Quality in celery as related to structure

LIBRARY OF THE
UNIVERSITY OF ILLINOIS
AT URBANA-CHAMPAICN

N.O.IM CIRCULATING

CHECK FOR UNBOUND
CIRCULATING COPY

UNIVERSITY OF ILLINOIS

Agricultural Experiment Station

BULLETIN No. 33G

QUALITY IN CELERY AS RELATED
TO STRUCTURE

BY C. B. SAYRE

URBAXA, ILLINOIS, Al'Cl'ST, 1929

CONTENTS

PAGE

REVIEW OF LITERATURE 559

STRUCTURE ()], THE CKLKRY PLANT 561

Kinds of Skeletal Tissues in Plants 563

EKiHT VARIETIES TESTED UNDER VARYING CONDITIONS OF

MOISTURE AND BLANCHING 566

COMPARISON OF VARIOUS SPECIMENS FOR QUALITY 566

Comparison of Varieties 567

LABORATORY TECHNIC 571

RESULTS OF MICROSCOPIC EXAMINATIONS 573

Mierochemical Tests 5S5

SUMMARY AND CONCLUSIONS 5S5

LITERATURE CITED.. 5SS

QUALITY IN CELERY AS RELATED
TO STRUCTURE1

BY CHARLES B. SAYKE, Formerly Assistant Chief in Olericulture

Celery to be of good quality must be crisp, tender, and string-
less, having a pleasant, sweet, nut-like flavor. Systematic improve-
ment in the quality of celery requires: first, a knowledge of the
structural characters of the plant that are correlated with quality
or lack of quality; and second, a knowledge of the conditions of tem-
perature, moisture, plant food, maturity, blanching, and varietal dif-
ferences that affect the structure of celery and its resulting relation
to quality. It is the object of this bulletin to present data and infor-
mation concerning these topics, particularly in regard to a micro-
scopic study of the interior anatomy and structural characters of
the celery plant that are correlated with quality.

REVIEW OF LITERATURE

Norton-"'* says that quality in celery is primarily a question of the
proportional relationships of the so-called parenchymous tissue to the
fibrous tissue. This is determined in part by varietal differences and,
within varietal limits, in the main by certain external conditions,
such as soil moisture, plant food, rate of growth, light, warmth, pest
control, and blanching.

Mills4* says that high quality in celery is based on its flavor,
crispness, and lack of stringiness. Flavor, he believes, is associated
in some way with the chlorophyll in the stalks. It is affected by the
rate of blanching and the type of celery, whether of the self-blanching
or green type. It is generally conceded that green varieties of celery
are of better flavor than varieties of the self-blanching type.

The degree of crispness, he says, is due entirely to the relative
water content of the plant. AYhe-n the cells are turgid the stalk will
be crisp, but when the parenchyma cells are not well filled with
liquid the stalk will be wilted and lacking in crispness. This lack
of crispness may be caused by too high a temperature, too much or
too little moisture, too little nitrogen in the fertilizer, checking of
growth dtie to disease, or by too rapid a growth which in turn ma-
lures the crop too early in the season, causing pithiness.

Pithiness. Mills believes, is caused by the maturing of the stalk
and depends partially ut least upon heredity, some strains of celery
tending to develop this trouble more than others. A pithy stalk

5()() BULI.KTIX Xo. 336 [August,

cannot be made crisp because the parenchyma cells have collapsed
and cannot be filled with water again.

Conditions of storage. Thompson"* says, may also affect crisp-
ness. Insufficient moisture in storage will cause wilting and lack
of crispness. Too high a. temperature in storage will cause rapid
maturity and result in pithiness.

According to Mills and many other writers, stringiness in celery
is due to the development of excessive woody tissue in and around
the fibrovascular bundles. Mills claims that in stringy celery these
bundles will be found enlarged and the woody cells will extend out
into the spaces where the soft-tissue cells should be present, often
extending out to the epidermis. Among the catises of this devel-
opment of excessive woody tissue he mentions strong light and too
little light: a check in growth owing to too much moisture, too little
fertilizer, or disease: banking too long; or an overrapid growth, caus-
ing a woody condition.

The writer cannot agree with Mills' statement that stringiness
of celery is due to an enlargement of the fibrovascular bundles, or
that woody cells extend out into the spaces where soft tisstie should
be present. The present paper includes data and microphotographs
which the author believes demonstrate that stringiness in celery is
due to other structural changes or differences.

White and Sandsten1'1' and also Austin and White1* ascribe pithi-
ness in celery to poor seed, especially to seed produced on pithy plants.
Their experiments indicate that pithiness is an inherited character
and they conclude that a rogueing out of the seed plot of all plants
showing pithiness will obviate this undesirable character.

Beattie2" says that pithiness in celery may frequently be prevented
by proper cultural conditions. Pithiness may be caused by too rank
a growth in the seed bed and also by a severe check in growth. An-
other cause of pithiness, he says, is in leaving the celery in the field
too long after it has become blanched. The outer stems in particular
tend to become pithy under these conditions.

Flavor in celery is affected by a number of conditions, Beattie
says. The strong, rank flavor of celery is caused by the presence
of chlorophyll and this is overcome by the blanching process. Blanch-
ing destroys the chlorophyll in the stems that are already grown and
prevents the formation of chlorophyll in the stems that are produced
during the blanching period. Flavor, he adds, is also influenced by
the variety. The green types are in general superior to the self-
blanching types.

There is a difference of opinion among writers as to the effect
of the soil on the quality of the celery it will produce. Watts111* says
that celery grown on muck soil is inferior in quality to celery grown
on upland soil. Thompson"* believes that muck soil will produce
celery of superior quality to that grown on mineral or upland soils.

Qr.u.iTY ix OKLKHY AS RKI.ATKD TO STIUVTI'KK 501

He adds that, to be of the best quality, celery must have a con-
tinuous growth, and this is more likely to occur on muck soil than
on mineral soils owing- to the better physical condition and greater
water-holding capacity of the muck soils.

STRUCTURE OF THE CELERY PLANT

The celery plant has a large fibrous root system and a very short
ste.m. so short- that the leaf petioles appear to arise at the base of
the roots. The leaves are compound and have long, thick petioles.

S&V'V^

>•- V, .• .-- , /

v"*y"^ /* ' "^js*^ * "•. X-J

.^&:?/^>u:V:

/5fe^!».>^>v_ '< ?>. V- -> --

YK;. 1. — CROSS-SECTIOX OF A POHTION OF CEI.EHY STALK
A, epidermis; i5. colleiichyinn in rib; C, paronchynia ; ]). phloem;
K, xylem (D and E together form the primary iibrovascular bundle) ;
F, secondary fibrovascular bundle.

The leaf petiole is the edible portion of the plant. There arc about
twenty leaves per plant, some varieties producing more leaves than
other varieties.

In cross-section the leaf petiole is somewhat crescent shaped.
The exterior side is somewhat ribbed or corrugated and the inner
side is smooth. The proportional thickness of the petiole varies with
different varieties, some varieties having exceptionally thick leaf-
stalks. A microscopic examination of a section of the celery leafstalk
shows that it is composed of a number of different kinds of tissues
(Fig. ll.

Entirely surrounding the section, there is an outer layer of tissue,
one cell in depth. This is the epidermis and is composed of cells

;,(i_> Bru.KTiN Xo. 336

which are somewhat thickened, particularly on the outer side. In
addition, a waxy substance called cutin is formed on the outer sur-
face. This forms a thin, waterproof film called the cuticle. The
cutini/.ed parts are clearly distinguishable when treated with chloro-
iodid of zinc.

Inside the1 epidermis is the collenchyma tissue which occurs as
separate strands in the "ribs" or corrugated portion around the ex-
terior side of each stalk; and also as a continuous zone about two
cells in depth next to the epidermis. This layer of collenchyma is
often deeper on the interior side of each stalk. The collenchyma
cells are thick-walled especially at the angles. The depth of this
collenchyma layer and the thickness of the cell walls vary with the
different varieties of celerv.

_~-A

"*>r v

\ '
\t

C

- 'j .

Fi<;. 2. — CHOSS-SECTIOX or CKI.ERY PETIOLE SHOWING TYPICAL ARRANGE-
MENT OF THE PRIMARY AND SECONDARY FIBROVASCULAR BUNDLES

A. collenchyma .strands in ribs; B. primary fibrovascular bundles;
('. secondary fibrovasular bundles.

The collenchyma strand in a rib is shown in Fig. 1. Arranged
as it is near the periphery, the collenchyma embodies the mechanical
structure that gives the greatest resistance or strength against bend-
ing stresses.

Thin-walled parenchyma tissue lies next to the collenchyma in-
ternally and constitutes the greater part of the stalk structure. The
cells in this tissue are large, irregular, and thin walled, as shown
in I-i'j;. 1. The parenchyma forms the more tender tissue in the stem.

•\t more or less regular intervals there is a line of bundles of
thick-walled cells. These are the fibrovascular bundles. There is one
line o] rather laruv bundles which are situated about one-third of
the distance in from the exterior side of the stalk. There is likewise

l'.<?:~i] Qr.M.rrv ix CKI.F.HY AS KKI.ATKD TO Sinn TTKK 503

a secondary line of a large number of small bundles located parallel
to and about one-sixth of the distance in from the interior side
of the stalk. A typical arrangement of the primary and secondary
bundle- is illustrated in Fig. 2, which shows an entire cross-section
of a celery petiole.

The fibrovascular bundles consist of three1 parts, an outer phloem,
an inner part or xylem. and a median layer of cambium. The cambium
layer is the cell-producing tissue. On the exterior side the cells produced
add to the phloem, and on the interior side of the cambium the cells
formed add to the xylem.

Since a vascular bundle is in part composed of very delicate cells,
it apparently should be protected by some coarser and more resistant
tissue, and hence tends to associate itself with mechanical strands.
A gutter-shaped >heath of thick-walled cells (the phloem sheath)
acts like a splint behind which the delicate parts, particularly the
cambium, of the vascular bundle find necessary shelter.

The phloem consists of three elements, the sieve tubes, companion
cells, and the phloem parenchyma. The walls of these cells remain
cellulose. The cell walls of the phloem sheath are thickened.

In the xylem are found the tracheal vessels and xylem parenchy-
ma. Each tracheal tube is formed by the absorption of the end or
transverse walls in a vertical row of cells and. at the same time.
there is an enlargement in all dimensions of the cells composing the
tube. An unequal thickening and lignification of the vertical walls
of these tracheal tubes then occurs. The thick places in the walls
strengthen the tube and the thin places make an easy passage for
water and materials in solution. The thickened places in the walls
are in the form of rings or of a spiral coil. The ring- or hoop-shaped
thickening makes what is called an annular tracheal tube. The tubes
that have spiral thickening are called spiral tracheal tubes. Both
kinds of tracheal tubes can be seen in Fig. 3.

The tracheal tubes are comparatively large in diameter, as shown
in Fin'. 1. The large, thick-walled, oval-shaped elements in the xylem
cross-section are the tracheal tubes.

The row of small secondary vascular bundles seems to be quite
similar to the primary vascular bundles, except that the secondary
bundles are smaller and the phloem sheath is not so well developed.

Kinds of Skeletal Tissues in Plants

In the larger and more complex plants certain tissues supply
strength and rigidity and serve as a skeleton for the plant body
as a whole. Four kinds of plant skeletal tissues are recognized. These
are: ill the collenchyma, (2i the bast-fiber tissue. 1 3') the wood-
fiber tissue, and (4i the stone cell or stereid tissue. It is to be ex-
pected that one or more of these tissues would be found in the celery

X<>. 33G

]»lant and would give toughness and stringiness to it. Furthermore,
differences in I he development of these tissues might be expected to
he correlated with differences in quality or lack of quality in differ-
ent specimens of celery.

According to Stevens7" the collenchyma is the first kind of skeletal
tissue formed in plants. It appears in stems a short distance below
the growing apex before the bast and wood fibers and stone cells have
begun to be formed. Its chief characteristic is the existence of thick-
ened cell walls at the angles where three or four cells join. In most

Fl(i. 3. — LoXDITfDINAL SECTION OF CELERY SHOWING THE ANNULAR
AND SPIRAL TKACHKAI. TUBES IN THE DENSER TISSUE

OF THE XYLEM

Xoto the absence of any collenchyma strand at the upper
cdjjp. This absence of collenchyma is due to the fact that this
section was cut between the ribs.

plants the cell walls remain cellulose thruout. Collenchyma may occur
as a continuous zone or in separate strands.

Since the collenchyma is formed where growth is still taking place
it mu-t be capable of growing or of stretching. In this elongation of
the cell- the ends shove past each other and the points become dove-
tailed and interwoven as shown in Fig. 4. This splicing or inter-
lacing ot the cells in the collenchyma increases the strength of this
tissue.

Ilaberlandr* says, "Organs which are still undergoing elongation
must therefore make use of a mechanical tissue which is itself capable
of active extension by means of growth. The tissue which satisfies
these requirements- is collenchyma. The fact that collenchyma regtt-

/.V<1 QrAi.rry i.v (Y.i.KKY AS KKI.ATKD TO STHUCTVUE ,">(>.">

larly forms the skeletal system of growing organs does not preclude
it from serving as the permanent, mechanical tissue in many fully
grown herbaceous structures such as petioles." He further states that
"while bast fibres and wood fibres perform the task of strengthening
fully grown organs, neither of these types of mechanical element is
suited to the needs of young organs which are still growing in length."
This, of course, would be the case with celery.

The author wishes to call special attention to the foregoing para-
graph because it seems that in all the literature in regard to quality

]•"!<;. 4. — PORTION OF LONCI'ITDIXAL SKCTION rJ'iinr A CKI.KKY Km

Th(> ihick-wallcd. pointed cell tissue at the bottom is the collen-
chyina strand in the rib. Note how these cells arc pointed and inter-
laced as contrasted with the thin-walled, square-ended Tells ot the
parenchyma above.

in celery the collenchyma tissue has been overlooked or not taken
into consideration. Yet from the standpoint, of plant anatomy it seems
most logical to expect that the collenchyma would have a very im-
portant bearing in the texture and toughness of the celery petiole.

According to Stevens, modifications of the cells in plant skeletal
tissues may occur by ll) thickening of the cell walls; (2l chemical
alterations of the cell walls by the changing of the old material and
depositing of new material; (3) transformations in the physical con-
dition of the cell wall involving changes in its hardness and elasticity.

With a view to determining which, if any. of these three modifi-
cations occur or are correlated with differences in quality of celery,
very careful microscopic examinations and microchemical tests were
made of the structure of celery plants of varying degrees of quality.

Xo. 336

An effort was al>o made to determine what circumstances of culture
and environment cause these modifications in the celery plant.

EIGHT VARIETIES TESTED UNDER VARYING
CONDITIONS OF MOISTURE AND BLANCHING

This investigation included a study of eight different varieties
of celery and a. comparison of the effect of different environmental
factors on some of these varieties. The different cultural conditions
included a series of fertilizer treatments, a comparison of three
methods of blanching, and the effect of supplementary irrigation.

Specimens of each variety and from each different condition of
environment were carefully studied in an effort to ascertain what
factors or conditions affected the quality of the product. Some of the
leading commercial varieties were included. The self-blanching
varieties were Garrahan's Easy Blanching, a special strain of
this popular commercial variety, White Plume, and an ordinary
strain of Easy Blanching. Varieties of the green type were Giant
Pascal and Emperor. The pink varieties were Button's Giant Red.
Sutton's Superb Pink, and Rose Ribbed Self-Blanching.

Three different methods of blanching were used in order to de-
termine their effect on the quality of the celery. Both White Plume
and Garrahan's Easy Blanching were blanched by each method. The
plants were banked with earth, blanched with boards, and with com-
position paper (R. &: D. celery blanching paper).

Celery requires abundant moisture during the growing season.
Furthermore, succulence and tenderness in vegetables is often limited
by the moisture supply. A comparison was therefore made of the
effect of supplementary irrigation in growing each of the eight varie-
ties, in order to determine if extra moisture would improve the
quality even tho the celery was grown in a humid region in a season
of normal rainfall. The water was applied by means of a Skinner
overhead irrigation system. The irrigation was tised thruout the grow-
ing season whenever it was considered that a little extra water would
be beneficial to the plants.

COMPARISON OF VARIOUS SPECIMENS FOR QUALITY

Quality in any vegetable is somewhat indefinite and therefore diffi-
cult to measure accurately, especially in the finer degrees. In all of
the various state and federal standard grades for celery, stringiness
and flavor are left out of consideration. In fact, the only factors of
quality that are specified are pithiness and degree of blanching. In
order to provide a better basis for comparison of quality the author
has assumed certain additional characteristics as essential to good

in.>'.'\ QrAi.iTY IN CKI.KKY AS KKI.ATKD TO STUTC TruK ">(>7

quality. As stated on page 559. celery to be of good quality must be
crisp, tender, ami stringless, and have a pleasant, sweet, nut-like
flavor.

Accordingly, in order to compare the quality of the various speci-
mens of celery, observations were made of the texture, succulence,
and flavor of each specimen by an eating test. Texture was recorded
as ttti/ltr, tough, or xtringy. Succulence was recorded as succulent,
or /lithy. Flavor was reported as nut-like, sirctt, rank, or pungent.

Specimens were considered tender when the stalk could be broken
or bitten thru readily without any undue compression before the
tissues would break and without tearing apart in ragged strings.
Tough specimens required considerable pressure to break the epider-
mis and to masticate a portion of the stalk. Stringy specimens were
usually tough and had tough, string-like tissues that would tear out
readily and were much tougher than the remaining tissues. Specimens
we're termed pithy if they were of somewhat open texture with air
spaces or dry. spongy tissue in the central portion. Crisp celery is
both tender and succulent. By succulent is meant especially juicy.

The terms used in describing flavor may be defined as follows:
The specimen was termed pungent if it tasted "hot" with a tendency
to "bite" the tongue; rank was used when the flavor was strong and
grass-like without any sweetness; nut-like was used when the flavor
was rich and somewhat like a fine-flavored nut.

To a certain extent, especially in regard to minor differences in
flavor, personal preference would be a slight factor in a definite
quality rating. To insure greater accuracy in the comparisons, five
or more specimens from each lot were sampled and the average im-
pressions recorded. After the plants had been trimmed as for market,
three outer stalks were removed and the fourth sampled in particu-
lar for the quality tests. In the comparison of varieties the opinions
of six advanced students are considered with the author's in the
quality rating. In the other lots the rating is based entirely on the
opinion of the author.

Comparison of Varieties

Self-Blanching Type. Three' varieties or strains of the self-
blanching type were included in the comparisons. The rating of each
variety may be summed up as follows;

Garrahan's Eaxij Blanching. — Crisp and tender. In some cases
slightly stringy, but the strings are brittle and not tough or tenacious.
Flavor sweet and very palatable. This would be classed as a variety
of good quality and the best of the three varieties of the self-blanching
type.

Wliitc l^lunx .- -This variety is tough and stringy, the strings being
very tenacious. The flavor is somewhat rank and (mite pungent.

,-,(;<; BVI.LKTIX Xo. 336

Kverythiim- considered, this is a variety of very inferior quality. In
flavor it \vas the poorest of any of the varieties tested.

KdKH Klanchimj. — The ordinary strain of Easy Blanching was
similar to Garrahan's Easy Blanching but was a little more stringy
and less sweet than Garrahan's.

Green Type. The two varieties of the green type that were tested
were of excellent quality.

(Hunt Pascal. — This variety is very crisp and tender. Strings even
on the largest stalks arc not tough or tenacious but are brittle. The
ilavor is very palatable, being sweet and somewhat nut-like. This
variety seems to meet all of the requirements for high quality.

Emperor. — Emperor is another excellent variety. Like Giant Pas-
cal the stalks are thick and very crisp and tender, without tough
strings. The flavor is sweet and somewhat nutty. It is difficult to de-
cide which of these varieties ranks highest in quality. The opinion
of six students was that they are both of excellent quality, and the
choice was equally divided between the two varieties, but the author
has a slight preference for Giant Pascal.

Pink Type. There was a striking contrast in the quality of the
three varieties of the pink type that were tested. The two extremes,
highest quality and poorest quality, were included in this group.

Suit on' 's Giant Red. — This variety was very crisp and tender with
a total absence of tough strings. The flavor was especially palatable,
being sweet, rich, and nut-like. All who sampled this variety agreed
that it was of excellent quality in all respects.

button's Superb Pink. — This is another high quality variety, being
very crisp and tender, without tough strings, and having an especial-
ly palatable, sweet, nut-like flavor. However, in a careful compari-
son, the Giant Red was ranked as slightly superior in flavor and suc-
culence.

Pose Pibbed Self -Blanching. — In texture and succulence this vari-
ety showed the poorest quality of all varieties tested, but in flavor
it was superior to White Plume, which was the next poorest variety.
Rose Ribbed Self-Blanching is very tough and stringy. The strings
are especially tenacious. The flavor is fair, being slightly sweet.

Rating of Eight Varieties. Following is the author's rating of
these eight varieties on the basis of quality:

1. Sutton's (limit Red 5. Garrahan's Easy Blanching

2. Giant Pascal 6. Easy Blanching

3. Emperor 7. White Plume

4. Sut ton's Superb Pink 8. Rose Ribbed Self-Blanching

There is little choice between the first four varieties. All would
be classed as excellent in quality. Garrahan's Easy Blanching would
be classed as good in quality, and the last three would be rated poor.

7.''?.''] (Jr.xi.iTY IN CKLF.KV AS KKI.ATKI> 10 STIUVUKF. 509

It is worthy of noU' that the first lour varieties are of the slower
Maudlin^ types, two of them being green varieties and two j)ink
varieties. This is in accordance with the prevailing opinion that the
slower Main-hint: type is of somewhat superior quality to the self-
blanching type.

Effect of Fertilizer Treatments on Quality. Some Easy Blanching
celery that had been grown by another investigator under twenty-
two different fertilizer treatments was tested for quality. Five of the
best -looking celery plants from each fertilizer plot were carefully test-
ed and compared as to quality. It was hoped to find some indication as
to what influence, if any, the different kinds and amounts of plant
food might have on the quality of the celery. The cultural conditions
were the same on each plot except in regard to the fertilizer treat-
ment, so that any consistent difference in quality could be ascribed
to the effect of the plant food. The seedling plants were purchased
from a distant grower; they were in transit and arrived during one
of the hottest periods of the summer. The soil was dry and the
weather very warm and dry when the seedlings were transplanted
to thi' field plots. In addition, there were no facilities for irrigating
them.

The above conditions were, of course, exceedingly unfavorable
for the best development of the plants. Consequently the entire series
was of inferior quality. However, since the plants were available,
they were included in the writer's study of quality. Xo consistent
effect upon either texture or flavor could be detected as resulting from
any form or amount of plant food used in the experiment. It is the
belief of the author, however, that the unfavorable conditions under
which this crop was produced affected the crop so adversely that the
results were not at all indicative of the effects which plant food may
have upon the quality of celery.

Effect of Method of Blanching on Quality. To determine what
effect the method of blanching may have on the quality of celery,
(iarrahan's Easy Blanching and White Plume were subjected to three
different methods. Adjacent sections of the same row of each variety
were differently blanched. The cultural conditions were identical ex-
cept for the blanching. All the celery was harvested at the same time.
Following is a record of the quality tests of the blanched product:

( rdrt'dttdn'x E<i*}i Blanching. — 111 Blanched With Boards. Flavor
excellent. Sweet and palatable. Crisp, tender texture. A few strings
but they were brittle and not tough and tenacious. (2i Bl(inch<'/l With
/i*. >.(• I). Blunchiny P(ip<r. Flavor rather flat, but slightly sweet. Some-
what tough and stringy in texture. (3i BUtiichal \\'itfi So/7. Flavor
very flat and tasteless, not unpalatable but insipid because neither
sweet nor pungent. Very inferior in flavor to the plants blanched with
boards. Texture slight Iv tough and stringv.

No. 336

Flavor rank and
Blanched With li.
A1 I). Hluncfiinii I'd/x r. Flavor slightly rank and pungent. Very tough
and stringy. (Hi Iil(i)tcli< <l \\~it1i ^oil. Flavor very rank and pungent.
Very tough and stringy.

Apparently the method of blanching ha* no very marked effect
on the quality of celery from the standpoint of texture. However,
in the case of ( iarrahan's Easy Blanching, blanching with boards
seemed to result in better texture. If the blanching had been allowed
to continue for a longer period, it is possible that greater differences
in texture would have resulted, tho the entire lot was harvested at
the time it was considered to be at its best condition.

The effect of blanching on flavor was rather striking in the case
of ( iarrahan's Easy Blanching. The flavor was decidedly superior
in the lot blanched with boards and very inferior in the lot blanched
with soil. This is contrary to the general belief that blanching with
soil results in superior flavor. However, it should be noted that only
the self-blanching type was included in this test and it is possible
that the green type might have given different results. It is diffi-
cult to explain the insipid flavor in the lot blanched with soil. Per-
haps the soil caused a more rapid blanching and consequently a loss
of flavor because the process was continued too long.

The prevailing opinion that blanching with soil results in better
quality the author believes may not be well founded, differences in
quality being due to differences in varieties rather than to methods
of blanching. The green varieties are usually blanched with soil while
the self-blanching varieties are commonly blanched with boards or
blanching paper, and unquestionably the green varieties are of superi-
or quality.

Effect of Supplementary Irrigation on Quality. The following re-
port is a comparison of the effect of supplementary irrigation on the
quality of different varieties of celery. The unirrigated lots received
only the moisture from the season's rainfall, while the irrigated lots
received supplementary overhead irrigation whenever it was thought
that additional water might be beneficial to the plants. Otherwise the
cultural conditions were the same.

(i unit Pascal. Irrigated — flavor excellent; slightly sweet, nutty, and
very palatable; crisp texture but slightly stringy tho strings were
nor tough. I HuTKjatid -flavor excellent; sweet, nutty, and very pala-
table; crisp, tender, and succulent.

Button"* (Hunt lt«l. Irrigated — flavor excellent; sweet, nutty, and
very palatable; very crisp and tender in texture. Unirrigated — flavor
excellent; sweet, nutty, and very palatable: very crisp, tender, and

succulent.

lit?!)} QTAI.ITY i.\ CKI.KKY AS RKI.ATKD TO STIU'CTTUK 57 1

Garrahan's Eaxt/ Blanching. Jrrigatt d— flavor excellent; slightly
sweet; crisp and slightly stringy tlio strings .were brittle and no! tough.
I'nirrigati <{ — flavor slightly pungent; crisp, but somewhat stringy.

E/n/xror. Irrigated- -flavor excellent; sweet, nutty, and palatable;
very crisp and succulent. Unirrigated — flavor very good; not quite so
sweet as the irrigated lot; crisp and succulent, but slightly stringy
tlio strings were not tough.

White Plume. Irrigated — flavor rank and pungent ; very tough and
stringy. Unirrigated — flavor slightly pungent, but more palatable
than irrigated lot; texture tough and stringy.

Kose Kibbed Self -Blanching. Irrigated — flavor poor; slightly sweet:
very tough and very stringy. Unirrigated — flavor poor; rather flat,
but slightly sweet; quite tough and stringy.

Sutton's Superb Pink. Irrigated — flavor excellent; sweet, nutty, and
very palatable; crisp and tender in texture. Unirrigated — flavor good;
slightly sweet and nutty, but not as fine-flavored as the irrigated lot;
slightly tough and slightly stringy but strings were not very tough.

From the above results there appears to be a slight advantage
in favor of supplementary irrigation. With four of the seven varieties,
namely Garrahan's Easy Blanching, Emperor, Rose Ribbed Self-
Blanching, and Sutton's Superb Pink, the flavor was superior in the
irrigated lot. With one variety, White Plume, the unirrigated lot was
superior in flavor. With two varieties, Giant Pascal and Sutton's
Giant Red. the flavor seemed equally good on the two plots.

In regard to texture the supplementary irrigation seemed to im-
prove the quality of Ciarrahan's Easy Blanching and Emperor. How-
ever, two other varieties, Giant Pascal and Rose Ribbed Self-Blanch-
ing, were of better texture on the unirrigated plots. The other three
varieties showed no difference in texture due to the irrigation.

In connection with the above results it should be noted that
this celery was grown in a humid region in a season of normal rain-
fall and that the unirrigated lots made' a strong, vigorous growth simi-
lar to the plants that were irrigated. The early growth seemed to be
a little more rapid on the irrigated lots, but late in the season
there was no apparent difference in the amount of growth. Under
these conditions striking contrasts in quality from the additional
irrigation could scarcely be expected.

LABORATORY TECHNIC

In order to make microscopic examinations of the various speci-
mens of celery, samples were taken and the material prepared in the
following manner. An effort was math1 to select typical specimens
of each lot in the manner previously described. After the celery was

trimmed as for market, three outer leaves were removed and the
fourth one taken for sampling. A section about one inch long was
cut from thi> leaf petiole beginning one inch above the base.

Kach -eminent for examination was first labeled and placed in
a fixing solution. For this purpose a saturated solution of picric acid
in 70-percent alcohol was tised. Picric acid was chosen because of
its great penetration and because it would fix and preserve the entire
section without unduly hardening the tissues. Furthermore tissues
fixed in picric acid can. after the washing out of the acid, be perfect-
ly stained with any stain.

The picric acid was then washed out by soaking the material in
numerous alcohol baths. After the picric acid was entirely washed
out. the alcohol was removed by gradually introducing choloroform
and then paraffin, and the material was imbedded in paraffin molds.

Using a microtome, sections 9 microns in thickness were made
of each specimen. Cross-sections were made of each sample, and
longitudinal sections were also made of specimens that showed out-
standing differences in the quality tests. These sections were mounted
on slides.

At first a triple stain was tried consisting of safrinin. gentian
violet, and orange G, but this combination of stains did not bring out
satisfactory contrasts in the coloring of the different tissues. It was
found that staining the sections first in safrinin and then in Dela-
field's haematoxylin colored the sections very satisfactorily. All the
sections were thereafter stained by that process.

For the microchemical tests that were made to determine the
composition of the cell walls, sections were cut 30 microns in thick-
ness. They were then dropped into xylol to remove the paraffin. After
the paraffin had been dissolved, the sections were placed in 95-percent
alcohol.

Sections to be tested with phlorogluein were then placed in alco-
hol in which a trace of phloroglticin had been dissolved. After ten
minute- in this solution the sections were transferred to a drop of
writer on a slide and covered with a cover glass. A drop of hydro-
chloric acid was then applied at the edge of the cover glass, and the
effect was observed thru a microscope. As the acid came in contact
with liLrnificd tissue, this part became a bright violet-red, while the
cellulose was not stained.

Sections to be tested with chloroiodid of zinc were taken from the
alcohol bath previously mentioned and were placed in water. From
this the sections were placed in a few drops of chloroiodid of zinc
on a >lide and covered with a cover glass. The resulting staining by
the chloroiodid of zinc was then observed thru a microscope. With
this stain, cellulose walls are colored violet, lignified membranes are
stained brown, and cutinized membranes are colored vellow.

RESULTS OF MICROSCOPIC EXAMINATIONS

After being stained and mounted as described, specimens of each
sample of celery that \vas tested for quality were carefully examined
under a microscope to determine what differences, if any. in the struc-
ture of the plant were correlated with differences in quality. First
in order to become familiar with the character of the different tissue?-
and cell structures composing the celery stalk, a thorn study \va^
made of several longitudinal and cross-sections. A general descrip-
tion of this internal structure is given on pages 501 to 503.

••''^ "'^i^p^::'^'^y^^r^cr^

FK;. 5. — A TKNDEH SPKCIMKX OF STTTON'S SUPKHH PINK
Note th<> un!)rok(ni collenchyma strand in the ril) and tlic large
fibro vascular bundle.

After this preliminary study was completed, a detailed study
was made of the structure of specimens that had shown striking
differences in the quality tests. It was thought that any differences
in the morphology of the plants that might be correlated with differ-
ences in quality would be most readily discovered by the examina-
tion of markedly contrasting specimens.

Comparison of Bundles, Xylem, and Phloem. First, the number
of fibrovascular bundles in tender specimens was compared with the
number of these bundles in tough specimens, then the size of the
bundles was compared in each case; but there seemed to be no
differences either in the number or in the size of the bundle's that
were1 consistently correlated with either phase of quality. This is
illustrated by comparing Fig. 5. a microphotograph of a very tender
specimen having extremely large fibrovascular bundles, and Fig. 0,

,-)7 1 Bri.i.KTix Xo. 330 [August,

a very tough, stringy specimen having much smaller fibrovascular
bundles. The magnification is the same in each of these photographs,
so that the proportional size of the bundles is as shown.

Next an examination was made of the size of the xylem, and a
count was even made of the number of large trachea! vessels in the
xylem of specimens of different quality. Again the results were nega-
tive. The structure of the phloem sheath was then carefully examined
in regard to the size of the phloem, relative size and thickness of
these cell walls, and shape of the phloem sheath, that is, whether
it extended well down around the xylem in a horseshoe shape or

!•']<;. 6. — TOUGH, STRINGY SPECIMEN OF ROSE RIBBED SELF-BLANCHING
Note the badly broken cells in the collenchyma and the small size
of the fibrovascular bundle.

not, But in all these phases of the plant structure, the differences
noted did not seem to be correlated with differences in quality. Altho
considerable variation was observed in the structure of different speci-
mens, yet similar variations of the characters mentioned were found
in specimens markedly different in quality.

Toughness Not Due to Separation of Bundles From Other Tissue.
While seeking some other connection between the structure of the fibro-
vascular bundles and toughness of the plant, it was observed that
in some of the tough specimens the tissue surrounding the bundles
was torn away from the bundle tissue, and that in some of the tender
specimens the tissue surrounding the bundles was not sharply differ-
entiated or torn away but seemed to blend into the tissue of the
bundles. This suggested the theory that there might be a difference
in the hardness of the cell walls of the different tissues and that the
greater the difference in relative hardness of the different tissues the
more noticeable would be the harder tissues when one was eating

ix CKI.KKY

KKI.ATKD TO S

the celery. Hence such specimens would he judged as tough or
stringy, and therefore poorer in quality, because the contrast in hard-
ness of the different tissues would emphasize the toucher tissue. How-
ever, after examining more sections, it was found that some tender
specimens showed a distinct separation of the hbrovascular bundles
from the surrounding tissue, and that in some tough specimens the
bundle tissue1 seemed to blend into the surrounding tissue.

Another negative restilt that should be mentioned is that no bast
fibers, wood fibers, or stone-cell tissues were found in anv of the sec-

Fie. 7. — A PITHY SAMPLE OF (IAKHAHAN'S KASY BI.AXCIIIN*;
Note the open texture due to the rupture of large areas of paren-
chyma cells.

tions. This eliminates these three of the four plant skeletal tissues
from further consideration in relation to quality of celery.

Relation of Collenchyma to Quality. Having eliminated the
above factors, attention was then concentrated on the one remaining
kind of plant skeletal tissue, the collenchyma. It was observed that this
tissue was most prominent in the ribs and that each rib was essentially
a large strand of collenchyma tissue. However, it was noticed that the
collenehyma cells were most clear and distinct in the cross-sections of
the tender specimens, as shown in Figs. f> and 9. This seemed to be
another negative factor until it was noticed that the reason the collen-
chyma was less distinct in tough, stringy specimens was that in those
specimens the collenehyma cells were badly broken, making this tissue
somewhat confused and indistinct as shown in Figs. (> and 8. The
broken cells would indicate mvater hardness of this tissue, for extreme-

BUI.I.ETIX Xo. 336

ly hard ti-sues invariably have a tendency to break or tear when they
are being sectioned instead of showing a clean cut. This suggested that
tin-re probably \vas a difference in the hardness of the cell walls of the
cnllendiyina of different specimens and that the hardening of this

Fie. 8. — TOUGH. STRINGY SAMPLE OF WHITE PLUME
Note the broken collenchyma cells, and also the glands or
oil ducts (A).

tissue would cause greater toughness and stringiness. A careful exami-
nation of all the specimens seemed to substantiate this theory, as
shown in the following detailed description of the microscopic appear-
ance of various sections.

Fn;. 9. — A TKNDEK IN.NKK STALK OF WHITE PLUME
Note the clear-cut, unbroken collenchyma in rib. Compare
tins with the .-ample shown in Fijr. 8.

QTAI.ITY IN CKI.KKY AS KKI.ATKD TO STIUTTVKE ">77

Examination of pithy specimens seemed to show a direct corre-
lation between pithiness and open spaces that were due to the rupture
or collapse of parenchyma (Fig. 7i. Apparently large areas of
parenchyma cells are torn apart or ruptured, leaving open spaces in
the interior of the stalk and thus causing pithiness.

Toughness Not Associated With Oil Ducts. In most of the speci-
mens a peculiar cell structure was found which was not definitely
identified. This has the appearance of a duct or gland, and it is
thought that these structures may be oil glands or oil ducts. If this
is the case, they probably have some relation to flavor, but this
connection was not established. These structures are marked "gland
or oil duct" in Figs. 8 and 12 and may also be seen in the other
pictures. These glands were found in both tender and tough speci-
mens and therefore did not seem to have any connection with tough-
ness of the plant.

Detailed Study of Selected Specimens. In the following detailed
description of the microscopic appearance of some of the various
specimens, a few individuals of contrasting quality are described first
and the remaining ones are given in numerical order.

No. 42. Rose Ribbed Self-Blanching From Unirrigated Plot. Very tough
and stringy. Flavor flat. Kloven primary fibrovascular bundles. Bundles some-
what small. The phloem sheath extends well down around the xylem and is
composed of small, thick-walled cells. Large tracheal vessels in the xylem.
Parenchyma cells large, irregular in shape, and thin walled. Inner row of 25
small secondary fibrovascular bundles. Collenchyma strands at each rib show
cells badly broken.

A portion of the cross-section of this specimen is shown in Fig. 6 and a
longitudinal section in Fig. 3. The latter view shows both spiral and annular
tracheal vessels.

No. 43. Sutton's Superb Pink From Irrigated Plot. Very crisp and tender,
with excellent sweet, nutty flavor. Thirteen primary fibrovascular bundles,
which are comparatively largo. Thirty-eight small secondary bundles in a row
near the inner edge. Large number (about OS) large tracheal vessels in xylem.
Lartro phloem sheath, somewhat bow shaped and composed of small dense
cells. Parenchyma cells are largo, irregular, and thin walled. Large collenchyma
strand in rib. Collenchyma colls show distinct thickening at corners and cut
cleanly without breaking. Kpidermis distinct and clearly shows thickening of
cells on outer side (the cuticle).

A cross-section view of this specimen is shown in Fig. o.

No. 28. White Plume Blanched With Boards. Very tough and stringy.
Flavor rank and pungent. Thirteen primary fibrovascular bundles of medium
si/e. Forty-nine small secondary bundles in a row near the inner edge. Medium
number (average 40) large tracheal vessels in the xylom. Large phloem sheath,
somewhat bow shaped, and composed ol small thick-walled cells. Strand ot
collenchyma tissue in each rib and also a layer of collenchyma about two
cells wide just below the epidermis on the inner surface of the stalk. The
collenchyma cells are thick walled and badly broken, particularly in the
strands of the ribs, indicating that they must be especially hard. .lust below
the- collenchyma strand in each rib there is a very prominent oil duct or

;»7S Bru.KTiN Xo. 336

trlaiid. Similar glands arc scattered thru the parenchyma. Tin; parenchyma
has unu.Mially large cells which are irregular in shape and thin walled (Fig. 8).

No. 31. Inner Stalks of White Plume. Tender, with strings present but
brittle and not tough. Flavor mildly rank. Xine primary bundles which are
medium in size. Large trachea! vessels in the xylem, but only about three-
fourths as many of the trachea! vessels as in Xo. 28, an outer stalk. The
phloem sheath is large and forms a semicircle around the xylem. The phloem
cells are small and thick walled. The parenchyma cells are large, irregular in
shape, and thin walled. There is a very dear-cut collenchyma strand in each
rib. The collenchyma cells show a distinct thickening at the angles and cur
cleanly without breaking or tearing, which indicates that they are not hard.

Fi«. 10. — X"o. 2, A TOUGH, STRINGY SPECIMEN OF EASY BLANCHING

X'ote the small fibrovascular bundle and the shading off of the
xylem. also the badly broken collenchyma.

Directly below the collenchyma strand in each rib is a distinct gland or oil duct.
Similar glands are scattered thru the parenchyma (Fig. 9).

No. 45. Pithy Specimen of Garrahan's Easy Blanching. Not very tough
or stringy, but stalk is pithy and spongy and entirely lacking in crispness.
Flavor slightly sweet, but rather flat. The microscopic examination showed 12
large primary fibrovascular bundles and 33 small secondary bundles. The
bundles were clearly differentiated from the surrounding parenchyma. The cell
walls of both the xylem and phloem tissue seem strikingly thicker than the
cell walls of the surrounding parenchyma which make the bundles stand out
so clearly. '\ here is a medium number of large tracheal vessels in the xylem.

1 he phloem sheath is large and forms a distinct semicircle around The xylem.

Die parenchyma cells are very large and thin walled. There are large open
space* in the parenchyma tissue and a distinct collenchyma strand in each
nli. The collenchyma cells are thickened at the angles and some of these
cells are broken but they are mostly intact, indicating that this tissue is only
slightly hardened. A little below each collenchyma strand is a distinct gland,
and other glands arc scattered thru the parenchyma (Fig. 7).

Longitudinal sections of this specimen show that the large open spaces in
tin parenchyma extend for considerable distances lengthwise of the stalk.

7.'>.J'<1 Qi'Ai.iTY IN CEI.KKY AS HKI.A'IKD TO STIUVTI UK ,">79

No. 2. Easy Blanching. Tough and stringy. Slightly sweet flavor. The
microscopic examination reveals 9 primary tibrovascular bundles and 2S small
secondary bundles. The primary bundles are comparatively small. 'The tracheal
vessels in the xyleni are of large size, but comparatively few in number. The
xylem seems to blend into the adjoining parenchyma. The phloem sheath is
wide, does not extend well around the xylem, but forms an arc on the outer
side. The phloem cells are small and thick walled. The parenchyma cells are
large and thin walled. Scattered thru the parenchyma are distinct glands or oil
ducts. There is a large collenchyma strand in each rib. The collenchyma cells
are small and thick walled, especially at the angles. Much of this tissue is
broken and torn, indicating that it is of too hard a texture to cut readily
(Fiir. 10).

A longitudinal section of this specimen shows that, the collenchyma cells
are long and pointed with a. spliced or interlaced arrangement. The paren-
chyma cells are somewhat square or rectangular, and the phloem cells are long
with rather square ends. MOST ol the tracheal vessels show annular rings and
a few have spiral reinforcement.

No. 5. Easy Blanching. Very tough and stringy. Slightly sweet flavor.
Nine primary fibrovascular bundles and 28 secondary bundles. There is a collen-
chyma strand in each rib, composed of thick-walled cells which are badly
broken, indicating very hard texture.

No. 6. Easy Blanching. Very tough and stringy. Slightly sweet flavor.
Twenty-one primary fibrovascular bundles, medium in size. The phloem sheath
is crescent shaped and composed of very thick-walled small cells which are
distinctly differentiated from the parenchyma. There are about 40 large tracheal
vessels in the xylem. The collenchyma in the ribs is very badly broken, in-
dicating hard texture. There is also a layer of collenchyma about three cells
deep just inside the- epidermis on the concave1 or inner side of the stalk.

No. 9. Easy Blanching. Slightly stringy, but fairly crisp and tender.
Slightly sweet flavor. Eighteen primary fibrovascular bundles medium in size.
Twenty-nine secondary bundles. Phloem sheath wide and crescent shaped and
composed of small, thick-walled cells. Medium number of large tracheal vessels
in xylem. Several oil glands scattered thru parenchyma. Distinct collenchyma
strand in each rib. The collenchyma cells are distinctly thickened at the cor-
ners. The clearness of the section thru this tissue and comparatively few broken
collenchyma cells indicates that the collenchyma was not as hard in texture
as in Xo. 6. There is a distinct collenchyma layer about two cells deep just
beneath the epidermis on the inner side.

No. 15. Easy Blanching. Slightly stringy, but a trifle tender and somewhat,
pithy. Sweet, flavor. Fourteen primary fibrovascular bundles, rather large in
^\7.(\ Many large tracheal tubes in xylem. Phloem sheath crescent shaped and
composed of very thick-walled, small cells. Parenchyma cells large, irregular, and
thin walled with many ruptured, making open spaces in this tissue. Distinct
glands scattered thru the parenchyma. Large collenchyma strand in each rib.
Collenchyma cells distinctly thickened at corners, and a few of them broken
(Fig. 11).

No. 16. Easy Blanching. Very tough and stringy. Rank, pungent flavor.
Twenty-one large-sized primary fibrovascular bundles and -1(5 secondary bundles.
Lame number of tracheal vessels in xylem. Phloem sheath rather wide and
horseshoe shaped, extending well down around xylem. Parenchyma cells large,
irregular, and thin walled. Several glands scattered thru the parenchyma. Col-
lenchyma strand in each rib composed of thick-walled cells badly broken, in-
dicating very hard texture.

;>so

BULLETIN Xo. 336

No. 19. Easy Blanching. Somewhat tough and stringy. Slightly sweet
tlavi>r. Sc\ i •ntrrii rather small primary fibrovascular bundles and 30 secondary
bundles. All tisMies similar to those described in Xo. 16 except that the oil
duels seem more abundant and the collenchyma tissue is not quite so badly

•'CT5^

/i i i- vju ; vp i.'**-*7 >«v »*.^ jLW*./ ^ A • ., • fc

J'*r*^ > l^^^^&V^-V^-'^-V
,e^^ - , '-A -^ N "i^U^^^S^W4- , :* \ >
^#: * I-^^V^^^O '% %%

• .>H'-:-^K-;^^^';>r ^-

YIG. 11. — Xo. 15. A SOMEWHAT PITHY SPECIMEN OF EASY BLANCHING
SHOWING MANY RUPTURED PARENCHYMA CELLS

broken and torn, indicating that while it is quite hard in texture it may not
have been so hard as in Xo. 16 (Fig. 12).

No. 20. Easy Blanching. Very tough and stringy. Slightly sweet flavor.
Xineteen primary fibrovascular bundles, quite small in size. Tracheal vessels in

x^r^x^Sr^^^^i,-^: ^'CG^' i'/tr:-%^

'^^^Sl-^^^^^^?^":^;,^^^^^

,'i — '"1 X ,. ^.,v * ^ •,-;• j9S^s:"-' "•'•"^ -vi'iH- • N " * ~-*v V-- • • v?o* ^ *^2^ ' !

OUAI.ITY IN (Y.LF.KY AS liKI.ATKI) To SllU

->M» vco^-C^V M^^VJr V^^fSks.
^^^^S^^^-^^\

$$&s$$$j^

^MJ^iik

e c^> ^r; Veti^^^gj

\ -v-v K>- T \>9» r i^i^^/nfcWhLz&E

Fn;. Kv — A VERY TOUCH AND STKINCY SPKCIMKN or EASY HI.ANCHINC,
WITH S.MALI, FiHHovAscn.AH BUNDLES

An uinisually lurjir number of glands sctittt^rcd tliru tho parenchyma, Dcnso
collrnchyma strand in each rib, this tissue being badly broken, the cells; break-
ing in sharp angles as tho very hard in texture.

-. . - - .

SW v-1

•.- -.-i--, :-•-,> >. ".'.' '"\ - )-*' > -V' --,.— ^ -^^^B

idc

;,s_> Hn.i.KTiN No. 336 [Anyii*t.

No. 22. Easy Blanching. Very rouirh and stringy. Rank. pungent flavor.
Nineteen primary fibrovascular bundles, somewhat small in size. Medium
number nf tracln'al vessels in xylrm. Phloem sheath forms wide semicircle
around the xylem. Phloem cells small and thick walled. Numerous glands
sratten<l thru the parenchyma. Collenchyma tissue considerably torn and
broken, indicating hard texture of cell walls. Also a layer of collenchyma about
two cells deep just below inner epidermis (Fig. 13).

No. 23. Garrahan's Easy Blanching Blanched With Boards. Slightly
stringy, but strings are not tough. Fairly crisp. Flavor sweet. Fifteen primary
fibrovascular bundles, large in size. Thirty-seven secondary bundles. Xylem
rather large with numerous (about 70) large tracheal vessels. Phloem sheath in
semicircle around the xylem. The outer cells in this sheath are large and tend
to blend into surrounding parenchyma. Parenchyma cells very large, irregular,
and thin walled. A few glands scattered thru the parenchyma. The collenchyma
,-trand in each rib is not large. The collenychma cells are rather large. Very
few are broken, indicating that this tissue is not especially hard in texture
(Fig. 14).

No. 24. Outer Stalk of Garrahan's Easy Blanching Blanched With Boards.
Quite stringy and tough. Flavor sweet, slightly rank. Fourteen large-size pri-
mary fibrovascular bundles and 27 secondary bundles. Xylem large with many
lame tracheal vessels. Wide phloem sheath in semicircle around xylem. Small,
thick-walled cells on inner edge of sheath grading into larger cells on outer
edge. A few glands scattered thru the parenchyma. Rather small collenchyma
strand in each rib. Most of the collenchyma tissue broken and torn, indicating
hard texture of cell walls. Distinct collenchyma layer about four cells deep
just below epidermis on concave or inner side of stalk.

No. 25. Inner Stalk From Same Plant as No. 24. Crisp and tender.
Sweet flavor. Eleven primary fibrovascular bundles, quite small in size. Small
number of large tracheal vessels in xylem. Phloem sheath composed of dense,
small cells forming an arc on the outer side of the xylem but not extending
far around the xylem. Parenchyma cells large, irregular, and thin walled. Numer-
ous glands scattered thru the parenchyma. Large, distinct collenchyma strand
in each rib. The collenchyma cells are small and thickened at the angles. The
entire collenchyma tissue at every rib was intact, showing that it cut readily
and therefore the texture of these cells was not hard.

Fig. 1 shows a portion of the cross-section of this specimen. A longitudinal
section showed the collenchyma cells to be spindle shaped and spliced or
interlocked.

No. 30. White Plume Blanched With R. & D. Paper. Very tough and
stringy. Flavor rank and pungent. Fifteen large primary fibrovascular bundles
and 24 secondary bundles. Xylem large with numerous (65) large tracheal
vessels. Phloem sheath composed of small, thick-walled cells, and extends well
around the xylem in a, horseshoe shape. Collenchyma strand in each rib very
badly broken, cells very thick walled and broken in sharp fragments, indicating
hard texture. Layer of collenchyma about three cells deep next to inner
epidermis.

A longitudinal section of this specimen showed that the collenchyma cells
were spindle shaped but badly broken. Many annular and a few spiral tracheal
vessels were found.

No. 32. Giant Pascal From Irrigated Plot. Crisp texture but slightly
stringy, tho >t rings were not tough. Flavor sweet and nutty. Thirteen primary

OTAI.ITY IN CKLKKY AS RKI.ATKH T

X f .'...'* ^7 >*~v/. ^iJS..,* -\j ft-s-1 v) "^ *• *^*?

•"Vj •*--^-x''i>i>^t 'V^O&c

Mj|il|MS

,N -/- 'ViVV;. /• ; '-; >-'r^:vi-^

LllS&^^M

'•''mmm*.+^m

,; , rv^p^-rsv-^^y: -'-":'--:" -•'•'• v-^xt

I'"n;. 1.1 — No. :}.">. STTTON'S (!IANT J^KD

Note the hirjrr fibrovasciilar Innidlo and cxtcnsi\'e l>ut unljroken
collc-nchynnt slr;in<l. Tlic Inviir si/c of tin Imndlc made it necessary
io take T\V<> photographs in ordi r to .-h<i\v Imth tlie lihroviiscular
liiindle and t lie eolleliclivina.

.-heath wide and en .-cent shaped. Handles set in to blend into the surrounding
parenchyma. Parenchyma ci-IL- very larire and thin walleil. C/omparativcIy
few glands scattered thru the parenchyma. I)i>tinct collenchyma la\'( r next
to epidermis on conca\"e -ide. Collenchyma .-trand in each rib fairly intact,
tho some ci Us were bri>ken, indicating only ,-liuhtly hard texture.

5s i

Bri.I.MTIN Xo. 336

No

. 33. Giant Pascal. Inner stalks from Xo. 32. Very crisp and tender.
nutty. flavor. Nineteen small primary fibrovascular bundles. Xo inner
secondary bundles. Very few (about 15) large tracheal \"essels in xylem.
sheath crescent shaped and quite wide. Bundle cells much thicker
nd therefore distinctly differentiated from surrounding parenchyma. The
were much smaller than the bundles in Xo. 32. Very clear-cut collen-
strand in each rib. The>e cells show distinct thickening at the angles
without breaking or tearing, indicating that the cell walls were not
Narrow layer of collenchyma next to epidermis on concave side.
inal section of this specimen (Fig. 4) shows; that the collenchyma
spindle shaped and interlocked while the parenchyma cells are some-
are. The cells of the phloem sheath are very long and somewhat ob-
shape. The tracheal tubes show very distinct, annular rings.

Kic;. 16. — KMPEKOH CELERY STALK (Xo. 39), A CKISP, SUCCULENT SPECI-
MEN WITH LARGE FIBROVASCULAR BUNDLES

No. 35. Sutton's Giant Red From Unirrigated Plot. Very crisp, tender,
rind succulent. Sweet, nutty flavor. Eleven very large primary fibrovascular
bundles. Thirty-live secondary bundles. Altho the bundles are larger than
those lound in any previous specimen, they seem to blend into the surrounding
parenchyma. Very numerous (about 90) large tracheal vessels in xylem. Phloem
sheath bow shaped and narrow in proportion to length. Parenchyma cells
iaru'c and thin walled. (.Hands or oil ducts scattered thru the parenchyma.
Hat her extensive collenchyma strand in each rib, comparatively thin-walled
cells lor collenchyma, showing distinct reinforcement at angles. The collen-
chym.-i strands mostly intact with very few cells broken, indicating that tex-
ture is not hard (Fig. lo).

No. 36. Sutton's Giant Red From Irrigated Plot. Very crisp and tender.
Sweet, nutty flavor. Twelve large primary fibrovascular bundles and 2S second-
ary bundles. All tissues very similar to Xo. 35. Longitudinal section shows
iinular and spiral tracheal vessels; also showed collenehyma cells spindle
shaped and spliced or interlocked.

Qr.u.iTY ix CKLKHY AS RKI.ATKD TO STKIVITKK f)S~)

No. 39. Emperor From Unirrigated Plot. (Yi>p and succulent but slight-
ly striimv, tlio strings arc not touch. Sweet ilavor. Thirteen lartie primary
fibrovasrular bundles, and 22 secondary bundles. Medium number of lame
iracheal vessels in xylem. 1'hlovm sheath bo\v >haped and rather narrow.
Parenchyma cells very lar.ni> and thin walled. Numerous inlands scattered thru
the parenchyma. Rather small collenchyma strand in each rib. Comparatively
larire collenchyma cells with distinctly thickened corners. Cells mostly intact
but somewhat broken, indicating a hardened texture ( Fin;. l(i).

Microchemical Tests

To determine the composition of the cell walls in the various
tissues, a number of sections were treated with phloroglucin and with
chloroiodid of /inc. as described on pa lie 572. The following specimens
were tested by these methods: Nos. 2. 24. 25, 29. 35. 36. 38. 39. 40,
42, 43, 44, and 45. This assortment embraced all varieties and in-
cluded both tough specimens and tender ones, and inner and outer
stalks. Both longitudinal and cross-sections were tested. The results
were identical in each case. Consequently one description will suffice
lor all.

There were only two kinds of tissues that showed lignifi catkm ;
namely, the tracheal vessels and the oil glands or ducts. In all speci-
mens only these elements were liquified and the inner tender stalks
showed the same nullification as the outer tough stalks. This was
similar in all varieties. There were no additional liquified elements in
the tough, stringy specimens. The spiral and annular reinforcement
in the tracheal vessels in all specimens seemed particularly liquified.
The tracheal tubes in the secondary bundles showed the same ligni-
fication as the tracheal tubes in the primary bundles.

Neither chemical test showed any trace of lignin in the collenchy-
ma. Therefore, there was no nullification in the ribs of the celery.
In the tests with chloroiodid of xinc the collenchyma, parenchyma,
and phloem and the other xylein elements except the tracheal tubes
all stained clear violet, indicating pure cellulose walls. The outer
side of tlu1 epidermal cells (the cuticle I stained clear yellow, indi-
cating that it was culinized.

SUMMARY AND CONCLUSIONS

This investigation resulted in some conclusive evidence in regard
to certain changes in the morphology of the celery plant that are
correlated with quality. Tin1 choice of the horticultural variety is
of ureat importance in producing high-quality celery, and the en-
vironment under which it is grown may cause modifications in the
structure of the plant which affect its quality. Haberlandt:!* says,
"Generally speaking, both the qualitative and the quantitative de-
velopment of the mechanical system are included among the heredi-

;,si', BCLI.KTIX Xo. 336

tary characters of the species. Nevertheless, a certain amount of di-
rect accommodation to external conditions on the part of the -mc-
cjianical tt/xttrn may take place during the life of the individual
plant."

In determining the nature of the changes in the mechanical sys-
tem which arc correlated with toughness and stringiness in celery,
some negative results seem especially significant, particularly in re-
gard to the relation of the fibrovascular bundles to stringiness.

1. Contrary to popular belief and to the opinion of some writers,
this investigation seemed to show conclusively that the size of the
fibrovascular bundles in celery is not a factor in causing stringiness.
One of the most tender varieties, Slit-ton's Giant Red, had unusually
large fibrovascular bundles (Fig. 15). Another tender variety. Giant
pascal, had medium-sized bundles. On the other hand, Rose Ribbed
Self-Blanching, a very tough, stringy variety, had medium-sized
bundles (Fig. 6). Another tough variety. White Plume, also had
medium-sized bundles I Fig. 8). Apparently the size of the fibrovas-
cular bundles is rather constant within a variety, but varies between
varieties, and large bundles are not correlated with tougher or stringier
varieties.

The fibrovascular bundles are smaller in the smaller inner stalks
on the same plant than in the large outer stalks but arc in propor-
tion to the entire size of the stalk.

The number of fibrovascular bundles in a stalk likewise seems to
have no relation to stringiness or toughness. There is a great varia-
tion in the number in different stalks of the same or of different vari-
eties and even in different stalks from the same plant, but there
seems to be no increase in the number of bundles in tough, stringy
specimens as compared with tender specimens.

2. Xo bast fibers were found in any specimens. Stringiness. there-
fore, cannot be attributed to the presence of this tissue. The elements
called glands or oil ducts, which were found scattered thru the paren-
chyma and which were lignified, seemed to have no relation to tough-
ness or stringiness. Probably these elements, if they are oil ducts,
have some relation to flavor, but there seemed to be no correlation
between the number of these ducts and the flavor of the specimen.

Since lignin is often associated with hard, tough tissue, and since
the tracheal vessels were the principal lignified elements, the relation
ot these vessels to stringiness must be considered. However, the
number of these lignified tracheal tubes seemed to be approximately
in proportion to the size of the fibrovascular bundles and therefore
seemed to have no relation to stringiness. Lignified tissues are not al-
ways of tough, hard texture. There are degrees of lignification and de-
grees of hardness of limiified tissue, and apparently the lignified ele-
ment- in the celery plant are not particularly hard or tough.

v IN ('F.I.KHY AS HKI.

3. Tin1 only celery tissue that seemed to have a very definite re-
lation to stringiness was the collenchyma. A superficial examination
of strhmy celery would support this belief, for, when a stringy celery
-talk is broken so that the strings tear out. it can be readily seen that
the strings are in the ribs where the collenchyma strands are found.
This is likewise in accordance with statements of plant anatomists
to the effect that collenchyma regularly forms the skeletal syMcm of
LLTowiim' organs, and often serves as the permanent mechanical tissue
in many fully urown herbaceous structures, such as leaf petioles.

Apparently the relative amount of collenchyma tissue in the ribs
is nor a factor in causing string-mess, for many of the tender speci-
mens had relatively large strands of collenchyma. The size of the
collenchyma cells and the thickness of the cell walls likewise seemed
to have no bearing on stringiness of celery. However, there seemed
to be a i:re at variation in the decree of hardness of the collenchyma
cell walls. Furthermore there was a definite relation between the de-
gree of hardness of the collenchyma tissue and stringiness of the celery
petiole. The collenchyma tissue in tough, stringy specimens seemed
to be of particularly hard texture, while in tender specimens, both
of tender varieties and of inner stalks of tough varieties, the texture
of the collenchyma tissue did not seem to be hardened. This was the
only morphological factor that seemed correlated with toughness and
strhminess in celery.

4. Pithiness is evidently correlated with a breaking down of the
parenchyma cells which leaves large open spaces thru the center of
the stalk.

5. Flavor seems to be influenced to a greater extent by the vari-
ety than by any environmental factors. Of course, proper blanching
is essential to good flavor. No morphological conditions could be found
that seemed correlated with flavor.

6. The relation between cultural conditions and quality were not
clearly developed in this investigation, but the findings regarding
differences in structure that are correlated with differences in quality
seem conclusive.

.")SS BULLETIN Xo. 336

LITERATURE CITED

1. AUSTIN. C. F., and WHITE, T. H. Second report on the cause of pithiness

in celery. Md. Agr. Exp. Sta. Bui. 93. 1904.

2. BK.ATTIE, W. R. Celery culture. Orange Judd Co. 1907.

3. HABERLANDT, G. Physiological plant anatomy. Macmillan. 1914.

4. MILLS, H. S. Quality in celery. Market Growers Jour. May 15, 1923.

5. XOKTOX, J. B. Concerning quality in celery. Vt. Agr. Exp. Sta. Bui. 203. 1917.

6. SANDSTEN, E. P., and WHITE, T. H. An inquiry as to the causes of pithi-

ness in celery. Md. Agr. Exp. Sta. Bui. 83. 1902.

7. STEVENS, W. C. Plant anatomy. P. Blakiston's Son & Co. 1911.

8. THOMPSON, H. C. Celery storage experiments. U. S. Dept. Agr. Bui. 579.

1917.

9. THOMPSON. H. C, Vegetable crops. McGraw-Hill Book Co. 1923.
10. WATTS. R. L. Vegetable gardening. Orange Judd Co. 1914.

UNIVERSITY OF ILLINOIS-URBANA