conduction of gases. Beneath
the supportingtissue the tube
entersthe chlorenchvmaand its
wall is no longercuticularized;it
here functionsas the substomal
It is of interest to contrast
with this permanentstructurea FIa. 9.- Echinocacdus cwqizeni. Sto-
formofstomaandsortofadjoin- mata and substomal canal which passes
through the supporting tissue (see
ing tissue which are a part of Fig. 7) to tte moredeeply placed chlor-
evanescentorgans,and whichalso enchyma (Fig. 6).

appear to have somewhatdifferent functions. I have reference

to the stomataof the leaves of Opunlia versicolor.The perma-
nent stomataof this Opuntia closelyresemblethose of Echino-
cactus but the temporarystomata,those of the leaves, are very
(Figs. 12-14). These stomata are superficially

FIG. 10.-- Echinncatuswislizeni. Same as Fig. 9, to show character of supporting


and theyopen directlyintothe substomal

chamber(Fig. 12). The substomalcanal
of the permanent organ, thereforeis lack-
ing. Associated with this form of stoma is - --- -

theabsenceof a sclerenchvmatous support- . -

ing,tissue,and, consequently,the exten--

sionofthechlorenchvma to theepidermis.
The heavy outer epidermalwall of the
older portions of Echinocactus is here
replaced by a delicate one. In connec-
tion with thisstructureof the leaves of
Opuntia appears their functionof pro- FIG. 11-- Erhinocatus wus-
lizeni. Semidiagrammatic
voting transpiration,and presumably sketch showing the rela-
tions of the substornal
the respiratoryactivitiesas well, and in canal. That portion of it
render it
it an
an important service,
important servtce,
which is in the supporting
tissue is heavily shaded;
as will be apparentfromthe resultsof that portionin the chloren-
chyma is represented by a
studies. ihe al


In all of the studies-on the transpirationof cacti whichwere

conductedoutdoorsthepolymeter method(Cannon :05a), described
elsewhere,was employed. A few,however,whichwillbe pointed

out later,were done in the laboratoryby weighingin a manner

FIG. 12.- Opuntia versicolor. Cross section of a leaf showing two stomata, one
of which was cut in two at rightangles to the guard cells and the otherparallel
to and to one side of them. The confluent substomal chamber is shown.
This section is to be contrasted with Fig. 13.

to be described. This change in methodwas made necessary

on accountof the highrelativehumidityprevailingat the time.


FIG. 13.- Opuntia 'ersicolor. Substomal canal and partly developed

supportingtissue of a young stem.
ExperimentNo. 1. Cereus giganteus
Percent of Amount in
Time Saturation Temperature Milligrams
9: 27 A. M. 42.0 940 F. 67.5'
10:37 A. M. 48.5 1050 F. 106.0

The cactuswas located in the shade of a palo verde(Parkin-

sonia microphylla)on a dry mountain-sidenot far above the bed
of Salvino Canyon, 18 miles east of the Laboratory. Higher

FIG. 14.-Opuntia versicolor. Cross section of a leaf to illustrateits delicate struc-


on the same slope were many othergiant cacti of large size, and
other typical desert plants such as Encelia farinosa,Fouquieria
splendens,as well as otherspecies of cacti.

1 The firstamount in each case is the absolute humidityof the atmosphere

of the bell glass when the experiment begins. The second amount is the
absolute humidity at the close of the experiment. The differencebetween
the two is the amount transpired.

The transpirationof the cactus was taken September9, and

as the high relativehumidityat the beginningof the experiment
shows,the effectsof the rains of Augustwere stillmanifest. The
rate which is 0.2 milligramsper minutefor 100 sq. cm. of tran-
spiringsurface,may be considereda high one, since withoutex-
ceptionthe greatestrate of transpirationof all the plants,whose
seasonal variationin rate has been observed,has been afteror
at the time of the summerrains.

Experiment2. Echinocactuswislizeni
Percent of Amount in
Time Saturation Temperature Milligrams
10: 28 A. M. 32.5 820 F. 36.0
2: 05 P. M. 35.5 910 F. 51.0

This experimenttook place on March 19, at the Desert Botan-

ical Laboratory.
The rate of transpirationof the entireplant per hour is 3.4
On September3 the experimentwas repeated when the fol-
lowingdata were derived:-

Experiment3. Echinocactuswislizeni
Percent of Amount in
Time Saturation Temperature Milligrams
12: 24 P. M. 35 1040 F. 60
3:14 P. M. 45 1090 F. 89

The rate forthe entireplant is 9.6 milligramsper hour.

This specimenof Echinocactus is growingon a westerlyslope
on the Laboratory Mountain. In its vicinityare found a few
giant cacti and Encelia farinosa, Lycium sp., and Parkinsonia
microphylla. Because of the desirabilityof preservingthis speci-
men the surfacewas not computedand therefore the rate cannot
be compareddirectlywiththat of Cereusgiganteusas givenin the
precedingexperiment. However,it happened that the two cacti
were of nearly the same size,- 10 cm. in height,- and a gen-
eral comparisonbetween the two can be made. The rate of

Cereus per hour was approximately33 milligramswhile that of

Echinocactus on September 3 was 9.6 milligrams. Whether
this differencein the rate is constantfor the two genera, or is
attributableto other and unknowncauses aside fromthe rela-
tivelyslightdifferencein surface,is not known.

TRANSPIRATION OF Opuntia versicolor

A specimenof Opuntia versicolorabout 20 cm. high, growing

near the laboratorybuilding,was studiedat various timesduring
the dry portionsof the year (1904), namely, in March, April,
June, and July. The observationsindicate that the reactionof
Opuntia to wateris verydifferent fromthatof otherdesertplants,
such forinstanceas Covillea tridetlata,Encelia farinosa,or Foa-
quieria splentdens(Cannon, :05a, p. 404), suggestinga unique
positionamong its associates and an importantfactor among
the varied ones that broughtabout the presentdistribution of the
groupto which it belongs.
As has been shown in anotherplace, the transpiration of Fou-
quieria splenden8,as well as thatof otherdesertplants and plants
of more humid regions (Burgerstein,: 04), under certain con-
ditionsincreaseswith an addition to the available water supply
(see Cannon, :05b; V. M\I.Spalding, :04, :05). Thus after
rains, but before leaves appeared, the rate of transpirationof
Fouquieria splendens increased about three fold; after leaves
had been formedand while they were developingthe rate was
relativelyverygreat. A similarconditionwas likewiseobserved
in Encelia farinosa,Covillea tridentata,and in other plants. In
Opuntia versicolor,however (see Figs. 2, 3), and probably in
other Opuntias the responseto the rains is indeed also positive
but in a verydifferent way. The cactus absorbs water greedily,
and as a consequenceit at once increasesin size, and its tissues
become turgid. But, so far as I observed,the rate of transpira-
tion did not increase proportionally.Indeed, laboratoryexperi-
ments,in which a small specimenof 0. versicolorwas attached
to a potometerby a long delicate tube so that the cactus could
be weighedat intervalsat the same time that its rate of absorp-
tion was being recorded,showed very clearly that under such

conditions thecactusmayabsorbwatermuchfasterthanit gives

it up by transpiration.
It shouldbe notedthatthe specimensof cactuswhichwere
experimented upon bothin the fieldand in the laboratory did
nothave an adequatewatersupplyat theirdisposalpreviousto
thetimesof theexperiments.As a generalthingnotuntilsome
timehas passed afterthe waterhas been absorbed,does new
growthappearwithits embryonic structureand its evanescent
leavesand thenonlydoestherateoftranspiration becomegreatly
increased. Duringthe periodsof droughtthe plantsmake but
littlenew tissue. These peculiarities
of Opuntiaversicolor were
observedrepeatedlyand will be presentedin the succeeding
resumeof representative experiments.
The transpiration of Opuntiaversicolor,-an entireplaint,
was as follows:-
March 25 . . . 51.0 milligrams in one hour
March26 . . . . 63.01 "
April 25 . . . . 19.9 "
June30 . . . . 27.5 "
July4 . . . 26.1 " "
DuringtheperiodfromMarchto Julytherainfallwas unusu-
ally small and the cactushad an insufficient supplyof water.
The ratesof March,April,and June,therefore, represent the
transpiratory activitiesof the plant in timesof drought.At
varioustimesin themidstof thedryseasonsFouquieria,Covil-
lea, and otherplantshad been irrigated and theeffects on their
transpirationwererecorded(Cannon,:05b). To learn how an
increasein thewatersupplyofOpuntiaversicolor wouldinfluence
itsrate,as wellas tolearnhowtherateundersuchcircumstances
wouldcomparewiththatduringdryconditions, it,also,was irri-
gated. On June27, whichwas a timeof drought, ninegallons
of waterwerepouredslowlyon the groundat the base of the
cactus,but it did notshowby an acceleratedrate (see the rate
of July4, above)thatit had absorbedanyofthewater. That it
had reallydone so, however, was indicatedby the factthatthe
planthad becomerigidbytheincreasedturgescence ofitstissues.

'Unfortunately a small branch was broken fromthe plant afterthis experi-

ment so that the winterand the summer rates are not comparable.
The transpiringsurface of the Opuntia was not estimated,
so thatits rate cannotbe compareddirectlywiththe rate of other
cacti or plantsof otherfamiliesalthoughthis,perhaps,is of minor
consequence. The importantfact was establishedthat the plant
does transpiremeasurable amounts of water even in the driest
timesand that it absorbs waterquite out of proportionto its rate
of transpiration.
I wish now to call attentionto a phase of the biologyof Opun-
tia versicolorwhich is also of great importancein the economy
of the plant but which has hithertoreceived little emphasis,
namely,to the rolewhichthe leaves play in transpiration.
On August18 the polymeterapparatus (Cannon, :05a, Fig. 4)
was adjusted to take the transpiration of a branchof the cactus
whichbore leaves and whichwas situated a fewmetersnorthof
the laboratorybuilding. The data derivedfromthis experiment
are as follows:-

Experiment4. versicolor
TranspirationofLeaves of Opurntia

Percent of Amount in
Time Saturation Temperature Milligrams
2:20 P*. 40 950 F. 62.
2:30 P. mI. 59 980 F. 101.

rThe branch transpiredat the rate of 234 milligramsin one

hour, or 0.91 milligramsper minutefor 100 sq. cm. of transpir-
ing surface.
As soon as the experimentwas finishedthe surfaceto the stem
was coated withvaselineand the experimentwas repeated. The
therefore,is thetranspiration
oftheleaves only.

Percent of Amount in.

Time Saturation Temperature Milligramis
2:42 P. M. 41.5 970 F. 69.
2: 52 P. mI. 52.0 100? F. 87.

The rate per hour for the leaves of the branch is 108 milli-
grams,or 0.42 milligramsa minutefor 100 sq. cm. of surface.
The surfaceof the stem alone was estimatedat 331 sq. cm.;
that of the leaves at 97 sq. cm. Thereforewith somewhatless

than one fourththe entiretranspiringsurface,the leaves alone

transpirednearlyone half the whole amount.
The high humidityat this timewas unfavorableto the further
use of the polymetermethod so that the experimentsupon the
transpirationof the leaves of the cactus were continuedwith a
special weighingapparatus in its stead. Since the resultsof all
of theseexperimentswere essentiallyalike, I shall referto one of
A branch of Opuntia versicolorwith leaves was placed in a
bottlecontainingwater which was so arranged,with a capillary
tube as well as the branch fastenedin the stopper,that the air
could enterand maintaina pressurewithinuniformwith that of
the room,while only an inappreciablequantityof vapor escaped.
In one hour,2: 15 to 3: 15 P. M., the branchlost 180 milligrams
in weight. The stein was then coated with vaseline and in one
hour,3: 30 to 4: 30 P. M., the loss of weightwas 100 inilligrams,
whichwas, of course,the transpirationof the leaves only.
There were 69 leaves on the branch whose entiresurfacewas
estimatedat 55 sq. cm. The surfaceof the stem alone was 65
sq. cm. Thereforethe leaves had about 45 percentof the entire
transpiringsurfaceand they gave off about 55 percentof the


The leading points in this paper and the conclusionsmay be

brieflystatedin thefollowingsummary.
1. The root systemsof Cereus giganteusand of Echinocactus
wislizeniwhich were studied and mapped, presentcharacteristic
differences.The root system of Cereus is in part superficial
and in part deeply placed. The root systemof Echinocactus is
superficialonly. There appears to be a relationbetweenthechar-
acterof the rootsystemsof theseplants and that of the habitats
in which theynaturallyoccur. For example, the formand the
extensionof the rootsof Cereus inhibitits occurrencein localities
wherethe underlying formationis of such naturethat theycannot
reach the usual or needfuldepth. We accordinglyfindthe plant
on rockymountains,or wherethe soil is deep, but in thislocality

it does not grow at all, or rarely,on the mesa wherethe rock-like

calliche forms a thick and nearly impenetrablestratumwhich
reaches almost to the surface. However, it may not be wholly
a problemof anchorage,since the morphologicalconditionmay
be associated witha physiologicalone, as for instance stubirriga-
tion or proper drainage which may be indispensablefactorsin
its water relations. Althoughthe characterof the root system
may thus be closelyconnectedwith the characterof the habitat,
certainfeaturesin the local distributionindicate that it cannot
be too narrowly insisted upon. For example, Cereus qigcynfteus
avoids northernslopes, althoughto all outward appearances the
structureand the water supplymay be quite the same as on the
Echinocactus presents quite a differentcondition of affairs.
The plant does not require unusual protectionagainst lateral
stresses. It grows most abundantlyin this localityon the mesa
wherethe soil is shallow. The rootsare so placed that theycan
neitheraffordsafe anchorage for a tall plant, nor absorb water
at the water level. There is thereforea directrelationbetween
the characterof the plant and that of the root system,on the
one hand, and the characterof the root systemand that of the
habitat, on the other. It should also be noted that the roots
of Echinocactus, which are very shallowly placed, permit the
plant to derive benefitfromrelativelysmall rains, but, by the
same token,that they preventit fromgettingwater other than
what fallson the area includedby them.
2. The strikingdisproportionbetween absorptionand tran-
spiration,which was observed in Opuntia versicolor,is thought
to be of great importancein accountingfor the distributionof
the plant (and perhaps of the family)in those parts whereevap-
3. A low rate of transpirationwas demonstratedin Optnfia
versicolorand Echinocactuswvislizeniduringperiodsof prolonged
drought. At the time of the summerrains the rate was greatly
increased and in all instances the increase was associated with
the renewal of growth.
4. A directrelationwas observedbetweenstructture and tran-
spiration. The matureportionsof Echinocactusand of Opuntia

versicolor are suited by the heavy outer epidermal wall, which

is cuticularized,as well as by the stomata of peculiar structure,
to resistrapid loss of water. This is the typeof structurethat is
to be foundduringthe periodsof drought. The embryonicpor-
tions of these cacti, and the evanescent organs, in which are
includedthe leaves of Opuntia, are well adapted to promotetran-
spiration. This is accomplishedin the embryonictissues by a
thin epidermalwall and by the undifferentiated portionsof the
outerpart of the cortexby whicha rapid transfer of wateris possi-
ble. The substomal tube functionsalso throughoutits entire
lengthas the substomalchamber. In the leaves of Opuntia not
only is the epidermalwall delicate,but the outer cortexis never
differentiatedinto sclerenchymaand chlorenchymaand there is
no substomalcanal. Such is the structureof the tissues at the
timeswhentherateof transpiration is mostactive.
5. The leaves of Opuntia versicolorplay an importantrole
in transpiration. In one instancewith somewhatless than one
fourththe entiretranspiringsurfacethe leaves transpirednearly
one half the whole amount. In anotherinstance about 45 per-
cent of the entiretranspiringsurface was foliar and the leaves
transpiredabout 55 percentof the total amount.


:04. Transpirationder Pflanzen. Jena.
:05a. A NewMethodof Measuringthe TranspirationofPlants inPlace.
Bull. TorreyBot. Club, vol. 32, p. 515.
:05b. On the Transpirationof Fouquieria splendens. Bull. Torreny
Bot. Club, vol. 32, p. 397.
'93. Botany oftheDeath Valley Expedition. Contr.U. S. Nat. Herb.,
vol. 4, p. 43.
:03. The Desert Botanical Laboratory. Public. CarnegieInst.,Wash-
E. S.
:05. Mechanical Adjustment of the Suguaro (Cereus giganteus) to
VaryingQuantitiesofStoredWater. Bot. Gaz., vol. 32, p. 57.
:04. The CreosoteBush (Covilleatridentata)in its Relation to Water
Supply. Bot. Gaz., vol. 38, p. 122, 7 figs.
:05. Soil Waterin Relationto Transpiration. Torreya,vol. 5, p. 25.