WNW”HWHIIHHHHI

Hill

 

PHYSIOLOGKZAL STUDIES ON
CEANOTHUS AMERICANUS
Thesis. for Degree of Ph. D.
Edwarcl: J. Petty
i 9 2. 5

 

ppppp

?dYbIOLUGIChL bTULIEL OR CEhnOTflUb ALEEICnhUb
'The Root Nodules and Some Characters of

the Causal O genism

A Dissertation Submitted in “artiel Fulfillment of
the Requirements for the Dewree of Doctor of
Bhilosoyhy in the Lichigen State College

of Agriculture aha Applied Science.

deurd Jacob retry, M. Sc.

June 1925.

r 3
b1

-‘ \ “ ‘7 :1“*‘.7'.‘..;;

I. Introduction . . . . . . . . . . . . . . . . . 1
II. Historical O o O o o o o o o o o o o o o o o 0 2
III 0 theribfifitfil o O o 0 o o o o o o o o o o o o o

A. Growth and distribution of nodules . . . . .

010107

1. Rate of growth and factors concerned . . .
2. Rode of brenchinq in nodules . . . . . . . 8
5. Distribution of nodules . . . . . . . . . 10
E. Correlation of nodulation with growth of
host . . . . . . . . . . . . . . . . . . . . 13
1. Field studies . . . . . . . . . . . . . . 13

("I

2. Greenhouse eXperinents . . . . . . . . . . 41

C. The causal organism . . . . . . . . . . . . 52
1. Isolation methods . . . . . . . . . . . 53
2. Some characters of the causal orgnnism . 57

5. Infection of roots nnd production of
nodules . . . . . . . . . . . . . . . . . 42
4. Re-isoletion of Causal organism .. . . . 46
IV. Discussion and conclusions . . . . . 47 and 49
Acknowledgments . . . . . . . . . . . . . . 51
Literature cited . . . . . . . . . . . . . 52

EXplanetion of plates I to X . . . . . . . . 54

In}
A
\.I
93
l" Ct
5M
\j 1
PF.

H‘Y’v

n‘bIOLOaICAL STUDIES on CDdKOinLS AfijllcnIUS

The Root nodules and Some Characters of

the Causal Organism.

I. Introduction

The structure and functions of the root nodules of non-
leguminous plants have engaged the attention of numerous in-
vestigators during a century. dowever, in no case has the
causal organism been clearly isolated and tested.

The nodules of Ceanothus americanus L. have been studied

 

by several investigators (6)* since Dr. Beal first called at-
tention to them.in 1890 (5); and Bottomley (4) claimed to have

isolated a strain of Escudomonas radicicola Beij. from Speci-

 

iens collected in America by Rosendahl and others. But Bot-
tomley (4) did not test the ability of his isolated organisms
to infect the host plant. Burrill and Hansen (6) shortly
thereafter failed to isolate any organism after many attempts
with a wide range of media; and on this account, the claims of
Bottomley have been doubted by many investigators.

The ,historic interest of geanothus americanus, as well as

 

its use in ornamental plantings, has made desirable a more de—
tailed knowledge of the physiology of this plant. Furthermore,
the general ecological significance of the root nodules of

Ceanothus takes on added interest due to the important role

 

discovered for analOgous structures on legume roots; and this

*The parenthetical indices refer to the authors cited in the
literature given after the conclusion of this paper.

- 1 -

2.
interest is not lessened by the fact that sinilar functions

have been attributed, even if on vague grounds, to the nodules
of such non-legumes as Cycas (8), Iodocarpus (9), Alnus (8),
Eyrica (4), Elaeagnus (8), and Ceanothus Species (4).

The observations reported in this paier are the results of
experiments undertaken with the purpose of discovering more
substantial bases for the claims made for nitrogen assimilation

of Ceanothus americanus L., and to determine some of the char-

 

acteristics of the causal organism of the root nodules.
II. Historical

When Beal reported the presence of nodules on Ceciothus

 

emericanus to the American Botanical Society (3), Atkinson be-

 

cane much interested in them and immediately began the study

of their structure and of the causal organism (2). He gave a
description of the mature nodules to which little has been added
by subsequent workers. Easing his findings upon the organism

as it occurs in the host cells only, he referred it to a new

species Frankie ceanothi.

 

Among the later investigators only Bottomley (4) and Epratt
(16) dispute this classification.

Upon the role of these nodules Atkinson makes no correlative
field observations on growth, and he dismisses the whole matter
of symbiosis wdth the statement that "so far as can be seen they
(parasite) cause no inconvenience to their host". He reports
no attempts at the isolation of the causal organism. He also
states that the older hyphae are devoid of protOplasmic contents.

Arzberger in 1910 (l) repeated the structural and taxonomic

studies of Atkinson without reaching different conclusions, but

3.
he also investigated.tke enzymatic action of nodule sap. He
found a decided ability of the juice to peptize certain proteids
(blood fibrin) and from this concluded that the empty hyphae
noted by Atkinson have in all cases had their contents digested
by a proteolytic enzyme, which, as he states, is probably se-
creted by the host plant. He concludes the "Symbiosis“ (nu-
tritive mutualism is most probably meant) "exists, which is
quite apparent in the early stage."

In 1915 Bottomley (4) reported briefly some studies on the
structure and branching of the nodule and he also presented
data bearing upon the classification and nitrogen-gathering
ability of the causal organism. He claimed that the branching
of nodules is lateral and not dichotomous. He also isolated

an organism which he identified as a strain of Pseudomonas

 

radicicola Beij., and from very limited eXperiments concluded

 

that it could assimilate very small amounts of atmospheric
(elementary) nitrogen. From these few tests and without any
inoculation trials he concluded that "it is therefore evident

that the root nodules of Ceanothus are definitely concerned

 

with nitrogen assimilation". Neither Arzberger nor any other

investigator of Ceanothus had, up until 1918, reported any ef-

 

fort to correlate general growth with degree of nodulation.

It was to be eXpected that investigators interested in leg-
ume nodule organisms would immediately try to confirm Bottom-
ley's conclusions. Especially Burrill and Hansen (6) tried
repeatedly to isolate the organism by employing a wide range

of the best Pseudomonas media; but they always failed to get

 

4.

anything remotely suggesting Eseudomonas radicicola from the

 

Ceanothus nodules, although the media used were always success-
ful in isolating various strains of Iseudononas from various
leguminous species. Since Bottomley had used only ordinary
"Pseudomonas culture solution", Burrill and Hansen judged that
Bottomley's conclusions were of doubtful taxonomic reliability,
and were more inclined to accept Ath'nson's identification as
the correct one.

These investigators were not able to grow the wild plants
or tneir seedlings successfully in sand cultures, nor does Bot-
tomley or any other investigator, except the present writer (14),
report the successful growth of this somewhat fastidious plant
when grown under artificial conditions. Burrill and Hansen
succeeded in sprouting only five per cent. of the seeds even
after treating ten minutes with concentrated sulfuric acid,
while the writer at the beginning of this investiaation (1920)
(14) succeeded in obtaining regularly from forty~five to seventy-
six per cent. with the sulfuric acid treatment and thirty-seven
to sixty-five per cent. by ordinary hand scarification with
sandpaper. Part of this wide divergence Kay be explained, in
the eXperience of tne writer, by locality of seed production,
and part Hwy be due to differences in methods of sterilization
and germination. In the paper referred to, the writer (14)
also reported effects of sterilized soils upon the Sprouting

and further rrowth of geanothus seedlings. These effects are

 

of importance in cultural work with this species, but they

need not be summarized here.

5.
fhus the role of the nodules, the correlation of their rate
and mode of growth with general growth, as well as the isola-
tion and identification of the causal organism, were all very
uncertain when the following experiments were begun. horeover,
the successful growth of sterile seedlings, and especially in
sterile sand cultures with.and without nitrogenous fertilizers
-- matters of the greatest importance in physiologic studies
of tnis nature -- had not been attempted.
III. EXperimental
A. Growth and Distribution of Kodules.

In order to correlate general growth with nodulation it is
first necessary to find a nmthod of determining the age and an-
nual growth rate of nodules. Otherwise, a positive correlation
between nodulation and growth in a perennial of this type would
lead to error in case a plant had only slight nodulation till
late in life. The mode of branching in nodules may also be
related to the determination of their age. The following ob-
servations on these points were made during two growing seasons.

1. Rate of Growth and Factors Concerned.

a. Annual Growth and the Glass Plate method of Observation.
About two hundred nodules on twenty different plants were

carefully eXposed. These were covered by glass plates ac-

cording to the method of thougal (9), to provide for later

examinations without disturbing the root system. Small

constrictions were noted on most nodules and these were

searched for in other plants. The significance of these

constrictions was at first not appreciated, but by counting

6.
them and comparing with the age of the plant as shown by
the annual rings of its hypocotyl, it was found that they
agreed in most cases where the oldest nodules of plants
from two to five years were considered. All discrepancies
were due to branching of the nodules in the first year's
growth, and a previous failure to note that in such cases
there is no constriction at the bases of the branches.

This discovery later proved to constitute a valuable
criterion of rate of nodule growth in different parts of
a given root system for different years and it later formed
the basis for comparisons of nodule growth in different
soils and habitats.
b. Observations on modules in situ.

During the autumn, winter, and spring of 1918-19 and
only autumn and Spring of 1919-20, observations were made
at intervals of two weeks. During the first half of the
first winter, which was a mild one, about thirty per cent.
of the nodules nearly doubled their growth as measured by
tnat of the summer, While forty per cent. only added from
one-fifth to three-fifths as much. The remainder grew
slightly or not at all. Those nearest the tap root grew
most rapidly.

No growth was noted during February or Harsh, but in mid-
April it was resumed again without stOpping, till the next
October when a severe winter set in. This did not allow
measurement again till January 1920 when no increase over

the October's measurements could be noted. late in the

following April slow growth was resumed again.
0. Observations on hodule Growth of Elants taken to the
greenhouse.

Similar observations were made daily on the nodules of
five plants brought to the greenhouse from the same habitat.
Here some fifty nodules made a more or less continuous
growth for nearly two years. Each increase in growth of
sufficient size to be measured took place during and im-
mediately following considerable periods (l-2 weeks) of
sunshine. No other factor could be satisfactorily corre-
lated with nodule growth in these five plants.

Summarizing briefly, these observations on Rate of
Growth indicated (1) that nodule growth continues beyond
the normal growth period of the tOps if the autumn and win-
ter are mild and that growth is stopped by very low temper-
atures, (2) that active nodule growth attends and extends
beyond a period of photosynthetic activity and therefore
most probably depends largely upon the products of photo-
synthesis, (3) that nodule growth is isodiametric or with-
out notable constrictions, when the conditions are favor-
able to general growth over long periods (one or more
_years), (4) that the annual constrictions of nodules make
possible comparisons of nodule and general growth in dif-
ferent parts of the same root system throughout the differ-
ent years of the plant's life,and.finally (5) that the ef-
fect of different habitats on nodule and general growth can

thus be more accurately judged.

2. Kode of Branching in Eodules.

In determining the age of a nodule, the branching must
be considered, as has been previously pointed out. All writ-
ers except Bottomley agree that branching is dichotomous or
polychotomous. Bottomley (4) states that as many as five
branches may be formed in one year but that they are succes-
sive. Such an interpretation of the mode of branching is
inconsistent with the formation of annual constrictions as
noted above, because branches do not occur between constric-
tions, except rarely in the first year's growth, and here
the coloration and other evidences of simultaneous branching
are too strong to escape notice. The following observations
were therefore made upon several hundred nodules to determine
upon what ground, if any, such an interpretation as hat of
Bottomley might rest. Three methods were followed: the

Q

first consisting of measurements of the length of di- and

H:

polychotomors branches; the second 0 longitudinal and cross

 

sections of incipient branchings, and the third, applicable
only in the case of the Oldest nodules, a comparison of ac-
tual age of the plant with the age of nodules as indicated
by the constrictions detailed above.

The results of these observations are as follows: Firs ,
in all cases (five hundred in number) of young nodule branches
neasured, their length was practically identical within the
group (lls. I and III); while in older branchings (six hund-
dred in number), uneven growth in a branch in some succeeding

year, or a total cessation of growth in one or more old

branches of a group, gave the appeararce of lateral out-
growth, inasmuch as these old branches were often thinner
and were crowded out of position by the more sturdy branch
of equal age which had greater capacity for growth.

The second method of exernration included about five hun-
dred nodules whose earliest branching stages were longitudi-
nally, or in some cases transversely, sectioned. Some of
these were stained after being passed through the usual
paraffin technique schedule. All showed the same nethod of
branching in which the growing point flattened or became
lobed, depending upon west the future position and number
of branches was to be. Each lobe therefore becomes a new
branch and thus they must be simultaneous in origin even if
they should not grow to equal length subsequently. The fig-
ure shown by Bottomley (4) is certainly that of a section
which does not follow the long axis of each branch (in a
pair), hence one seems older and.nore develOped than the
other. Only the careful following of a complete series of
such sections of a Whole nodule, or the use of younger stages
of branching could avoid misjudgment in this connection.

The general or longitudinal type (21. I,fig. 2) of infec-
tion and nodule formation which was rarely found by the writ-
er, produced as many as five branches simultaneously and in
a row. In these as in other cases, branching of the plerome
and stele is often accomplished before the apex of the nod-
ule snows much deformation.

In using the third method of determining the mode of

lO.
branching, the oldest nodules of plants from one to five
years old were found most suitable. The age of the plant
usually was first determined. The annual constrictions were
then taken into account. Thus in over two hundred cases ex-
amined at random the ages of branches agreed exactly.

When this test is applied in the reverse order, the first
part; viz., age of nodule, in a small percentage of cases,
appears greater than that of the plant by one year, as deter-
mined by the annual rings of the wood. This is due to the
fact mentioned previously that in the vigorous growth of some
young plants the main body of a nodule or of a branch divides
while it is still actively elongating. But in such cases
the constriction at the point of branching is not evident as
it is in all other cases where branching usually occurs early
in the growing season. Koreover, the uniformity in color of
such exceptional branches, with their branch of origin, per-
sists for at least two years.

Briefly stated, these observations furnish no support for

the claims of Bottomley that branching of Ceanothus nodules

 

is not siuultaneous; but on the contrary, they indicate that
one may misjudge the type of branching, due to (l) faulty
sections when not studied in series; or (2) failure of some
nodule branches to grow, while their contemporaries monOpo-
lize the food and push them out of position; ard finally (5),
that these facts may cause inaccurate age comparisons to be
made.
5. Distribution of hodules.

MicrOSCOpic examination of nodules has always disclosed

11.
the fact that the cortical cells are nearly filled with
starch. If the metabolism of the parasite requires starch
or starch-forming sugars and if it liberates materials of
benefit to the host, it would be interesting to see if nod-
ule distribution conforms to the most efficient location of
such a double function. The following observations on field
and greenhouse seedlings were made to determine whether nodu-
lation has some regularity or not, and if regular, whether
this can be correlated with other habits of growth or facts
concerning food distribution.

Over one thousand root systems were carefully studied,
over seven hundred of which came from three different types
of habitat soils, two hundred were greenhouse seedlings from
seven to nine months old grown in several types of soils,
while seventy were plants of all ages transferred from field
to greenhouse. The nodules were removed from these latter
plants which soon produced new nodules on new roots.

a. Of the seven hundred plants studied in the field two
; hundred of all ages were dug from a fine clay soil. These
had most of their nodules at a depth of from three to ten
; cm.; aid the largest, i.e., those of most rapid growth, were
3 located in the upper part of this stratum and mostly on or
wery near the primary root. Below a fifteen cm. depth the
nodules were small in diameter and of slow growth.

Three hundred plants dug from sandy soils showed a slight-

ly deeper distribution than was found in clay soil. Below a

fifteen cm. depth the nodules were slightly larger and more

12.

numerous than in the clay soil. a few were found at one hun-
dred fifty an. below the surface of the soil which is about
seventy-five cm. deeper than the deepest found in the clay
soil.

Over two hundred plants from a sandy loam soil had essen-
tially the same distribution of nodules as those in the sandy
soil. As found for the other soils, the lowest nodules here
showed the smallest diameters and shortest annual growths.

b. The seventy plants brought to the greenhouse conformed
in every way to the distributions just mentioned. The nodules
were removed and the longer roots were pruned to fit them for
planting in pots having from fifteen to twenty cm. of soil.
They formed many neW'roots and nodules, and the teps grew
luxuriantly. The bottom root growth was especially heavy
due to bottom heat and good aeration, but the nodules were
larger and more numerous in the upper half of the root sys-
tem. However, the differences between the lowest and highest
strata were not nearly so great as in the field conditions
of these plants.

0. Three hundred greenhouse seedlings naturally infected
by habitat soil showed most nodule growth in the upper two
to eight cm. These plants had ample Opportunity for a dif-
ferent distribution. Iriority of infection.and a possible
acquired inmunity are not factors here, because the higher
(El.VII) lateral roots bore nodules, while the lower roots
did not.

The explanation for this distribution of nodules as well

*3

D.

H
Q»?

as for that of the field plants, it would seem, does not lie
in the variation of oxygen supply as all were well aerated,
except those from the clay soils which had numerous nodules
at depths known to be poorly aerated. Hor does it lie in
the water supply for the regions favoring root growth most;
i.e., those of Optimum water supply were not the most highly
nodulate regions of the root systems.

d. In sectioning the root crown to determine age it was
always fbund to be excessively thick for a plant of such
size, age, and woody character. This extra thickness tapers
off very rapidly in the upper ten to fifteen cm. of its length.
Here the largest amounts of food are deposited as shown by
the thickness of the annual rings and here nodule growth is
always most rapid and plentiful.

It would seem then, that this region of food deposit would
be most favorable for the nourishment of parasites and at the
same time most favorable for the distribution of useful foods
or stimulants made by them, to roots and shoots of the host
plant. The high degree of correlation found between this
lodgment place for food and location of most nodule growth
amounts to certainty or causation. a corresponding degree
of certainty might be anticipated far a reciprocal distribu-
tion of file effect of the parasite upon the host if such
could be demonstrated.

Correlation of Hodulation with Growth.
1. Field Studies.

As previously noted, Atkinson (2) did not consider the

l4.

root nodule organisms of Cemnothus americanus detrimental

 

to growth, while Arzberger (1) probably intended to convey
the Opinion that a mutualistic symbiosis exists. Bottomley
(4) claimed that these nodules are definitely associated with
atmospheric nitrogen assimilation, without giving proof of
the Specificity for nodule production of the microdrganism
isolated. Burrill and Eansen (6) failed to isolate an organ-

ism, and on this account, as well as because even Pseudomonas

 

radicicola synthesizes only minute amounts of organic nitro-

 

gen lg vitro, they considered a mutualistic symbiosis involv-
ing nitrogen assimilation in these nodules, to be of very
doubtful existence.

If such a relationship does exist in natural habitats,
then careful comparisons among many plants of different de-
grees of nodulation should give some positive indications
that the possession of nodules is of'soue advantage to 9322‘
othus, as is found to hold in the case of leguminous plants.
In view of these considerations and of the foregoing detailed
methods of’uaking exact comparisons between the annual growth
of nodules and general growth, the fellowing observations-
were made on the total growths found in the field.

Actual correlative studies were preceded by a general
survey of ecologic conditions such as tepography, type of
soil, and associated plants. The plants were not dug at
random on account of the great age of some and on account
of lack of very young plants in the first habitat due to

excessive pasturing. Equal sampling in different types of

15.
soil and a fair representation of the different ages from one
to ten years were always sought so far as the growing materia
would allow. Before digging, the individuals were studied in
a general way as to size, vigor, color, and branching of stems,
ind their blossoming or fruiting habits. After careful digging,
the age of the plant, diSposition of nodules, and comparisons
of their annual growths with those of the annual rings in the
root crown, were made and in most cases rough weights were al-
so taden. The data are presented by habitats as follows;
A. First Habitat

This habitat is a low gently sleping hill whose flat tOp
consisted of fine calcareous clay and whose sandy sides were
gullied in such a way as to prevent the deposition of the clay
washed off the tOp. The eXposure is to the south and east.
Small clumps of peplar, white oak, hickory, and sumac occur
over this twenty-acre tract. but the Open spaces have been oc-
cupied during the last two decades by pear trees. The areas

occupied by Ceanothus are low in nitrogen (below 0.05 per cent.)

 

and variable in acidity (Eh. 6.5 to 6.9). The clay soils are
so close-textured that drouth effects often appear.

The numbers and ages of’plants dug frOm'this Habitat are
as follows; thirty were from five to six years old; eighty from
from seven to eight years old; seventy from nine to ten years
old,and'twenty were from ten to twelve years old. Their dis-
tribution according to age and relative smount of nodulation

is shown in Table I.

fie
We

Table I. Distribution of

...—

TTr
-AL‘

according to age, size, nodulation in b

16.

othus americs

V1113

itat l.

 

 

 

 

 

 

Age of Total ; Type of nodulation and relative size
:number: of plants dug.
seed- ; of :
:seed- : Eoorly : hedium; Heavy ; Very heavily
lings* :lings :nodulate:nodula—; nodula~z nodulat
: ; buall . tion. ; tion. :W.er large plants.
: :plants :Ledium : Large ;
: : ; plants: plants :
5 years** 10 : 5 z 6 ; l ; .....
6 " ; 2O : 2 z 5 : ll ; 2
7 “ : 50 ; 2 ; 7 : l6 ; 5
8 " z 55 z 5 z 7 ; 55 ; 10
9 " z 45 z 2 ; 5 ; 50 ; 8
10 " z 25 : l ; 5 z 15 ; 6
ll " ; ll 3 l : 2 : 5 : 5
12 " z 9 : l : l : 5 : 4
Eotals ; 205 z 15 z 56 : ll6 : 58
* Seedlings, as d: tin Uiflfid fro; flants n oe by stolen—
like growth f:om very old plants.
** ‘bout lOO seedlings from tIo to three years old are not
included, because atyéic al, due to shade and crowding.

It will be noted that in this total of two hundred and
five plants shamined, fifteen were poorly nodulate, thirty-
six were of medium nodulation, one hundred and sixteen were
heavily nodulate, Llile thi rty-eight Were very heavily nodu-

late. The poorly nodulate

1+
U

averare weig at the sane age

was about lalf of

ylants were all shall and their

that Of the

17.
next group, each of which had from one to three fair-sized
nodules (Pl. III) as old as the plant itself; i.e., infection
and nodule growth began soon after the syrouting of the seed.
The nodules of the smaller plants were so new that they indi-
cated late infection. The one-hundred sixteen heavily nodu—
late plants had from three to ten very large and much-branched
nodules on the main root and many smaller ones among the sec—
ondary roots. These plants averaged over twice the weight of
those in the medium nodulate group owing to profuse branching
of tOps (kl. II, tOp) and heavier root system. Young plants
of this group were frequently supposed to be several years
~older than the age shown by the number of annual rings and
bud scars. She very heavily nodulate group of thirty-eight
plants (11. II) had their nodules distributed as in the heav-
ily nodulate group, but the nodule branches were densely packed,
and these large dense nodules covered the tap root as well as
the large lateral roots to such an extent that the observer
would eXpect a detrimental effect on tOp growth if the rela-
tionship were purely parasitic. Yet the size and weight of
tOps alone ranged from two to three times that of the heavily
nodulate groups, while the root systems bore a similar ratio
down to a depth of over one metre.

In this habitat, rapidity of nodule and general growth go
hand in hand, and the total growths are vastly different in
the different groups based on amount of nodulation and age.
In general, the plants of clay soil were smaller than those

of sandy soil. They had fewer nodules, due possibly to a

18.
limitation placed upon infection by this close-textured soil
-- a phenomenon also noted for greenhouse seedlings when
planted in a clayey soil which had been compacted by wet
sterilization.

E. Second Habitat.

This tract of about six acres lies on the sides of a nar-
row vale or glen trending north and south. The soil consists
of fine sand only, and.is roughly covered on the east side by
brambles, sumac, and rock rose, while the west side is cov-
ered by red and white oak and hickories standing in rather
ogen foniation. The total soil nitrogen content of this area
(0.06 to 0.07 per cent.) is only slightly higher than that of
the first habitat. Both sides of this valley receive much
less sunlight than the first habitat and growth, in general,
is notably less in amount even though the humidity is slight-
ly higher.

The numbers of plants dug from the second habitat and.stud-
ied as in the first habitat are as follows; two hundred and
seventy-eight were from one to two years old; two hundred and
sixty-seven from three to four years old; one hundred and fif-
ty-eight from five to six years old; eighty from seven to
eight years old; and thirty-three from nine to twelve years
old. Their distribution, according to degree of nodulation,
is shown in Table II.

Amongst this total of eight hundred and sixteen plants

forty are poorly nodulate; two hundred are of medium nodula-

tion; three hundred and ninety-six heavily nodulate; and one

Table II.

according to

 

Age of: Total

.-'_

Distribution.of

383.

size,

19.

geanothus ameri canus

 

and nodulation in Eabitat 2.

 

Type of nodulation and relative size

 

 

 

:number ; of plants dug.
seed- Esegg- f Poorly :Medium : Heavy : Very heavily
lin s : linas jnodulate:nodulationznodulation: nodulate.
g " ~ : Small : medium : large ; Very large
: : plants : lflsnts : Iflants plants.
1 year: 111 : 7 : l9 : 65 : 22
2 years; 167 : ll ; 58 : 85 ; 35
5 " : 163 : 9 : 41 : 79 : 34
4 5 ; 104 ; 5 : 27 ; 44 : 28
5 5 z 81 ; 4 : 22 : 54 ; 21
5 3 . 77 : 5 : 20 : 55 ; 19
7 5 : 53 ; 1 : 15 ; 25 : 12
8 5 z 27 z .. ; 8 : l4 : 5
9 5 ; 26 : .. ; 9 : 16 ; 1
10 2 : 5 : .. : l : 2 : 2
ll 3 : l : .. : .. : .. : l
12 8 : l : .. : .. : l ; ..
Totals : 816 : 4O : 200 : 396 : 180
2:;—

 

 

hundred and eighty very heavily nodulate.

The distribution

is better here among the younger plants than in the first

habitat, because pasturing has apparently been very light

within the previous four or five years.

The differences among the younger seedlings were not so

apparent, as between heavily nodulate and poorly nodulate

20.
older plants (321. III), but careful measurement showed that
they were relatively almost as great as among the older
groups. In this habitat (Table II), precisely the same rela-
tions hold as those found in the first habitat.

The distribution over this area is much more regular than
in the first habitat and it is difficult to believe that these
plants are hereditarily so different amongst themselves, so
far as vegetative thriftiness is concerned, as to cause such
great variations in size independently of, or rather with,
the handicap of the nodules which are known.to store quanti-
ties of foods which they seem never to disgorge, and to har-
bor a parasite Which must feed on organic foods.

In this habitat, heavily nodulate plants blossomed more
COpiously and uniformly and fruited more heavily than the
poorly nodilate plants. The annual growths in thickness of
the tap root and attached nodules paralleled that of the
neighboring area, while the nomiles at greater depths than
fifteen or twenty em., made much less growth in length and
thickness. Some unthrifty seedlings of one season's growth
were found with very minute nodules or none. Some dead ones
with no nodules were found in the same areas where sufficient
moisture and sunshine were present. These seem to indicate a
survival value for nodules which could not, however, be es-
tablished with the small numbers; viz., seventeen non-nodulate
and dead, and twelve plants with minute nodlles, which bere-
ly survived their first winter.

Lummarizing the data from hese field studies it.is found

that:

2l.

1. Plants are larger, heavier, and more thrifty, in prepor-
tion to the number of large nodules found on their roots.

2. These nodules are found in the upper part of the root
system near to or upon the primary root.

3. seed bearing is heavier aid fruits ripen later in.tls
heavily nodulate plants than in poorly nodulate plants.

4. Very young seedlings which were heavily nodulate with-
stood winter conditions and competition better than poorly
nodulate plants, and a few very small non-nodulate seedlings

died fEAOIS.‘ 110 a ‘;£3I'e;jlt cal-13:8.

V
«A.

2. Greenhouse Inneriuents.

a.

A. tprouting of normant Field flants.

Ihe positive correlations be tween growth aid nodulations
found in the field studies soon raised the question whether
a correlation between forced growth or sprouting from the
dormant condition might not be found and whether the complete
removal of nodules would have a notable effect upon Sprouting
from dornency. Several series of field plants were tested
in the greenhouse. The plants were dug very early in the
spring and were tested in two series as follows: Series I
consisted of plants of different degrees of nodulation; Series
II consisted of fifty pairs of plants, each pair being alike
and one plant of each pair having all its nodules removed be-
fore Sprouting.

Series I.
The forty plants of fitis series were selected to represent

the various degrees of nodulation found in the field studies

22.

described above. nly plants four, five, and six years old
were represented on account of convenience in potting. Habi-
tat soils and greenhouse compost were used in an equal number
of cases, and bottom heat was used to shorten the dormancy
period.

The sprouting data are as follows: twelve heavily nodulate
plants Sprouted one week earlier than fifteen plants of medi-
um nodulation, and there was no difference as between habitat
soil and greenhouse compost. Thirteen poorly nodulate plants
were not only from one to two weexs slower in Sprouting than
the medium nodulate, but they had fewer and much smaller
shoots. Later field examinations to the number of over one
hundred, showed similar differences. The extremes of these
differences must give the advantage of an early start, equi—
valent to two or three weeks of growth in the Spring, before
shading in forested areas occurs.

Series II.

In order to see if living nodiles exercise a direct or im-
mediate effect upon sprouting, the following three sub-series
consisting of one hundred dornant field plants were Sprouted
in the greenhouse. These were carefully selected in pairs
of equal size, nodulation, age, and habit (218. II and III).
All nodules were carefully removed from one plant of each
pair before potting.

Sub-series A.
The first sub-series consisted of fifteen pairs of one-

year old seedlings carefully dug.and carried to the greenhouse

25.

in moist soils. After carefully hatching;the pairs and prun-
ing off the nodules with a sharp scalpel the whole root sys-
tem was potted. The pairs were numbered (Pl. V) and the de-
nodulated plant was labeled “2H". Six of these pairs were
heavily nodulate and nine were of medium nodulation. After
four weeks of sprouting, notable differences between nodulate
and denodulate plants of each pair existed. The leaves of
the nodulate ("+h", Pl. V) were in all cases somewhat larger,
more numerous, and more thrifty in.appearance than those of
the denodulate plants and the new stems (scarcely visible in
P1.V) were from two to five times as long in the nodulate as
in the denodulate plants. There was no exception in this
sub-series, all plants of which came from a small area of
uniform soil and other conditions. The nodulate plants
showed larger differences in growth according to nodulation
than the denodulate plants. No injuries to roots due to re-
moval of nodules could be found several weeks later, and all
wounds were covered by healthy callous tissue. In this sub-
series the nodules seem to exert a very noticeable effect
upon the rate of sprouting.

Sub-series B.

The second sub-series consisted of twenty pairs of plants
from three to six years old (Pl. II, lower figure). These
were heavily tOp- and root-pruned after carefully matching
pairs. The nodules were removed as in sub-series A, but the
plants were put into suitably larger pots holding from one to

two kilograms of soil. These, as in sub-series A, were care-

24.
fully top watered with rainwater and Kept at thirty to fifty
per cent. of saturation.

In a few Weeks the differences were even more notable as
between nodulate and denodulate plants than in the first sub-
series and the differences continued to increase for more than
two months. While the root systems of these were very severe-
ly reduced, the number of potential shoots or buds was prob-
ably even more reduced (to three nodes) by t0p pruning, thus
probably having the effect of exaggerating the differences in
the pairs due to the presence and absence of nodules. It is
scarcely probable that the extra wounding due to the careful
removal of nodules could make such a difference between plants
that had been so equally severely wounded in both tOps and
roots. Alligghnds were well calloused over.

Sub-series C.

The third sub-series consisted of older plants than those
of sub-series B. The fifteen pairs ranged from five to eight
years old and were pruned and otherwise treated as sub-series
B, except that three of the denodulate plants were watered
more heavily than their nodulate rates.

No differences due to extra water ar‘to richer soils could
be noted and the wrwle sub-series behaved exactly like the
second sub-series of which it say be considered a continuation.
The small fluctuations in water content throughout all these
sub-series is very probably of small consequence. But the
correlation of more rapid Sprouting with the presence of nod-

ules, or the converse,-- the consistent retardation.of Sprouting

N
01
O

accompanying the removal of nodules, is present throughout,
being more marked in the older and more severely wounded pairs.

From these data it seems clear that Ccanothus receives not

 

only a general benefit to growth from the possession of nod-
ules, but that there is an immediate benefit or advantage se-
cured during the sprouting period. Whether this effect is due
to hormones, to nitrogenous or carbohydrate foods stored or
made in the nodules is not as yet clear.
3. Sterile and Hodulate Seedlings.
The eXperiments and field observations already given do

not answer the question as to whether Ceanothus can grow nor-

 

mally without nodules nor do they indicate what the amount of
growth would be in case non-nodulate plants could be grown
artificially. In order to get some light on these questions
the following series of seedlings were grown in two different
soils. One was the soil of the second habitat previously
studied, while the other was a fertile, fine-textured green-
house compost containing O.21 per cent. total nitrOgen, thus
being three times as fertile in reSpect to nitrogen as the
habitat soil used.

seedlings were from seeds that had been carbonized by
treating ten minutes with pure sulfuric acid. These were
plaited in sterilized soils and transferred after a week's
growth to pots of sterile soil, by the use of a Specially
constructed planting tweezers with broad blades, which re-
moved the unbranched priiary root without injury.

For the production of nodulate seedlings (l4) the seeds

were planted in habitat soil taken from a region thickly

Cr-
(«00

COVered by nature wild Ceanothus plants. Io avoid soil ef-

 

fects, this sprouting soil was carefully washed fro r the
roots tmlen tfuusflldntlhu

The pots were well packed with moist soil and sterilized
at various pressures, bsettent results showing that heating

at a pressure of two pounds per sguare inch was sufficient to

prevent nodule formation. The following- eedlin s were grown
in the two soils negated; in the green -house 001.. cost, ei ghty-
eight were in sterilized soil and ei ht ty-five in untterilized
soil. Of the twenty-two per cent. of infected plants grown

in this soil only a few were accidental, the remainder having
been infected in the unsterilized habitat soil during germina-

tion. Sterile sprouted seedlings develOPed no nodules in this

soil, although it was well-infected by Iseudononas radio ieola

 

strains, as show; by the growth of red, white, sweet, and al-
sike clovers with nodules. Recalling Bottomley' s identifica-
tion of the causal organism, it is clear that even if it were

a :geudomonas the variety or strain which attacks Ceanothus

 

 

—‘

certainly wa not present in this soil. although flge nearest

Ceanotlus habitat was only about two miles distant, a wider

 

 

distribution of lseudomonas, by'wind, might be eXpected. Chis

 

leads to the idea that the cause 1 o~grsnism may be a less easily

distributed one and t11erefore not a variety of Iseudomonas ra-

 

dicicola, as Bottomley (4) claimed. The plants grown in the

 

COL 190st soil and their degree of nodulation are given in Table
III, their weights are given in Table IV, and their nitrogen

content, in Table V.

27.
Table III

Laboratory seedlings Found Infected

 

 

.m

 

 

 

 

:Total numberzflsnts with: neavily: Not
Plants and soils . plants : nodules ;infectedavisibly
; studied : :no nodules :infected
Fertile compost . : . :
Sterilized . 88 :16 . 5 :69
hot sterilized : 85 :17 . 2 :66
Totals . 175 ; 53 ; 5 z 135
Habitat soil , : . :
Sterilized :105 : 13 : 7 : 85
Not sterilized :161 3127 : 22 ; 12
Totals ; ~ 266 : 140 : 29 : 97
Totals, both soils: 439 : 173 : 34 : 252

 

 

Of the two hundred and sixty-six plants grown in habitat
soil, one hundred and five were in sterilized and one hundred
and sixty-one in unsterilized pots. Of the twenty per cent.
infected plants in the sterilized soil, all but two were in-
tentional. The data for nodulation, weight, and nitrogen con-
tent, reSpectively, are given for comparison,with those of the
plants grown in compost soils, in Tables III, IV, and V.

This table shows the considerable number of two hundred and
thirty-two non-nodulate plants which were grown for more than
eight months c0ntinuously. The heavily nodulate plants were
few and mostly in inoculated soil, whereas the medium nodulate
plants were sufficiently numerous for fair comparison with

other groups.

28.

The fact that all these plants had grown only one season
and that a Special type of infection not found in the field,
appeared in considerable numbers made necessary a slirhtly
different classification from that employed for field seed-
lings. This Special type or group named "many infections"
in Table III, had its roots literally covered by infections
which did not seem able, for some unknown reason, to develOp
into nodules. Very few infections (Pl. IX) had started to
produce minute nodules at the age of eight months, but judge
ing by their position on the older roots as well as on the
younger ones, these infections must have taken place as the
roots were forund. The only factors which correlate with this
failure to develOp nodules are bottom heat and thorough mixing
of the soil before planting, which, however, is not applicable
to the ten plants in sterile soils. Two of the five plants
with many infections in compost are accidental, while three
Were Sprouted in infected soil, and even though washed when
transPlanted, must have carried over some soil as inoculum.

Mag
ThisAalso explains the occurrence of the seven heavily infect-
ed plants in the sterilized habitat soil.

It will be noted that there is a decided correlation between
growth and degree of nodulation as found in the field studies.
Moreover, the non-nodulate plants are numerous enough to be

decisive as to how this species of Ceanothus behaves without

 

nodules, and how this behavior is affected by a very fertile
soil.

By comparison of non-nodulate plants of the fertile soil

29.
witn the heavily nodulate plants of the poor habitat soil, a
nitrOgen function of nodules may be indicated, whose magnitude
in terms of soil fertility can be directly read. And this
correlation is maintained to a high degree in the fertile soil,
where, according to eXperience with legumes, it would hardly

be eXpected. If nodules exercise a nitrogen effect upon growth
than it is apparent, from the habitat soil plant data, that in
one season a plant with many large nodules has a six to one ad-
vantage over the non-nodulate or sterile plant. The weights

of most of these plants are given in Table IV.

The weights of individuals did not vary enough from the
averages to caise overlapping of the groups. Unfortunately,
some plants were destroyed before the decision to weigh and
determine the nitrogen content, was reached.

It is remarkable that plants with many infections were not
fbund in the field studies. It is also remarkable that these
infections have the same degree of correlation with growth as
is found in the group of medium nodulate plants.

This ratio between non-modulate and nodulate plants, con-
sidered as a handicap, may eXplain the finding of the small
number of dead non-nodulate seedlings in the field which, as
noted previously, failed to survive the winter. Since the
soils had enough nitrOgen to make good growth, it is possible
that the parasite in the nodules excretes some hormone-like
body which enables the roots to take up amounts or types of
nitrogen not otherwise available. That nitrogen is here posi-
tively involved is sufficiently apparent from the data of

Table V.

Table IV.

Dry Height of Laboratory beedlings

 

 

"oil and groupszbescription :humber : Total : Average
; z o f ; we 1'. ght : we i ght
of plants : of plants ; plantszin grams : of one.
: : :air-dry. : plan .
Fertile compost; : 3 :
Group I :Large nodules: 9 : 7.520 : .855
Group II :medium " : 5 : 2.570 : .514
Group III :Small " : 1 z .445 z .445
Group IV : No " : 125 : 54.540 : .435
Group V :hany infec- : : :
‘tions (E1.IX : 2 : 1.570 : .685
Habitat soil : ; ; :
Group VI :Large nodules: 42 : 18.195 : .435
Group VII :hedium " : 40 ; 8.660 : .216
Group VIII :5mall " : 30 : 3.9L : .151
Group IX : "o " : 71. : 4.980 . .070
Group I :hhny infec- : z :
:tions (Pl.IX): 25 : 5.200 : .208

 

 

.4

Note; The decision to determine the dry weight of these
plants was not reached until after some plants had spoiled
subsequent to the regular nodulation studies of the root
Systems, hence the numbers of plants.are smaller than in

1

Table III.

Table V.

Nitroden in Laboratory Seedlings and Roots

 

Part of plant :Oven- 'Ash per Eer cent.*: Total

 

taken, and nitr05en; dry : cent. : nitrogen : nitrogen
estimation method ; mois- ; oven- : oven-dry ;oven-dry

used. ; ture ; dry : basis : plant, gms.
Ejeldahl method : 3 . .

Whole plants
Large nodules, : 6.60 z 5.60 : 2.05 ; 0.0171
Compost soil.

Whole plants
No nodules,
Compost soil.

0 c
Q
o
()3
O5
o

6027 3 2.14 : 000095

Whole plants
Lany infections,
Compost soil.

7.46 : 10.38 2.52 0.0172

.0
.0
.0

Whole plants
Large nodules, ; 8.60
Habitat soil.

12.60 2.45 : c.0107

Whole plants
Ho nodules, ; 8.80 : 11.50
Habitat soil.

N
0

CT.
DP

0.0018

Whole plants
Lany infections,
Habitat soil.

9.70 ; 2.45 : 0.0050

0 0
q
.
(>3
()1

 

 

* These are averages of closely agreeing duplicate deter-
minations. Blank determinations for nitrogen in reagents
were made and deducted, for tne above results.

The concentration of nitrogen in the air dry plants of
the compost soil averages lower than in the habitat soil,
but concentration of nitr0gen of the plants with many in-

fections is in the reverse order, in compost soil.

52.

It is remarkable that the non-nodulate plants of habitat
soil had the highest nitrogen concentration but less than
one-fifth of the total nitrogen found in the nodulate, and
about one-third the total nitrogen found in the heavily in-
fected plants of the same soil. In the rich compost these
differences are not so great, the total nitrogen of the non-
nodulate plants being slightly over one-half that of the nod-
ulate and heavily infected plants. The latter two groups
are identical even though the nodulate plants are far heavier,
as shown in Table IV. The non-nodulate plants of this soil
contain over five times the nitrogen found in the sane group
of the poor soil and also serve to emphasize the nitrogen re-
lation of nodules to the growth of the plant. But whether
this is one of stimulation, enabling the roots to absorb more
nitrogen from the poorer soil, so test the nodulate plants
of the poor soil were able to get as much nitrogen as the
non-modulate plants got from the richer (three times) soil,
or whether the nodules were the seat of nitrogen synthesis,
cannot be finally determined by these data alone. Nor do the
data of enhanced Sprouting given under greenhouse correlations
answer this question, whith will be further considered under

the physiology of the causal organism in the following division.

C. The Causal Organism
According to the descriptions of Atkinson (2) and Arzberger
(1), the organism is hyphal and branched. Atkinson referred

it to the genus Frankie of which he considered it a new Spe-

”-

 

cies, F. Ceanothi. Arzberger, as well as hurrill and Hansen

C“

I,
U o

(6), considered this classification more accurate than tha
of Bottomley (5), who placed the organism anon: the bacteria,

5
V

stating that its characters were those of lseudononas; he

 

maintained that its morphology (not published) as well as the

fact that it showed fixation of a heapheric nitrogen in min-

ute amounts, placed it among the strains of Is. radicicola.

 

Spratt (16) who studied nodule fornation, acce ted this clas-

*1
sification. Bottomley had not tried the infecting abilities

\4

0 w

of the organism leciated, and as there wa‘ some probability
that a pleomorplism night he f‘und which would clear up these
conflicting views, the following caperinents upon isolation

l‘

and infection were directed toward this end.
1. Isolation LLethods .

Laterial, consisting of many nodulate plants which were
taxen frOm a sandy soil, was potted and allowed to grow in
the greenhouse as soon as the resting period was broken.
This period lasted in preportion.to the lateness in the au-
tumn at which they were dug. Shose dug at mid—October re-
mained dormant three weeks, While those dug in mid-hoveuber
from the same areas remained dornant from six to seven Weeks
under similar greenhouse conditions.

a. The bacterial media consisted mainly of agars at first,

9715‘. I‘

m

but later several liquid media were also employed. The

m
3

may roughly be divided into stronrly nutrient and dilute
groups as follows:
Strongly nutrient:
(l) Iotato Agar with two per cent. dextrose.

(2) nutrient Agar with one to two per cent. peptone and

one to two yer cent. sugars.
(Z) nutrient Agar with one and one-half ger cent. pep-
tone mid two per Cent. glycerin.

Dilute media:

 

(l) agar with nutrient salts and Ceanothus juice.

(2) Agar with nutrient salts and asgaragin.

(3) Solutions of salts and asparagin and sugar,(aoetly
Coons' synthetic medium).

(4) Ceanotnus iuice, sterilized or filtered. Conbina-

U

 

tions of these; e.g., Ceanothus juice boiled or filtered

 

(raw) was also used with slants of the various agars for
proPagation and for nodule tissue cultures.

The Eotato Agar was prepared in the usual way by steepino
diced potatoes and straining the broth into the agar, after
which dextrose was added.

The fiutrient Agar Was prepared from ground beef in the
usual way, except that the raw juice was heated to boiling
and partly cleared with whites of eggs before being added

at 60° to 70° c. to the agar which contained the peptone.
In some cases filtration through flainel was necessary.
Glycerin and sodium chloride were added to the cold meat juice.

The Cevnothus juice was preyared from roots and subterra-

 

nean Sprouts in about equal prosortions, ten grams of this

material being ground in a mortar and 50 cc. water being

\J

x

- , rm 05Hw . w 1
added later. The nutrient salts wereAsouium phosphate

(nadg£04) and Lagnesium sulfate in the {reportions used by

Coons(7).

55.
In general, the relative values of these media are in pro-
portion to their dtnsity and nitrosen content; the Kutrient
gear with glycerin being the best, Iotato agar and heparagin
agars or solutions being intermediate, and the nutrient salt

and Ceanothus juices being weakest in their ability to bring

 

out many and large colonies. at the beginning, a nutrient
agar with tannin in low concentrations was used but a compar—
ison with nutrient agar growth showed the advisability of
discontinuing its use.

The two media finally used continuously were nutrient gly-
cerin agar, as above described, and the Asparagin agar of
Coons. Qhe latter tends to remain clearer after growth of
colonies, permitting a better view of what happens under the
colonies. It is also free from the tyrosine coloration pro-
duced by several organisms in the Nutrient glycerin agar,
although this was reduced some by reducing the amount of pep-
tone.

b. Procedure in Isolation.

Rodules whidh were unbranched and several years old were
washed and in some cases sterilized by chlorazine or by mer-
cury bichloride, for from one-half to one minute, then dried
under sterile conditions.

At first these were crushed before plating, but this
brought out too many soil organisms from the unsterilized
nodules, while the sterilized nodules resulted in blank plates.
Peeling of nodules was then resorted to, nder a large reading

glass. Bhis still brought forth a wide variety of organisms

56.
with no characteristic colonies. Then the tips of dry nodules
were broken off and the nodules were methodically peeled in
a revolving pair of tweezers, mounted vertically on a pinion.

This resulted in sterile plates and tissue slants. Few spe-
cies were obtained and these varied greatly. They were all
too large and too unlike the nodule organ em in vivo, to be
fifteen
tried out in the production of nodules. nevertheless, '
of the most likely ones were tried in sterile culture,, de-
scribed later, but without producing nodules.

Finally, the nodules were lightly scraped under a stream
of sterile nutrient salt solution and only the tips were
used. This brought out multitudes of similar colonies, about
the first of April 1925. These were immediately multiplied
on slants from which in a few days a heavy growth of inoculum
was produced.

The inoculum was prepared by softening the growth on agar
slants with a few cubic centimeters of sterile nutrient salt
solution (H/lfifi concentration) described later. This was
stirred carefully with a transfer needle, then decanted into
100 cc. of similar solution which was poured upon the roots
of the sterile seedlings, using special precautions to pre—
vent contamination. In the earliest tests, the roots of the
seedlings were placed into such inoculum for several minutes
previous to planting.

In plating, the nodules were first crushed in a sterile
mortar, under a glass cover using about 10 cc. of sterile

salt solution for from two to five scraled nodule tips. One

L7.

dilution only was needed, aid miis was usually made by pouring
a drOp or two of the mixed first tube, which itself had only

a few drops of crushed nodule juice, from the mortar, into a
second tube. Contaminations were few in the second dilution
which usually contained from fifty to two hundred colonies

of the causal organism whose description from prelininary
study follows.

2. Some Characters of the Causal Organism.

Korphologically, the organism in culture, is a minute rod
0.15 micron to 0.20 micron in diameter at first. This nay
increase to 0.50 micron or more at full growth. Its length
is from two to five or more microns, depending on age and me-
dium of growth. It shows motility, but flagella are difficult
to demonstrate. In fact, it is'douhtful whether true flagella
occur, as the polar projections nay be formed as the rods
break apart. Threads are formed, axial seem to branch in age,
tnus corresponding with fresh threads in the growing nodule
tip where septa or disjunctions of cells are hard to demon-
strate, and wnere the diameter say be greater and more vari-
able,-- the ends especially becoming thickened or club-shaped
at maturity. In vitro, these Cluhbed ends have rarely been
and branching is else not connon. From agar colonies
the younger cells are sctewhat gran positive, hut in vivo
they are strongly so.

The colonies are roughly circular and sliyhtly raised at
the edges, especially on dilute Ledia; while on strong nedia
they beco e almost hemispherical, and change from the first

Opalescent bluish tinge to a dark brown color in eight to

38.

ten daJs. Occasionally they are associated intiuately with a
filamentous fungus, whose threads they follow and with which
gruuous root-lire processes are forged, loOXing like gnarled
roots. Zhese benetrate the agar or they radiate out on the
surface, forming star-line surface colonies. Once it was
found associated with a yeast, but this conbinaticn could not
successfully be transferred.*

bone phisiological characters were examined in the follow-
ing engeriments:

a. Actions on Starches.

rne materials used in the starch reaction tests were:

rotate Starch, Corn starch, and Ceanothus module Starch.

 

These Were placed in $0 cc. Erlenmeyer flasks and after the
addition of 15 cc. of heavy inoculum and three per cent. tol—
uol, were incubated at 880 c. for ten days with the follow-
ing results;

(1). Elask with potato starch and boiled inoculum, no
action.

(2) Flask with potato starch and inoculum, almost complete
solution of all grains.

(3) Flask with corr starch and sterile inoculum, no action.

(4) Flask with corn starch and inoculum, very decided dis-
integration of most grains, due to a centripetal action re-

sembling the drillinr of nary holes toward.tie center of the grain.

Q

* It was felt that more extensive bacteriological studies of
the causal organism than are here given, should not be permitted
to delay the completion of this thesis. such studies are, how-
ever, being made for future publication.

(O

f ,3

(5) Ceanothus nodule starch with sterile inoculum gave

 

no result.

(6) Ceaiothus nodule starch with noculum s1.:.owed tvro times

by *4

 

O

of solution; ne a hollowing-out of the grains from one point
which was also noted in the potato starch eXperihent, the oth-
er action was one of equal solution over the whole surface of
thegnfin.
(7) Raw filtered nodule juice alone seemed to act somewhat
less on the nodule starch than when the inoculum was added.
b. helation to Atmospheric NitroSen.
bince Bottomley (4) had claimed nitrogen fixation with a
limited series of tests on liquid media and since the organism
here described seemed to thrive far better on solid media,
fourteen different cultures of known composition (the controls
give almost exactly the sums of ingredients) were set up as
follows:
Kutrient Glycerin Agar (surface inoculated):

(1) Three flasks of the pure organism.

(2) Three flasks of the organism inoculated with a combin-
ation of the organism and a fungus, the commoner of
the two previously mentioned.

Coons' Synthetic Agar (surface inoculated):

(b) Two flasks of the pure organism.

(4) Two flasks of the above-mentioned combination.

(5) Ewe checxs of each combination (i.e., two of Lutrient

agar and two of Coons' synthetic) were also run.

40.

These had an equal amount (5 cc.) of sterilized
inoculum added.

She 150 cc. agar used in each flask was solidified asainst
the sides of 600 cc. Zjeldahl flasks, by Spinning them in a
Vertical position in a centrifuge. This greatly increased
the surface area for the aerobic growth of the organism.
After fifteen days' growth they were subjected to the modi-
fied Ejeldahl analysis as used by Bristol and lags (6) with
the following results (faole VI):

Eere somewhat larger gains are shown in the nutrient glycer-
in than were found by Lottomley (4) who used loO cc. yer cul-
as a cu'nst _

ture flask, ' ' 150 cc. here used, the surface Using
much greater, and :robably constituting a factor in synthetic
activity. Ihese mains are also larger than those feund by
Lfihnis and Eillai (14) for Azotobacter in a glycerol medium.

The associated funjus seems either to reduce the action
of the bacterium on gaseous nitrogen or else to possess some
activity which liberates nitrogen in some gaseous form from
the organic nitrogen yresent. ho provision was made in these
exnerinents to retrieve any combined nitrogen lost in this
manner, nor to detect the reduction of organic nitrogen, hence
it may not be hnown from these exocrinents alone whether one
or both of these actions took place.

IyLDhe Coons' Agar which had only about one-fourth as much
nitroden, the gains lie too near the ergerinental error to be
significant. She acidity of tne media was between 6.7 and
6.8 93 attnmabeginning<flTtim engerinentzuuiied.changed

only slightly at the end.

41.
Cable VI.

‘3’: 3.9.... _' . ‘4. -‘ , I .-‘- ’~ ,.V-: ..
nitroten fination by the Jew sinus Lrtgfllsfl

... _-—- ~~

 

 

.- ...-—

.. . i , Nitrogen in :Gain over Aver-
'.‘ A»~.1s1 \-l ‘1 ' Q C ’ a C O, _
“8““ t1“- W-ltme - 001mOld-Culturefimis: age 0 Controls.

 

 

nutrient Glycerin : : :
.5358 I‘ :

Plus: 1. Pure cul- : : 0.0596 : 0.0045
ture

Flssx 2. lure cul- : : 0.0596 : 0.0045
ture

Flask 0. Pure cul- : : 0.0605 : 0.0050
ture

Blues 4. fwo organ-z : 0.0507 : 0.0054
ib‘n‘xs

ilasx 5. Duo organ-z : 0.0565 : 0.0055
isns

Flasn 6. fwo organ-z : 0.0590 — : 0.0007
isns

Elasx 7. : 0.0552 : :

Flash 5. : 0.0574 : :
Average : 0.0555 : :

Coons' Synthetic : : :

Agar

Elasx 1. Bure cul- : : 0.0159 : 0.0006
ture

FlaSK 2. Eure cul- : : 0.0158 : 0.0005
ture

FlaSA 5. inc organ-z : 0.0159 : 0.0006
isns -

Elasx 4. Two organ-z : 0.0159 : 0.0006

0
v , w
l blAk b

i
.J

F.
H F-

S

f

F .

C

O

O

O
O 0

FJ

L“

N

.0

{‘1
U:
k
n
G:
O
O.
C
I
F’
C“!
,5

Average : 0.6155

 

In digesting these cultures it Was noted that the controls
were much more resistant to combustion thafl the cultures of

either the organisms alone or of two organisus together, the

‘xb.
digestion of the controls requiring about two nours longer,
and about 25 cc. more of tne digesting acid. Inis seems to
indicate cOnsiderable disintegrating Power of the causal or-
ganism as it grows on agar.

r

0. Infection of Roots of Ceunotnus and froduction of Hodules.

 

a. hethods.

Plants whose roots are to he tested by infection must be
sterile throughout. Such plants were produced as follows:

Tall culture dishes were prepared as germinators, by adding
a layer of sand, previously digested with hydrochloric acid
and thoroughly washed (free of chlorides). To this was added
a sterile salt solution originally composed of sodium acid
phosphate, calcium chloride, and magnesium sulphate in equi-
molecular probortions, the total having an h/léb concentra-
tion with no iron gresent. These were sterilized at fifteen
pounds' pressure in tne autoclave. needs WLlCh had been car-
bonized with strong sulfuric acid were then, under aseptic
precautions, planted in these germinators at distances of one-
half inch or more from each other. Bone seeds were internally
infected with fungi by an insect which deposits its eggs Witn-
in the seed, but while these seeds would not sprout, yet, since
the carbonized seed coat furnished a good medium for growth,
the fungi would emerge, eporulate and become a menace to the
sterile plantlets. The latter were tested out on nutrient
agar till the sterile ones could he certainly judged by appear-
ance.. The sterile seedlings were then planted into sterile

culture tubes or dishes having a nutrient solution of the

48.
same composition as that of the germinators except that each
culture (except in the earliest tests) had added to it enough
calcium carbonate of M/loo solution to bring the total calci-
um 00ncentration to h/4OU strength, and enough ferrous sul-
phate to bring its own final concentration to h/eooo or
h/lo,OOO. At first greater dilutions of iron were used but
seedlings becane etiolated after four or five weehs' growth.
The calcium.which was at first added in the form of carbonate,
although present in fine suspension did not enable the plants
to thrive due, perhaps, to the toxic effect of the magnesium
Which apparently could not be balanced by the ample amount of
calcium present. Later, when the calcium was introduced, most-
ly as chloride, the difficulty disappeared.

The seedlings were at first tranSplanted with a strong trans-
fer needle, but this was too hazardous. A Special transplant-
ing forceps was designed and made of nichrome wire in such a
way that its jaws could not crush the tender plantlets, but
would still hold them finhly. ho plants were lost by drOpping
or brearing of‘roots after these forceps were used, and the
element of time was greatly shortened thus making aseptic con-
ditions more certain.

After transplanting, the roots were flooded with 5 to 10 cc.
of the inoculum, prepared in sterile solution as previously
described under the head of isolation methods, but during the
testings of the first isolations the rootlets were submerged
in the inoculum several minutes before transplanting.

The numbers of inoculation cultures with the causal organism,

and the results obtained are as follows~

44.

b. Inoculation Cultures.

About fifteen different organisms Which were isolated dur-
ing the year 1924-25 were tested out in tube cultures without
results. That these organisms did not possess the pr0per mor-
phology was of course known at the beginning. They also pre-
sented no constant picture on the poured plates, but it was
felt that the morphology of the causal organism might be dif—
ferent in vitro from that in vivo, and that it might appear
in isolation plates only sporadically due to the possibility
that the media used might not be suitable to its growth, or
that it might have a dormant stage. It is not here considered
feasible either to describe these non-infecting organisms or
their cultures (usually consisting of six tubes with two checks
each).

The cultures in Which infection took place and in which
some nodules were produced were the following:

In large test tubes (57 mm. wide and.500 mm. long) nine
cultures (two controls) were made of tne pure culture and nine
(two controls) of the organism associated with a fungus (as
previously described under the physiolOgy of tne organism).
These tubes had 90 grams of acid-washed sand and 35 cc. of the
non-nitroeenous solutions. Growth was always good in the
enecas. A similar set was also furnished with calcium nitrate
instead of chloride and carbonate as described above. These
made especially good growth and as compared with the growth of
wild plaits under competition, tneir growth was phenomenal,

indicating that these salts with the sand are not a handicap.

 

45.

In capsules which had been used nostly as checks in the
Special testing of previous organisms, the same prOportions
of sand and non-nitrate solution tests were also carried out.
These contained older plants which were growing slowly and
consisted of the following numbers:

12 nodulate cultures (all but three finally sterile; i.e.,
no other organisms present), of which six had the pure culture
inoculation and six the combination inoculation of fungus and
causal organism;

10 non-nodulate controls (eight originally and finally
sterile), wnich had sterilized inoculum added, to equalize
tne nitr0gen content of the culture;

Finally, thirty-four pot cultures were set up having three
plants each. The tube cultures all had one plant each, While
tne capsule cultures had from one to three plants each. The

soil in the pots was not sterilized but was carefully taken

 

from sandy subsoil several miles from any Qeanothus habitat.
Of these cultures, eighteen were inoculated and sixteen were
used as controls. Half of the inoculated pots were supplied
with pure culture inoculum and half with the combination in-
oculum, as previously described.

Dhe results (to June 10th) are as follows:

fwo of the pure culture tubes produced nodules while sev-
eral others have infections. hot all have been completely
examin 8 d 0

all inoculated capsule cultures were infected and of the
twelve, nine produced large nodules, five of WLlCh were in

pure culture (21. IV) and finally free from other organisms,

4e.
wnile of tne four which had had the combination inoculum in-
troduced, two were finally flee from other organisms and two
were not free, having become contaminated by an alga.

Ten uninoculated checns were all sterile; i.e., they had
no infections and no nodules.

It Seems that nodule growth in sand cultures is very slow,
and that the seedlings in tube cultures are too young to have
develOped nodules in tne four weers' growth Miich they have
made. This is borne out by the fact that in a s ecial series
not above detailed, nodules were formed in four weeks in every
case (fifteen in number) where soil infusion fr m heavily in-
oculated habitat soil was used upon sterile garden soil. All
controls of raw or sterilized soils examined (twenty in number)
nave been.fbund free of infections and nodules.

Platings made from the inoculated sand cultures showed that
tne organisM‘was present in very shall numbers, indicating that
it does not thrive under these conditions. Since it is of im-
portance to re-isolate from the nodules apparently produced by
pure culture inoculation, tne following trials were made:

4. Re-isolation of the Organism.

a. Four large nodules (Pl. IV) were carefully scraped and
oiushed as previously described under isolation of the organ-
ism. Twelve plates were poured including six dilutions. Both
nutrient glycerin and Coons' synthetic agar were used.

b. These platings in two days resulted in.LLn‘ snail colo-
nies having the peculiar blue color of the original pure cul-

tures. Their morphology and staining reactions are identical

47.
with the original from which the inoculum was Prepared. One
nodule from the combination inoculum was separately plated,
but it showed only the characters of those derived from pure
inoculum nodules, thus indicating that this fungus does not
survive, and that it is not, in sudh exnerinents, essential
in iniection.

Whether the previous passage through agar media had any
hastening effect on the growth of these colonies, or whether
the media were better adapted than the original media of iso-
lation of the same formula, is not known; but they develoned
in about half the tiam, or at twice the rate of the original
colonies.

Dnus it is Clear that the organism.here isolated, whatever
may be its final classification, has the ability to infect

Ceanothus americanus under suitable conditions, and that it

 

can be re-isolated from uncontaminated cultures, and from the
nouules produced by infection with a pure culture.
General Discussion.

as was stated in the historical part, no correlative eco-
logical data on growth, as related to nodulation, had been
given by a previous writer. The determination of the age,
and if possible, the annual growths of the nodules had to be
made before accurate correlation with general growth could be
made. Year by year, age determination was also necessary,
inasmuch as the amount of nodule tissue is not always prepor-
tional to the age of the plant for a given number of nodules

and in a given soil. A plant at first lightly nodulate may

48.
become heavily nodulate after it is several years old and yet
not nake up in growth sufficiently to correlate accurately
with its numbers of nodules and their anount of nodule tissue.
When such exceptiOnal cases are prOperly excluded by the ac-
curate determination of annual growths and age of nodules,
the tabulations show great regularity and strong correlation
of general growth with nodule growth, as presented in Tables
I, II, and III. The more accurate dry weights of greenhouse
seedlings bear out the relationShip found in tne field, and
in addition the total nitrogen of greenhouse plants gives
SOme indication that a nitrogen assimilating function in the
nodules is concerned. The uniformly small size and low nitro-
gen of non-nodulate plants rules out heredity, and makes a
Special nitrogen avidity of roots for soil-nitrogen less
probable.

The sprouting of nodulate versus denodulate plants rein-
forces this indication, but inasmuch as the root systems in
sore pairs were heavily pruned and did not again become normal
in extent, eitner a stimulating action upon the roots is in-
dicated, whereby greater ability to remove nitrogen from the
soil is acquired, or, what is more probable, the youth of
roots makes them more efficient in absorption. These consid-
erations make a definite decision as to the exact total action
of nonulation upon general growth impossible.

The preliminary studies on the morphology and physiology
of tne causal organism, make further discussion of its taxonomy

unprofitable till more details are available, but the tendency

49.

to accumulate nitrOgen in vitro lends very strong if not final
support to the theory that the nodules are the seat of some
nitrogen accumulation. however, even if this should be proven
by the later analysis of nodulate versus non-nodulate plants
as grown in sand cultures (Where no combined nitrogen is pres-
ent in nutrient solution), the swecial stimulation to general
growth, by Shall amounts of nitrogenous or non-nitrOgenous
bodies, made only in the nodules or in enlarged infections,
is not excluded.

The following principal conclusions may be drawn from the
observational and eXperiaental data presented:

Conclusions.

1. A strong correlation between growth.and amount of root

 

nOdulation is shown in Ceanothus-anericanus by field (wild)
plants of different ages and by greenhouse seedlings of one
season's growth.

2. The poorest growth of non-nodulate seedlings was made
in.tLe poorest soil while good growth was made in tne nitro-
gen ricn soil. The growth of the latter was as heavy as that
made by very nodulate plants in poor soil and conversely, indi-
cates that there is either a nitrogen accumulation or a root
stimulation, or both, due to the nodules, which is eouivalent
to the adVantage accruing from a nitrogen fertile soil.

3. Dormant plants brought into the greenhouse and forced
into Sprouting showed earlier Sprouting where nodules were
abundant; somewnat later sprouting, wnere few in number; and

retarded Sprouting where nodules were removed.

50

4. Ine causal organisdlisolated and prelininarily de-
scribed is able to dissolve nodule and other starches. It is
also able to cause tne fixation of very shall alounts of ni-
trogen in vitro, which supports the theory that the nodules
are the seat of nitrogen accumulation.

5. The causal organism was re-isolated and its morpholog-
ical characters were found to have remained constant in vitro.
These Characters varied s0newhat from its morphology within
the nodules.

6. Ceanothus seedlings made very good growth in sterile

 

conditions, if supplied with certain concentrations of salts.
Those supplied with nitrate and sufficient iron Lade the most
rapid growth.

7. The general conclusion follows that a nitrogen assini-
lation function is rendered far more probable by these ob-
servations, but another special stimulation to growth by the
root nodules is not thereby ruled out, and both functiOns may

exist within the nOdules of Ceanothus species.

 

ACKLJOWLEIJGI :31.” :38

The writer is deeply indebted to Doctor :.a. Bessey,
wnose Constant kindly interest, assistance, and advice
in all matters pertainiig to the advancement of this
problem, made possible the completion of tnis thesis.

To Doctor R. E. Eibbard tne writer is much indebted
for critical advice and supervision in the physiological
phases. To Doctor R. U. Snyder the writer also owes
much for advice and encouragement in the bacteriological

pllb-S ES 0

VI. LITEnnfUnE CITED.

1. Arzberger, E.G. fhe fungous root tubercles of Ceanothus

 

americanus, Elaeagnus argentea, and Irrica cerifera.

 

 

 

ho.Bot.Gard.Rept. 21:60-1U3. 213. 6-14. Dec. 1910.

DD
0

Atninson, G.F. Ihe genus Franaia in tne United States.
Bu11.20rr.Bot. Club. 12:171-177. Pl. 128. 1892.

5. Real, Charles. Tubercles on Ceanothus americanus.

 

Bot. Gaz. ig: 232. 1890.

4. Bottomley, W.B. The root nodules of geanothus americanus.

 

ann. Bot. 23:605-611. P1. 28. 1915.

5. Bristol, B.M. and 8.3. Page. A critical enquiry into the
alleged fixation of nitrogen by green algae. Annals of
Applied biology'lgjb78-408. Pls. 20-21. 1923.

6. burrill, 2.J. and Hansen R. Is sgmbiosis p ssible between

legume bacteria and non-legume plants? Univ. of 111., Agric.
Eng. Sta. bull. 202. Bls. 1-17. July 19l7.

7. Coons, G. n. Factors involved in the growth and the pyc-

nidium fermation of Plenodomus fuscomaculans. Journ. Agr.

 

Res. §:757. 1918.

8. Hiltner, L. Ueber die bedeutung der Hurzelknollchen von
nlnus glutinosa fur die Stickstoffernanrung dieser Pflanze.
Landw. Vers. Stat. 38:155-161. 1898.

9. thougal, W.B- On the mycorhizas of forest trees.

Am. Jouro BOto 1:51-74. P180 4-6. 19140

10.

ll.

12.

14.

Wiley, E.h. (Editor). Official and Irovisional Lethods
of Analysis. U.L. Deyt. Agric. Bur. Chem., Bull. 107
(revised). 1910.

Hobbs, F. und Eiltner, L. Die endotroPhe Mykorrhiza
von Podocarpus and ihre physiologische Bedeutung.
Landw. Vers. Stat. 1:415-425. 1898.

retry, 2.3. Germination and growth of seedlings of

Ceanothus americanus as afiected by heated soils.

 

Kich. Acad. Sci. Rept. 22:135. 1920.

Spratt, E.R. A comparative account of the root nodules
of the Leguminosae. Ann.Bot. ggtlB9-200. 218. 13-14.
1919.

Lohnis, F. and Pillai, n.K. Ueber Stickstofffixierende
Bakterien III. Centr. Baht. Par., Abt. II, 39:781-799.
1908.

EXP AHATIOH OF ILATES

Plate I. Modules of different ages, showing branching and
positions relative to roots.

Figs. 1 and 27 represent normal roots.

Figs. 2, 5, 6, and 7, and 54 represent common nodules one
year old or less.

Figs. 4, 5, 9, 10, l2, 14, 19, 20, 26, and 28 represent
nodules from 2 to 5 years old.

Fig. 8 represents a tetrachotomous nodule.

Fig. 9 represents a branched nodule, one branch presenting
a pentachotomous branching.

Figs. 11, 15, 19, 31, b2, and 55 represent nodules in
positions which are unusual for roots.

Fig. 11 shows a fasciated nodule extending over a whole
infection.

Figs. 15, 16, 18, 21, and 25 represent very young nodules,
Figs. 17, 21, 22, 2b, 24, 29, and 51 represent infections
wnich can develop into visible nodules within a short

time. ‘
Figs. 20, 28, and 55 represent trichotomies.

Plate 11. Upper part of Plate represents fruiting tOps of
two plants, 8 years old from habitat 1.

Fig. 1 represents a heavily nodulate plant which had 4
large nodules near the crown and several other smaller
ones near the upper part of the prinany root.

Fig. 2 represents a very large plant with 11 very large
nodules on the primary root and its large branches near
the crown. Over a score of medium-sized nodules were at-
tached to smaller roots near the Larger nodules. The wood
of this plant is over twice that of plant in Fig. 1.
Lower part of Plate represents part of sub-series b, tOp
and root pruned and forced to shoot in the greenhouse.
Kembers in each of these 5 pairs were equal in size, nod-
ulation, and age. The minus member of each pair had its
nodules removed when potted. Sprouts shown were developed
in three weeks. Fairs 2 and 7 were very heavily nodulate.
These plants range from 5 to 6 years old.

Plate III. Seedlings l and 2 years old, as dug from habitat i
2, plot 5, fertile soil (Phot. 1 x). , |

Figs. 1 and 5 are poorly nodulate plants two years old.

Fig. 2, a two-year old heavily nodulate plant which grew
within a few feet of those in rigs. 1 and b.

Fig. a moderately heavily nodulate one-year old plant,
nearly as large as two-year old plants in Figs. 1 and 2.

Figs. 5 and 6, practically non-nodulate plants 1 year old.

bk) 0

Figs. 5 and 6, practically non-nodulate plants one year
old. Fi5. 6 and two minute nodules near tne distal end
of'the primary root.

Plate IV. seedlings 5rown in sterile sand culture.

Middle plant from two organise inoculum. Small plant has
nodule on dead tap root. (Circular spots on stem of right
hand plant and among leaves are bubbles. Glass rods were
nec ssary to Keep some roots below the surface of the Water
in the dish). Some of these nodules are more branched than
those naturally produced.

Plate V. hhole, one- and two-year seedlings of moderate nodu-
lation forced 5 weeks in greenhouse (Sub-series a). Y—mbers
of pairs were almost identical in size, shape, an. amount
of nouulation, but pairs Varied in size and shape. The
“minus"plunt of each pair was denodulated. The differences
were sli5nt, but easily measurable and they were consistent.
(Ihot. 1/5 XJ.

Ilate VI. A series of Iec nnous seedlinws inclutln5 nodulate

cnld innd- nodiLLate ifilah.c,..lthlni tLK: sane ;»ots. 3:12 e [llnlts
.Lll «0 Lu ‘1. , 'u, 7 , O ’ (.5 , .LU ,_L_L, £,.l’lC; .Lg 5.330 llL'x “"”-"—-.. llpulllet‘ yOt 6
nub SICK”; LAID-{Age CU ...J' J .. e ; Lani 4.5-5- ’(3 ‘1. ......L... Int.) 14;, (A. ... "Q‘i'v ll\J‘\L1L .LC 8 o
unallest plants in pots 5,d,6,7, NJQ lo hwd practically no

nodules after 5 months' Lrowtn. Intermediate plants had, as
a rule, intermediate nodulation. (Inot. 1/5 I).

rlate VII. Ihree ilants from pot 9 shown in Ilate VI. These
illustr—te 5 of tne Nei5ned groups of Sable IV,-- Fig. 1 rep-
res outing Group VII I; El5. 2, Group VII; and Elg. 5 Group
VI; (Pnot. 2/0 X).

Ilate VIII. Root systems of plants shown in pot 2, Ilate VI,
(2/5 x). Plant 1 at left has sever; youn5 but lar5e nodules
indicating late infection. Plant 2 has a few small youn5
nodules. Plant 5 has no easily visible nodules, but it has
many old infection , shown in detail (2K in Plate IX. This
anomalous condition was found in a relatively snbll number
of plants (Group I, Table IV). This group of plants made as
500d growth as the youn5 heavily nodula te plants.

Plate II. Part of root system of plant 5, Plate VIII, showing
many infections LUd incipient nodules (namnified 2 x). In-
fections and nodules are indicated a I. At Io a young root
has penetrated an infection.

Ilate L. Lore detailed (4:) photograph of nodules shown (“V)
in Ilate IV.

"N r . o _‘_ _ _ . 4:" “ _, _ r‘ a - ‘ _, w .-.. 1‘ _ ..‘ J. .3 ,v .C " s, 1'
£15138 fill. 131:": b€l‘( top ll.»,11_1‘s ) alto est-AL IL. ( .J Dru-'1'". J—l~-‘~1 9)
placement 0; sterile sand cult res.

I

P LA TE

 

Plate II.

 

 

 

“l¢*d II:

 

 

l! \‘

'5'." ,\ \
' a " ‘
. \ .tt'eééfi!’

g
Y‘ ‘

 

 

PLATE V

 

 

PLATE V I

 

PLATE VI I

 

VIII

PLATE

20. UN. ”4:

:.z.:..'E.=.. .. ,r__‘ If;

 

PLATE IX

 

PL!) T“ X.

 

 

 

 

 

V v
:2 *5?
ad '

”.-
I

. f.

9b...

.t. ‘4

.
‘0 .u
A.?O-..Ȣ.~5 .. . . .
‘. . , -
WRQNJA§_.;3 _ . _ . .
o... QEKP . u {5.07 . . ... . . , . . . . .
3.. s, .0 . 7 w. . ... . . ,. . . .‘ I . . . . . ‘ o.oVI<o.11‘~.¢.¢....cnn‘.
.. .1 . . , h... xwtusduv. .051... .90. ......uuvvk}
. , .

I...

 

I
I
'l
I
III
I
I
I'll
III
I
III
I
III
I
II
I
l
l
I
I!
I

IllllllllllllHllHlllWlIll

03174 4737

3 1293