Full text of "Journal of agricultural research"

See other formats

1, 432 L07

JOURNAL OF
AGRICULTURAL
RESEARCH

Volume I

OCTOBER, 1913— MARCH, 1914

DEPARTMENT OF AGRICULTURE

28736° — 14 — S

WASHINGTON, D. C.

Published by Authority of the Secretary of Agriculture

EDITORIAL COMMITTEE

Karl F. Kellerman, Chairman
Edwin W. Allen
Charles L. Marlatt

CONTENTS

Page

Foreword. B. T. G allow ay . i

Citrus Ichangensis, a Promising, Hardy, New Species from

Southwestern China and Assam. Walter T. Swingle . i

Cysticercus Ovis, the Cause of Tapeworm Cysts in Mutton.

B. H. Ransom . 15

The Serpentine Leaf -Miner. F. M. Webster and T. H. Parks. . 59

The Occurrence of a Cotton Boll Weevil in Arizona. W. DwtghT

Pierce . 89

The Diagnosis of Dourine by Complement Fixation. John R.

Mohler, Adolph Eichhorn, and John M. Buck . 99

Three Undescribed Heart-Rots of Hardwood Trees, Especially of

Oak. W. H. Long . 109

Individual Variation in the Alkaloidal Content of Belladonna

Plants. Arthur F. Sievers . 129

The Pubescent-Fruited Species of Prunus of the Southwestern

States. Silas C. Mason . 147

Selective Adsorption by Soils. E. G. Parker . 179

A Bacterium Causing a Disease of Sugar-Beet and Nasturtium

Leaves. Nellie A. Brown and Clara O. Jamieson . 189

The Calliephialtes Parasite of the Codling Moth. R. A. Cushman . 2 1 1
Polyporus Dryadeus, a Root Parasite on the Oak. W. H. Long . . 239

The Foot-Rot of the Sweet Potato. L. L. Harter . 251

Environmental Influences on the Physical and Chemical Charac¬
teristics of Wheat. J. A. LE ClERC and P. A. Yoder . 275

A Drought-Resisting Adaptation in Seedlings of Ho£i Maize.

G. N. Collins . 293

Some Diseases of Pecans. Frederick V. Rand . 303

A Twig Blight of Quercus Prinus and Related Species. Della E.

Ingram . 339

New Potato Weevils from Andean South America. W. Dwight

Pierce . 347

An Undescribed Species of Gymnosporangium from Japan. W. H.

Long . 353

The Presence of Some Benzene Derivatives in Soils. Edmund C.

Shorey . 357

Journal of Agricultural Research

Vol. I

Page

Indicator Significance of Vegetation in Tooele Valley, Utah. T. H.
Kearney, T. J. Briggs, H. L. Shantz, J. W. McLane, and R. L.

PiEMEiSEL . 365

Citropsis, a New Tropical African Genus Allied to Citrus. Walter

T. Swingle and Maude Kellerman . 419

Winter Spraying with Solutions of Nitrate of Soda. W. S. Bal¬
lard and W. H. Volck . 437

Tyloses, a Study of Their Occurrence and Practical Significance in

Some American Woods. KloisE Gerry . 445

The Cambium Miner in River Birch. Charles T. GrEENE. _ _ 471

A Study of Some Imperfect Fungi Isolated from Wheat, Oat, and

Barley Plants. Edward C. Johnson . 475

The Origin of Some of the Streptococci Found in Milk. T. A.

Rogers and Arnold O. Dahlberg . 491

Crystallization of Cream of Tartar in the Fruit of Grapes.

William B. Alwood . 513

The Reduction of Arsenic Acid to Arsenious Acid by Thiosulphuric

Acid. Robert M. Chapin . 515

Index . 519

ERRATA

Page 2, footnote, line 1, an gust is” should read “angustis.”

Page 2, footnote, line 2, “ate” should read “late.”

Page 28, line 30, “ Taenia ovis (Cobbold, 1869) Ransom, n. comb., 1913,” should read
“Taenia ovis (Cobbold, 1869), Ransom, 1913.”

•Page 98, line 5, “Figs. 3 and 7. — Side and dorsal views of female” should read “Figs.
3 and 7. — Side and dorsal views of male. Figs. 4 and 8. — Side and dorsal views
of female.”

Page 176, line 19, “Prunus havardii W. F. Wight, n. comb.” should read “Prunus
havardii (W. F. Wight), n. comb.”

Page 421, footnote, line 2, “locularis” should read “ locularibus. ”

Page 421, footnote, line 7, “locularis” should read “loculare.”

Page 425, figure 4, “gabonensis” should read “gabunensis.”

ILLUSTRATIONS

Plate

I. Citrus ichangensis Swingle: The type specimen from Hsing-

shan District, Hupeh Province, China . 14

II. Fig. 1. — Cysticercus ovis from lamb which had been fed eggs
of Taenia ovis. Fig. 2. — Cysticercus cellulosae. Fig. 3. —

Taenia ovis. Fig. 4. — Taenia hydatigena (T. marginata)
from an imported sheep dog. Fig. 5. — T. hydatigena (T.
marginata) from a dog which had been fed Cysticercus

tenuicollis . 58

III (colored). Figs. A and B. — Portions of muscle of sheep
showing Cysticercus ovis (undegenerated) in situ. Fig.

A. — Section of hind leg showing two “deep” cysticerci.

Fig. B. — Hind leg showing three “superficial” cysticerci.

Figs. C and D. — Heart and portion of diaphragm of sheep
showing Sarcocystis nodules likely to be mistaken for de¬
generate cysticerci. Fig. E. — Sheep heart showing nu¬
merous small degenerate cysticerci ( Cysticercus ovis) . 58

IV. Fig. 1 . — Carcass of sheep showing a degenerate cyst of Cysti¬
cercus ovis at the point indicated by the penknife. Fig.

2 . — Degenerate cysts of Cysticercus ovis in muscle of sheep ; ,

portion of carcass shown in Plate III, figs. A and B . 58

V. Leaves of different species, showing the work of the serpen¬
tine leaf-miner (Agromyza pusilla). Fig. 1.— Mines in a
leaf of rape. Fig. 2 — Mines in leaves of white clover.

Fig. 3. — Mines in leaves of alfalfa . 88

VI. Figs. 1, 2, 5, and 6. — Anthonomus grandis ihurberiae: Type
specimens. Figs. 1 and 5. — Side and dorsal views of
male. Figs. 2 and 6. — Side and dorsal views of female.

Figs. 3, 4, 7, and 8. — Anthonomus grandis . Typical
specimens. Figs. 3 and 7. — Side and dorsal views of
male. Figs. 4 and 8. — Side and dorsal views of female.

Fig. 9. — Thurberia thespesioides: Section of boll, showing
cell of Anthonomus grandis ihurberiae. Fig. 10. — Thur¬
beria thespesioides: Seed, showing cell of Anthonomus
grandis ihurberiae. Fig. n. — T hurberia thespesioides: Boll ,
showing egg puncture of Anthonomus grandis thurberiae. . 98

VII. Fig. 1. — Polyporus pilotae: A sporophore on the end of a
white-oak log from Arkansas. Fig. 2 . — Polyporus pilotae:

Rot appearing in the butt of a white-oak log from Arkan¬
sas, showing the holes and white cellulose areas charac¬
teristic of this rot in a cross section of a living oak. Fig.

3. — Polyporus pilotae: Radial-longitudinal view of a
white-oak log from Arkansas, showing the honeycomb
type of the rot with the white cellulose lines and elliptical
hollows. Fig. 4. — Polyporus pilotae: Rot occurring in a
log of Castanea pumilairom Arkansas; A , concentric layers
of the rotted wood; B, white cellulose fibers. Fig. 5. —
Polyporus pilotae: Cross section of a chestnut log from
New York, showing the central circular rotted zone.

Fig. 6. — Polyporus pilotae: Radial-longitudinal view of
the rot in a chestnut log from New York, showing the
white pocketed stage . 12S

Journal of Agricultural Research

Vol.I

Page

PifATiJ VIII. Fig. i. — Polyporus pilotae: Radial-longitudinal view of the
rot in a chestnut log from New York. Fig. 2. — Polyporus
berkeleyi: Radial-longitudinal view of the rot in white-oak
timber from Arkansas, showing the string and ray form
characteristic of its second stage. Fig. 3. — Polyporus
berkeleyi: A sporophore on a white-oak root from Arkansas.

Fig. 4. — Polyporus frondosus : A sporophore on roots of

white oak from Arkansas . 128

IX. Fig. 1. — Prunus texana: Better quality of fruit. Fig. 2. —
Prunus texana: Fruiting bush, 2 meters in diameter.

Fig. 3. — Prunus texana: Seeds; three scraped clean of

pile . 178

X. Fig. 1. — Prunus texana hybrid, hort. var. Stuart: Fruit and
leaves. Fig. 2. — Prunus texana hybrid, hort. var. Stuart:

Tree in first leaf. Fig. 3. — Prunus texana hybrid, hort.

var. Johnson: Fruiting branch . 178

XI. Fig. 1. — Prunus andersonii: Plant, showing taproot. Fig.

2. — Prunus andersonii: Flowering branch. Fig. 3. —

Prunus andersonii: Types of seeds . 178

XII. Fig. 1. — Prunus andersonii: Tangled thickets, the more
common form. Fig. 2. — Prunus andersonii: Treelike
specimen, 3 meters high. Fig. 3. — Prunus eriogyna , n.
sp.: Erect, large-leaved form of plant . 178

XIII. Fig. 1. — Prunus eriogyna^ n. sp.: Common form of plant.

Fig. 2 . — Prunus eriogyna , n. sp. : Variable fruits and seeds.

Fig. 3. — Prunus eriogyna , n. sp.: Fruiting branch . 178

XIV. Fig. 1. — Prunus eriogyna , n. sp.: Seedlings. Fig. 2. — Pru¬

nus fasciculata: Growth in flood-swept wash . 178

XV. Prunus minutijlora: Fruiting branch . 178

XVI. Prunus havardii: Fruiting branch of the type specimen . 178

XVII. Fig. 1. — Sugar-beet leaves inoculated with Bacterium apia-
turn . Fig. 2 . — Sugar-beet root inoculated with Bacterium

aptatum . 210

XVIII (colored). Nasturtium leaves showing bacterial leaf spots 10

days after inoculation with Bacterium aptatum . 210

XIX. Fig. 1. — Bean leaves inoculated with Bacterium aptatum
from leaf-spot of sugar beet. Fig. 2. — Nasturtium leaves
inoculated with Bacterium aptatum from leaf-spot of
sugar beet. Fig. 3. — Bean pods inoculated with Bac¬
terium aptatum from leaf-spot of sugar beet . 210

XX. Calliephialtes sp. Fig. 1. — Female. Figs. 2 and 3. — Char¬
acteristic positions assumed by the insect in oviposition.

Fig. 4. — Male . 238

XXI. Fig. 1. — Polyporus dryophilus: A median-longitudinal sec¬
tion of a sporophore on Quercus alba from Arkansas, show¬
ing the granular core and the white mycelial lines in the
central and rear portion. Fig. 2. — Polyporus dryophi¬
lus: Side view of the ungulate type of sporophore on
Quercus calif ornica from California. Fig. 3. — Polyporus
dryophilus: Median-longitudinal section of the globose
type of sporophore on Quercus garryana from California,
showing the large granular core and prominent white
mycelial lines. Fig. 4. — Polyporus dryadeus: Median-

Oct., X9i3-Mar,, 1914

Illustrations

VII

Page

longitudinal view of a young sporophore on Quercus tex-
ana from Texas, showing the fibrous, nongranular nature
of the context. Fig. 5. — Polyporus fulvus Fries: Median-
longitudinal view of a sporophore on Quercus sp. from
Sweden, showing the granular core characteristic of P.
dryophilus. Fig. 6. — Polyporus vulpinus : Median-longi¬
tudinal view of sporophore on Populus sp. from Sweden,
showing the granular core characteristic of P. dryophilus.

Fig. 7. — Polyporus dryophilus: Front view of the appla-
nate type of a sporophore on Populus tremuloides from
Colorado, showing the faint zones on the pileus where the
hairs have disappeared. Fig. 8. — Polyporus dryophilus:
Median-longitudinal view of sporophore on Populus tre¬
muloides from Colorado, showing the granular core origi¬
nating between the sapwood and bark and extending

into the center of the sporophore . 250

Pirate XXII. Fig. i. — Polyporus dryophilus: Radial-longitudinal view of
the rot occurring in Quercus sp. from Furope and said to
be the rot produced by P. dryadeus. Fig. 2. — Polyporus
dryadeus: Cross section of a small root of Quercus alba from
Maryland, showing the mottled appearance of the diseased
wood in the middle stages of the rot. Fig. 3. — Polyporus
dryophilus: Radial-longitudinal view of the rot appearing
in Quercus alba from Arkansas, showing the advancing
line of rot in a branch. Fig. 4. — Polypoms dryadeus: Up¬
per surface of a sporophore on roots of Quercus texana from
Texas, showing the rough tuberculate pileus. Fig. 5. —
Polyporus dryadeus: Rot occurring in an apparently sound
root of Quercus alba from Virginia, showing cross section
of a diseased root immediately adjacent to the point of
attachment of a large sporophore of P . dryadeus , 1 foot high
and 1 foot wide. Fig. 6. — Polyporus dryadeus: Cross sec¬
tion of a diseased root of Quercus alba from Virginia, show¬
ing the nearly sound, living upper half of the root and

the badly diseased lower half . 250

XXIII. Parts of sweet-potato plants, showing the presence of
pycnidia: A, On the stem just above the ground; B , on

the root . 274

XXIV. Portion of sweet-potato vines several feet from the hill,
showing the results of a natural infection of the foot-rot
fungus . 274

XXV. Microscopic characters of the foot-rot fungus: A , Section
through a pycnidium on the root; B, section through a
pycnidium on the stem; C, hyphae, from artificial cul¬
ture; D and E , chlamydosporelike bodies found on the
host and in some culture media; F, pycnospores; G,
club-shaped bodies often found in pycnidia; H, ger¬

minating pycnospores . 274

XXVI. Two sweet-potato plants in pots, demonstrating the parasit¬
ism of the foot-rot fungus: A, Inoculated; B, not inocu¬
lated . 274

XXVII. Nine-day-old cultures on synthetic agar: A, The conidial

stage of Diaporthe batatatis ; B, Plenodomus destruens . ... 274

VIII

Journal of Agricultural Research

Vol.I

Pi,aT3 XXVIII. Sweet potatoes inoculated with Plenodomus destruens: A,
Inoculated at the end; B , a section of A showing extent
of rot; C, inoculated at the side; D, section of C showing

the extent of rot .

XXIX. Fig. i.* — A seedling of Hopi maize with mesocotyl 18 cm.

long. Fig. 2. — The root system of a plant of Zuni maize
dug from a field near Zuni, N. Mex., showing the well-
developed, single seminal root and the comparatively

feeble nodal roots .

XXX. Fig. i. — A hill of Hopi maize containing 15 plants grown
under conditions of extreme drought at the base of the
First Mesa near Polacca, Ariz. Fig. 2. — A plant of Hopi

maize .

XXXI. Fig. 1 . — A field of Zuni maize near Zuni, N. Mex. Fig. 2 . —
A hill of Zuni maize. Fig. 3. — A hill of Hopi maize
making luxuriant growth under conditions of extreme

drought .

XXXII. Fig. 1. — A single plant of Navajo maize grown under irri¬
gation at Shiprock, N. Mex. Fig. 2. — The basal portion
of the plant of Navajo maize shown in figure 1, with

leaves and husks removed . .

XXXIII. Fig. 1 . — Pecan nuts infected with the anthracnose fungus by
spraying with a distilled water suspension of conidia,
showing the appearance nine days after inoculation.
A , Four check nuts, two punctured with sterile needle
and two unpunctured. B, Four nuts inoculated upon
the unpunctured surface of the hull. C, Four nuts
inoculated after puncturing the surface of the hull with
a sterile needle. Fig. 2. — Three of the infected nuts
shown in figure 1 after further development of the

acervuli .

XXXIV. Yellow Newtown apples infected by needle puncture with
conidia of the anthracnose fungus from pecan and apple,
showing appearance four days after inoculation. Fig.
A . — Check apples punctured by sterile needle. Fig. B. —
Apples infected by needle punctures with strain 150 from
the apple. Fig. C.— Apples infected with strain 123 from
a diseased pecan hull. Fig. D. — Apples infected with

strain 125 from a diseased pecan hull .

XXXV. Yellow Newtown apples infected by needle puncture with
conidia of the anthracnose fungus from pecan and apple,
showing appearance four days after inoculation. Fig.
A . — Check apple punctured by sterile needle. Fig. B. —
Apple infected with strain 125 from the pecan nut.
Fig. C. — Apple infected with strain 123 from the pecan
nut. Fig. Z>. — Apple infected with strain 150 from the
apple. Fig. E. — Apple infected with strain 146 from the
pecan leaf. Fig. F. — Apple infected with strain 158, a
reisolation of strain 125 after passage through the apple. .
XXXVI. Crown-gall (caused by Bacterium tumefaciens Sm. and
Town.) on pecan nursery trees from southern Mississippi.
Fig. 1. — The soft type of gall. Fig. 2. — The hard type of
gall .

Page

274

302

33s

338

Oct., i9i3-JVIar., 1914

Illustrations

Page

Plate XXXVII (colored;. Fig. A. — A pecan leaflet infected with the brown
leaf-spot fungus (Cercospora fusca, emend, sp.) from pure
culture. Fig. B. — A pecan leaflet infected with the
anthracnose fungus ( Glomerella cingulata (Stonem.) S.
and v. S.) from pure culture. Fig. C. — View of upper
surface of a pecan leaflet recently infected with the
nursery-blight fungus ( Phyllosticta caryae Peck) from
pure culture. Fig. D. — A pecan kernel infected with
the kernel-spot fungus ( Coniothyrium caryogenum, n. sp.)
from a pure culture, showing the appearance eight days
after inoculation. Fig. E. — A pecan kernel with the
kernel-spot from natural infection. Fig. F. — A pecan
nut infected with the anthracnose fungus from pure cul¬
ture. Fig. G. — The nursery-blight fungus upon syn¬
thetic agar after two weeks. Fig. H. — The nursery-
blight fungus on corn-meal agar after two weeks. Fig.

I. — Viewsof the upper and lower surfaces of pecan leaflets,
showing an advanced stage of the nursery-blight. Fig. J. —

The brown leaf-spot fungus on synthetic agar after four
weeks. Fig. K. — The brown leafspot fungus on corn-

meal agar after four weeks . 338

XXXVIII. An oak ( Quercus gambelii) inoculated with Diplodia longis -

pora at X when dormant . * . 346

XXXIX. Injury caused by potato weevils. Fig. 1. — A section of a
potato from Peru, showing the larva of Rhigopsidius tucu -
manus in its burrow. Fig. 2. — A section of a potato,

showing the burro wings of Rhigopsidius tucumanus . 352

XL. Rhigopsidius tucumanus Heller. Fig. 1. — Dorsal view. Fig.

2. — Ventral view . 352

XLI. Figs. 1 and 2. — Premnotrypes solani Pierce. Fig. 1. — Dorsal
view. Fig. 2. — Ventral view. Fig. 3. — Trypopremnon
latithorax Pierce. Dorsal view . 352

XU I (colored). Sketch map showing the distribution and relative
areas of the different types of vegetation in Tooele Val¬
ley, with detail showing depressions covered with salt-
flat vegetation alternating with ridges bearing greasewood-

shadscale vegetation . 418

Xlylll. Fig. 1. — Salt-flat vegetation bordering Great Salt Lake with
a greasewood-shadscale ridge in the foreground, a pure
stand of Salicornia utahensis at the right and hummocks
covered with Allenrolfea occidentals in the background.

Fig. 2. — Sagebrush association and islands of Kochia

vegetation in the upper part of Tooele Valley . 418

XLIV. Sagebrush (Artemisia iridentata). Fig. 1. — A good stand and
growth, showing the typical appearance of this association
where the conditions are relatively favorable. Fig. 2. —

Plants showing the root habit; photographed at the edge
of a deep “ arroyo ’ * where the soil had recently caved in . . 418

XLV. Fig. 1. — Sagebrush land which has recently been burned
over, showing scattered, dead plants of Artemisia triden-
tata (no living ones), a dense growth of the annual grass
Bromus tectorum , and scattered plants of Gutierrezia
sarothrae. Fig. 2. — An advanced stage in succession on
sagebrush land which has been under cultivation, with

Journal of Agricultural Research

Vol. i

Page

numerous young plants of Artemisia tridentata and a
dense herbaceous covering of Bromus tectorum and alfilaria
(. Erodium cicutarium). Fig. 3. — Sagebrush reestablished
on land which has been in cultivation and the original,

undisturbed sagebrush vegetation . 418

Plate XL VI. Fig. 1. — Line of contact between the sagebrush association

and the Kochia association, showing the characteristically
sharp demarcation of the two types. Fig. 2. — A typical
view of the Kochia association, with plants rather far
apart and very uniform in size and appearance. Fig. 3 . —

Plants of Kochia vestita , 4 or 5 inches high, and the grass
Poa sandhergii, which is usually associated with the Ko¬
chia in land that is not grazed . 418

XL VI I. Fig. 1 . — Typical shadscale vegetation, consisting of a nearly
pure stand of A triplex confertifolia , showing much dead-
wood, as is usually the case, but the stand is denser than
in much of the area occupied by this association.

Fig. 2. — Transition area between the shadscale and the
grease wood-shadscale types of vegetation. Fig. 3. — Salt
grass ( Distichlis spicata) covering the whole of the depres¬
sion to the right with the exception of a colony of Allen-

rolfea in the middle distance . 418

XLVIII. Fig. 1. — Salt-flat vegetation, Allenrolfea community.

Fig. 2. — Salt-flat vegetation, showing plants of Salicornta
uiahensis . Fig. 3. — Grass-flat vegetation, Sporobolus-

Chrysothamnus community, showing a species of rabbit

brush, associated with tussock grass . 418

XLIX. Citropsis Schweinfurthii grafted on grapefruit stock ( Citrus

decumana ), showing vigorous growth made in 2% years. . 436

L. Fig. 1. — Yellow Bellflower apple tree in full bloom on
April 16, 1912, showing effect of spraying with a solution
tion of nitrate of soda plus caustic potash on February
2 previous. Fig. 2. — Unsprayed check tree for com¬
parison with figure 1 . 444

LI. Fig. 1. — A branch from a Yellow Bellflower tree in full
bloom on April 10, 1913, showing the effect of spraying
with a solution of nitrate of soda plus caustic soda on
February 3 previous. Fig. 2. — A branch from an un¬
sprayed check tree for comparison with figure 1 . 444

LII. Fig. 1. — Split radial face of a creosoted hickory block,
showing tyloses in a large vessel. Fig. 2.— Tangential
section of Aesculus octandra (yellow buckeye), showing
two tyloses which have grown out of one medullary-ray
parenchyma cell. Fig. 3. — Cross section of valley oak,
a white oak, showing young tyloses next the bark in

vessels . 470

LIII. Fig. 1. — Cross section of a white oak, showing fully devel¬
oped tyloses in the large vessels. Fig. 2. — Radial-lon¬
gitudinal view, quarter-sawed surface, of the white oak
shown in figure 1, showing complete closing of the vessel.

Fig. 3. — Cross section of sapwood of pignut hickory,
showing fully developed tyloses. Fig. 4. — Radial view
of mesquite, showing “gum” droplets and formation
often stimulating tyloses

470

Oct., 1913-Mar., 1914

Illustrations

Plate

Page

LIV. Cross section of cow oak, a white oak, showing normal and
abnormal tyloses. Fig. 1. — Wound tyloses induced by
the felling of the tree and the sudden cessation of sap
flow. Fig. 2. — No tyloses; empty vessels. Fig. 3. —

Young and well-developed normal tyloses . 470

LV. Fig. 1. — Cross section of a diffuse porous wood, yellow pop¬
lar or tulip, showing scattered tyloses. Fig. 2. — Cross
section of a ring porous wood, osage orange, with vasi-
centric parenchyma, showing abundantly developed

tyloses . 470

I/VI. Fig. i.~ Cross section of western white pine, showing ray
tyloses, closed vertical resin canal in young sap wood, and
nuclei visible in epithelial cells of canal which is begin¬
ning to split open. Fig. 2. — Tangential section of Nor¬

way pine, showing ray tyloses . 470

I/VTI. Fig. 1. —Cross section view of shortleaf pine, showing open
and partly closed vertical resin canals. Fig. 2. — Heart-
wood of Sitka spruce, showing closed vertical canal . 470

I/VIII. Open and closed horizontal canals in sapwood. Fig. 1. —

Open canal in tamarack. Fig. 2. — Partly closed canal
with distended epithelial cells in Douglas fir. Fig. 3. —
Young canal which has never opened in western white
pine. Fig. 4. — Open canal in red spruce surrounded by
thick- walled epithelium. Fig. 5. — Partly closed canal
in red spruce. Fig. 6. — Closed canal in Engelmann

spruce . 470

LIX. Fig. 1. — hog from collection of woods in the Forest- Products
Laboratory — a specimen of the material used in this study.

Fig. 2. — Specimens of woods showing creosote penetrance
in sapwood and heartwood as affected by tyloses. Speci¬
men A. — Red oak. Specimen B. — White oak. Specimen

C. — Pignut hickory . 470

LX. Fig. 1. — River birch with bark removed, showing larval
mines of Agromyza pruinosa. Fig. 2. — Section through
wood of river birch, showing “ pith-ray flecks' * produced
by the work of Agromyza pruinosa . 474

LXI. Fig. 1. — Agromyza pruinosa: Larva and details. Fig. 2. —
Agromyza pruinosa: Pupa. Fig. 3. — Agromyza pruinosa:

Adult male. Fig. 4. — Agromyza pruinosa: Abdomen of
adult female, showing ovipositor. Fig. 5. — Sympha agro-

myzae: Adult . 474

LXII. Fig. 1. — Wheat seedlings from seed inoculated with spores
of Helminihosporium gramineum and from seed exter¬
nally sterilized. Fig. 2. — Barley seedlings from seed
inoculated with Helminihosporium gramineum and from
sterilized seed. Fig. 3. — Wheat seedlings from seed in¬
oculated with spores of Fusarium culmorum from oat
seedlings and from seed externally sterilized. Fig. 4. —

Barley seedlings from seed inoculated with spores of
Fusarium culmorum from oat seedlings and from seed
externally sterilized. Fig. 5. — Oat seedlings from seed
inoculated with spores of Fusarium culmorum from oat
seedlings and from seed externally sterilized

490

xii Journal of Agricultural Research voi.i

Page

Plats LXIII. Root systems of wheat seedlings grown in 6-inch pots from
seed externally sterilized and from seed inoculated with
Helminihosporium gramineum from wheat . 490

TEXT FIGURES

Citrus Ichangensis, a Promising, Hardy, New Species from Southwestern
China and Assam.

Fig. i. — Citrus ichangensis , n. sp.: A, Pistil after the petals and stamens
have dropped but before the style has fallen off; B, stamen as
seen from one side ; C, two seeds deformed by mutual pressure . . 1

2. — Citrus ichangensis , n. sp.: Fruit showing the very low, broad,

apical papilla circumscribed by a shallow furrow . 2

3 — Citrus ichangensis , n . sp . : A , Cross section of a large fruit ; B , seeds . . 3

4. — Citrus ichangensis: A , Calyx of dwarf wild form and pedicel with

bracts; B , young ruit; C, flower bud and pedicel with bracts. . 4

5. — Citrus ichangensis: Flowering branch from the type specimen. ... 5

6. — Citrus ichangensis: Flowering branch of dwarf wild form . 6

7. — Citrus ichangensis: Flowering branch from a paratype in the her¬

barium of the Arnold Arboretum . 7

Cysticercus Ovis, the Cause of Tapeworm Cysts in Mutton:

Fig. i. — Cysticercus ovis: Hooks . 17

2. — Cysticercus ovis: Head and neck . ig

3. — Cysticercus ovipariens C=C. ovis): Fragment of head . 19

4. — Cysticercus ovipariens l=C. ovis): Hooks . 19

5. — Cysticercus ovipariens ( —C . ovis) : Papillae on caudal bladder. . . 29

6. — Hooks of Taenia ovis, T. hydaiigena , T. solium , T. balaniceps , and

T. krabbei . 30

7. — Sexually mature segment of Taenia ovis . 31

8 . — Sexually mature segment of T aenia hydaiigena . 31

9. — Gravid segment of Taenia ovis . 32

10. — Gravid segment of T aenia hydaiigena . 32

11. — Surface of caudal bladder of Cysticercus ovis showing papillae. . 33

12. — Surface of caudal bladder of Cysticercus tenuicollis showing trans¬

verse furrows . 33

13. — Cysticercus ovipariens (~C. ovis): a , Hook; b, cyst containing

cysticercus cut across . 38

The Serpentine Leaf -Miner:

. Fig. i. — The serpentine leaf -miner (Agromyza pusilla ): a, Adult; b, side
view of head; c, side view of thorax, showing characteristic
color pattern; d , dorsal view of abdomen, melanic phase; e,
outline of thorax, showing location of characteristic bristles. . . 59

2. — Map showing known distribution of the serpentine leaf -miner

throughout the world . 61

3. — Map showing distribution of the serpentine leaf-miner within the

United States . 62

4. — Alfalfa leaf with eggs of the serpentine leaf-miner in situ, a, Egg,

greatly enlarged; b , same, in situ, with parenchyma of leaf
partly dissected away . 66

5. — Larva of the serpentine leaf -miner, lateral view . 68

6. — Puparium of the serpentine leaf-miner, ventral view . 68

7. — Mouth armature of larva of the serpentine leaf-miner . 70

8. — Diagram showing the range in temperature throughout the year

at three widely separated localities at which observations were
made on the serpentine leaf-miner . 73

9. — Diaulinus begini , a parasite of the serpentine leaf-miner. At left,

hind leg of Diaulinus websteri . 78

10. — Larva of Diaulinus begini . 79

11. — Pupa of Diaulinus begini . 79

12. — Chrysocharis parksi , a parasite of the serpentine leaf-miner, a,

Middle and hind legs of Chrysocharis ainsliei . 80

13. — Zagrammosoma multilineata , a parasite of the serpentine leaf-

miner . . 81

Oct., 1913-Mar., 1914

Illustrations

XIII

The Serpentine Leaf-Miner — Continued : Page

Fig. 14. — Pleurotropis rugositkorax , a parasite of the serpentine leaf-miner. 81

15. — Agromyza angulata . 84

16. — Puparium of Agromyza angulata , with lateral view of anal appen¬

dages at left . 84

17. — Agromyza coquilletti . 85

The Occurrence of a Cotton Boll Weevil in Arizona :

Fig. 1. — Anthonomus grandis , var. thurberiae: Prothorax . 91

2. — Anthonomus grandis Boh.: Prothorax . 91

3. -—. Anthonomus grandis , var. thurberiae : Head and beak: A , Female;

B, male . 92

4. — Anthonomus grandis Boh.: Head and beak: A, Female; Bt male. 92

5. — Anthonomus grandis , var. thurberiae : Antenna of female . 93

6. — Anthonomus grandis Boh. : Antenna of female . 93

7. — Anthonomus grandis , var. thurberiae : A, Front leg; B, middle leg;

C, hindleg . 94

8. — Anthonomus grandis Boh.: A, Front leg; B , middle leg; C, hind

leg . 94

9. — Anthonomus grandis , var. thurberiae: Wing . 95

Individual Variation in the Alkaloidal Content of Belladonna Plants:

Fig. 1 . — Diagram showing the percentage of alkaloids in the leaves of indi¬
vidual belladonna plants at the Arlington Experimental Farm,

Va., during the seasons of 1911 and 1912 . 144

The Pubescent-Fruited Species of Prunus of the Southwestern States:

Fig. 1. — Map of the southwestern part of the United States, showing the
range of Prunus andersonii, Prunus fascieulata, and Prunus
eriogyna , n. sp . 149

2. — Map of Texas, showing the known areas and probable range of

Prunus minuiiflora and Prunus texana . 151

3. — Prunus texana Dietr. : A , Section of calyx; B , detail of calyx lobes,

showing glandular margins; C, section of calyx from flower of the
horticultural variety Ramsey, P. texana X Wild Goose plum. . 155

4. — Prunus andersonii Gray: A, Petal; B, section of a flower; C, calyx

showing ciliate margins; D, E, dried fruit; F , G , stone . 165

5. — Prunus eriogyna , n. sp.: A, Section of calyx; B, detail of portion

of calyx with petals, from outside, showing glandular ciliation
of lobes; C, twig showing angular habit of branching, leaves and
fruit attached . 169

6. — Prunus fascieulata Gray: A, Section of staminate flower, showing

abortive ovary and minute hairs on interior of calyx; B , calyx
cup, pistillate form, showing abortive stamens; C, detail of calyx
lobe; £>, fecundated ovary; E, F, G , fruits, three forms; Hy /, /,
seed, dorsal, ventral, and side views . 17 1

7. — Prunus minutiflora Engelm. : A , Section of flower of pistillate form,

showing well-developed pistil and abortive stamens; B, section
of flower, staminate form, showing well-developed stamens and
abortive pistil; C, detail of calyx lobes and petals . 173

8. — Prunus microphylla Hems. : A , Section of staminate flower, showing

well-developed stamens and abortive pistil; Bt detail of calyx
from outside; C, twigs showing leaves and fruit; Dt fecundated
ovary . 176

Selective Adsorption by Soils:

Fig. 1. — Curves showing the effect of concentration on the selective adsorp¬
tion of potassium from solutions of potassium by Norfolk sandy

loam and by Marshall silt loam . 185

2 . — Curves showing the effect of the presence of sodium nitrate and
calcium phosphate on the selective adsorption of potassium from
solutions of potassium chlorid . 187

A Bacterium Causing a Disease of Sugar-Beet and Nasturtium Leaves:

Fig. i. — Bacterium aptatum from a 2-day beef-bouillon culture stained with

carbol fuchsin . 195

2. — Filaments of Bacterium aptatum taken from the condensation water
from a 2-day-old agar culture; stained with carbol fuchsin: a,
Segmented; 6, unsegmented . 195

XIV

Journal of Agricultural Research

Vol.I

A Bacterium Causing a Disease of Sugar-Beet and Nasturtium Leaves — Contd. : page
Fig. 3. — Process of cell division as seen in an 18-hour-old hanging drop cul¬
ture of Bacterium aptatum . 195

4. — Bacterium aptatum showing flagella from a 2-day-old agar culture;

stained with Loeffler’s flagella stain . 196

5. — Camera-lucida drawing of a portion of a cross section of sugar-beet

leaf inoculated with Bacterium aptatum . 206

The Calliephialtes Parasite of the Codling Moth:

Fig. i. — Calliephialtes sp.: Ventral view of terminal abdominal segments,
showing relative position of elements of ovipositor, a, Valves
of sheath; b , lance; c » lancets; d, cerci . 216

2. — Calliephialtes sp.: Lateral view of terminal abdominal seg¬

ments, showing relative position of elements of ovipositor,
o, Valves of sheath; 6, lance; c, lancets; c?, cerci . 216

3. — Calliephialtes sp. : Lateral view of tips of elements of ovipositor.

a, Sheath; b , lance; c , lancet . 217

4. — Calliephialtes sp.: Ventral view of male genitalia, a, Sheath;

b, penis; c, clasper; d, genital palpus; e , cardo . 217

5 . — Calliephialtes sp . : Ventral view of clasping organ of male genitalia.

a , Basal portion; b , clasper; c, genital palpus . 217

6. — Calliephialtes sp.: Egg . 219

7. — Diagram showing relation between incubation period of eggs of

Calliephialtes sp. and average mean temperature at Vienna,

Va., 1912 . 220

8. — Calliephialtes sp.: Dorsal view of newly hatched larva . 221

9. — Calliephialtes sp. : Ventral view of head of newly hatched larva . . 221

10. — Calliephialtes sp.: a, Full-grown larva; b, face . 221

11. — Diagram showing relation between temperature and larval period

of males and females of Calliephialtes sp. in the cocoon at
Vienna, Va., 1912 . 224

12 . — Calliephialtes sp . : Prepupa of female . 225

13. — Calliephialtes sp. : Beginning of exuviation of female pupa . 226

14. — Calliephialtes sp.: Pupa of female and tip of abdomen of male

pupa . 226

15. — Diagram showing relation between pupal period of Calliephialtes

sp. and temperature . 228

The Foot-Rot of the Sweet Potato:

Fig. 1 . — Graphic representation of growth on rice at different temperatures . 2 70

A Drought-Resisting Adaptation in Seedlings of Hopi Maize:

Fig. 1. — Diagram of seedling maize plant, giving terminology of parts _ 294

2. — Diagram showing the average size of seedlings of Chinese, Boone

County White, and Navajo maize planted at different depths. . 296

Some Diseases of Pecans:

Fig. 1 . — Cross section of pecan leaf recently infected with the nursery-blight

fungus ( Phyllosticia carycte Peck) . 306

2 . — Horizontal section of leaf recently infected with the nursery-blight

fungus . 31 1

3. — Cross section of a leaf infected with the brown leaf-spot fungus. . 314

4. — Diagram showing measurements in length of 200 conidia . 318

5. — Diagram showing measurements in width of 200 conidia . 319

6. — The anthracnose fungus upon corn-meal agar: A, Acervulus; B,

conidia; C, ascus . 328

7. — Diagram showing ascospore measurements of the anthracnose

fungus : A , Length of 1 50 ascospores ; B , width of 1 50 ascospores . 328

8. — Diagram showing conidial measurements of the anthracnose fun¬

gus: A , Length of 150 conidia; Bt width of 150 conidia . 329

A Twig Blight of Quercus Prinus and Related Species:

Fig. 1. — Diplodia longispora: A section of a pycnidium . 340

2. — Diplodia longispora: Stages in development of spore. A , Macro-

phoma stage; B, Diplodia stage; C, Diplodia spore with two
septa . 340

3. — Diplodia longispora: Sclerotial bodies formed in artificial media. 343

4. — Diplodia longispora: A section showing grouping of pycnidia , . . 344

Oct., 1913-Mar., 1914 Illustrations xv

A Twig Blight of Quercus Prinus and Related Species — Continued: page

Fig. 5. — Diplodia longispora: Types of germination. A, B, Germ tubes

from end of spore; C, germ tube from side of spore . 344

6. — Diplodia longispora: A portion of mycelium showing the coalescing

of the hyphae . 345

7. — Diplodia longispora: A portion of mycelium with chlamydospore-

like bodies . 345

New Potato Weevils from Andean South America:

Fig, 1. — Premnotrypes solani Pierce: Lateral view of prothorax and beak. 348

2. — Premnotrypes solani Pierce: Frontal view of beak . 348

3. — Trypopremnon latithorax Pierce: Lateral view of thorax and beak. 349

Indicator Significance of Vegetation in Tooele Valley, Utah:

Fig. 1. — Curve showing the relation between the salt content (in percent¬
ages of the dry weight of the soil) and the specific electrical
resistance (in ohms) of the soil when saturated with water. _ 368

2. — Monthly distribution of precipitation at Tooele, Utah (mean for

15 years) . 369

3. — A representative 10-meter quadrat of the sagebrush association,

showing the location of each individual of Artemisia tridentata
and of Gutter rezia sarothraet these being the only woody species
present . *. . 379

4. — Artemisia tridentata (sagebrush): A, Detail showing the wedge-

shaped, 3-toothed leaves by which this plant is easily recog¬
nized; B, a small plant growing where hardpan occurred, show¬
ing the deflection of the taproot from a vertical to a horizontal
direction after reaching a depth of 5 inches . 381

5. — A small plant of sagebrush ( Artemisia tridentata ), showing the

deeply penetrating taproot and good development of superfi¬
cial lateral roots typical of this species . 384

6. — A representative 10-meter quadrat of the Kochi a association,

showing the location of each tuft of Kochia and of each matlike
colony of Poa sandbergii . 390

7. — Kochia vestita: A , Detail showing the narrow, hairy leaves; B, a

plant showing the shallow root system and the propagation by
root shoots . 392

8. — A representative 10-meter quadrat of the shadscale association,

showing the location of each individual plant of Atriplex con -

Solia , the only woody species present,, and of Opuntia sp. . 395

x confertifolia (shadscale): A , A typical plant, showing the
thick, vertical taproot and the widespreading lateral roots; B,
detail of a fruiting branch , showing the shape of the leaves and

of the bracts, or scales, which envelop the fruits . 398

10. — Sarcobatus vermiculatus (greasewood) : A , Detail showing the nar¬
row, rather fleshy leaves; B, a plant showing the excellent root

development . . . 404

11. — Allenrolfea occidentals: A, Detail of a fruiting branch, showing the
cylindrical, fleshy, practically leafless stems; B , a plant show¬
ing the large taproot and rather scanty lateral roots character¬
istic of this species . 409

12. — A representative 10-meter quadrat of the Allenrolfea community
(salt-flat association), showing the location of each individual

plant of Allenrolfea occidentals > the only species present . 410

13. — Diagram showing the characteristic root development of the dom¬
inant species of each of the principal types of vegetation of
Tooele Valley and indicating the average conditions of soil
moisture and salinity of the corresponding types of land . 412

Citropsis, a New Tropical African Genus Allied to Citrus:

Fig. 1. — Citropsis Schweinfurthii: A branch showing 3-foliate and 5-foliate
leaves, leaflike petioles, and rachis segments; also paired and

single spines in the axils of the leaves . 419

2. — Citropsis Schweinfurthii: Young seedlings germinated in Wash¬
ington, D. C., from seed from Budongo Forest, Uganda, Africa.

A, Young seedling, showing the first pair of leaves, succeeded
by alternate simple leaves, and finally unifoliate leaves; B and
D, young seedlings, showing the first foliage leaves, which are
opposite; C, a single one of the pair of first foliage leaves . 422

XVI

Journal of Agricultural Research voti.oct., 1913-Mar., 1914

Citropsis, a New Tropical African Genus Allied to Citrus — Continued: page

3. — Citropsis Preussii: Flowers after petals and stamens have fallen;

leaves, one trifoliate and one having the terminal leaflet borne
on a winged segment of the rachis . 424

4. — Pistils of four species of Citropsis. A , Citropsis Preussii; Bt Citrop¬

sis mirabilis; C> Citropsis Schweinfurthii; and D, Citropsis gabu -
nensis . 425

5. — Citropsis Schweinfurthii: Nearly mature fruit; A, side view, show¬

ing calyx and disk; B , section showing four cells with pulp
vesicles and three seeds . 427

6. — Citropsis Schweinfurthii : Cluster of flowers, showing stamens ar¬

ranged to form a staminal tube . 429

7. — Citropsis Schweinfurthii: A trifoliate leaf from the type specimen,

showing double spines in the axils and pronounced serrations
of the leaflets toward the tips . 430

The Origin of Some of the Streptococci Found in Milk:

Fig. 1. — Cells of streptococci, showing variation in size and morphology. . 494

2. — Types of cells of streptococci . 495

3. - — Curve showing the typical rate of fermentation of dextrose and

glycerin . 496

4. — Frequency curves showing acid formation in dextrose broth . 502

5. — Graphic representation of the characters of cultures of streptococci

from milk and from bovine feces . 503

6. — Graphic representation of the characters of cultures of streptococci

from the mouths of cows and from infected udders . 504

7. — Diagram showing the fermentation reactions of two types of udder

cultures of streptococci . 506

8. — Diagram showing a possible grouping of the milk cultures of strep¬

tococci . 507

FOREWORD

The recent advances in the theory and practice of agriculture have come
almost entirely from scientific research applied to agricultural problems .
Accumulated results of centuries of painstaking studies have been drawn
upon , and it has become evident that further improvement in agriculture
calls for continued investigation of the most accurate and thorough nature.
The first recognition of the economic value of progress in these investiga¬
tions as well as the initial application of theories to practical problems
comes usually from specialists . Indeed , only in rare instances is the
significance of the results of scientific research apparent to farmers , since
newly discovered fads are seldom directly applicable to agricultural
conditions .

The suggestive or the indirect value of reports of new work is usually
of paramount economic importance; it is the purpose of the Journal of
Agricultural Research , therefore , to record investigations bearing directly
or indirectly upon economic conditions of agriculture. It is hoped that
permanence of record of new data may be secured by sending the Journal
in its entirety to special libraries and institutions which make suitable
exchanges and that a liberal distribution of the reprinted papers to inter¬
ested specialists may enhance the usefulness of the separate articles.

The first few issues will contain papers from the Department of Agri¬
culture only. Plans , however , are now being perfected in accordance with
the tentative suggestions made to the Secretary of Agriculture by the execu¬
tive committee of the Association of American Agricultural Colleges and
Experiment Stations so that articles prepared and submitted by investi¬
gators in the State agricultural colleges and experiment stations will eventu¬
ally be included in the Journal.

B . T. GALLOWAY ,

Assistant Secretary of Agricidture .

Washington , D. C.,

October r, 1913. r

JOURNAL OF AGRIffllTURAL RESEARCH

DEPARTMENT OF AGRICULTURE

Voiv. I Washington, D. C., October io, 1913. No. 1

CITRUS ICHANGENSIS, A PROMISING, HARDY, NEW
SPECIES FROM SOUTHWESTERN CHINA AND ASSAM

By Walter T. Swingle, '

Physiologist in Charge of Crop Physiology and Breeding Investigations ,

Bureau of Plant Industry

INTRODUCTION

A study of the wild relatives of the orange begun a few years ago in the
hope of finding new material for use in hybridization or as stocks has
resulted in bringing to light a number of very interesting wild species,
some of them new and many of
them very little known. One of
the most remarkable of these is
a wild Citrus, native to south¬
western China. This species is
cultivated in the vicinity of
Ichang, and it bears a very large
lemonlike fruit that is of suffi¬
ciently good quality to cause it
to be shipped to markets several
hundred miles distant. It grows
wild farther to the north and at
a higher altitude than any other
species of Citrus and is undoubt¬
edly very hardy, which makes it
of great promise for use in breed¬
ing cold-resistant citrous fruits.

Because of its unusually large
seeds it promises to yield very vigorous seedlings and to be, in conse¬
quence, a useful stock on which to graft oranges, lemons, and other
cultivated species of the genus.

Mr. Augustine Henry collected excellent material of this species
around Ichang, China, from 1885 to 1888. His specimens are found in
many herbaria under the name 11 Citrus medica L., var.” The best
specimens, however, are those collected by Mr. E. H. Wilson, first in
1900 to 1903 for Veitch & Sons, and again in 1907 for the Arnold
Arboretum, this latter material comprising an abundance of flowering
specimens, young fruits, and also ripe fruits in alcohol.

Fig. i. — Citrus ichangensis , n. sp.: At Pistil after the
petals and 3 tamens have dropped but before the
style has fallen off; from a para type in the herbarium
of the Arnold Arboretum; E. H. Wilson No. 2230A;
2i times natural size. B , Stamen as seen from one
side; from a paratype in the herbarium of the Arnold
Arboretum; E. H. Wilson No. 2230A; 2^ times nat¬
ural size. C, Two seeds deformed by mutual pres¬
sure; from a paratype in the National Herbarium;
A. Henry No. 3423 (?), bottle A; natural size.
(Drawn by J. M. Shull.)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(1)

Vol. I, No. 1
Oct. 10, 1913
G— 1

Journal of Agricultural Research

Vol. I.No.i

The director of the Arnold Arboretum, Prof. C. S. Sargent, has very
kindly turned over to the writer all this valuable material. Thanks are
also due to Mr. E. H. Wilson for furnishing very full notes about his
specimens and for his observations on the use of this species as a substitute
for the lemon.

In China this species occurs in an undoubted wild state in the hills of
the Upper Yangtze Valley from Ichang west and southwest in Hupeh,

Szechwan, and Kweichow,
growing at altitudes of
1,500 to 6,000 feet. In
Assam a closely related
but slightly different form
is found at an altitude of
5,000 to 6,000 feet in the
Khasi Hills. Doubtless
other similar forms occur
to the eastward in that
province and in Upper
Burma as well. The spe¬
cies thus ranges over a
region at least 1 ,500 miles
long and some 500 miles
wide.

This plant is reported
in all parts of its range as
growing in a truly wild
state and is cultivated on a
small scale around Ichang
along the Yangtze River,
where the fruit is called
the “ Ichang lemon” by
foreigners.

TECHNICAL DESCRIPTION OF CITRUS ICHANGENSIS

Fig. 2. — Citrus ichangensis , n. sp.: Fruit showing the very low,
broad, apical papilla circumscribed by a shallow furrow; from
a paratype in the National Herbarium; E. H. Wilson No. 4736;
natural size. (Drawn by J. M. Shull.)

Citrus ichangensis is strikingly unlike any other Citrus native to China
and is easily distinguished from all its congeners. Its technical descrip¬
tion is as follows : 1

i Citrus ichangensis, sp. nov.— Citrus foliis augustis, latitudine 4 plo vel 6 plo longioribus, petiolis
atealatis, obovatis vel oblongis ad basin abrupteattenuatis, laminis ovato-acuminatis, vix petiolis aequanti-
bus, fioribus grandibus, 5-meris, staminibus 20, connatis, polyadelphiis, seminibus numerosis, grandibus.

Frutex vel arbor 1-10 metralis (plerumque 1-5 met.); rami juniores angulati saepe spinosisimi, 2-4 mm.
diameter. Folia angusta, 60-135X15-33 mm. (plerumque 80X 1 15-10-30 mm. ), petiolis late alatis, laminis
saepe aequantibus vel superantibus, obovatis ellipticis vel oblongo-spathulatis ad basin abrupte attenuatis,
apice regulariter rotundatis vel truncatis vel subcordatis; laminis ovato-acuminatis plus minusve caudatis
apice leviter emarginatis, ad basin regulariter rotundatis vel obtuso-cuneatis. Flores grandes, 20-35 mm.
diam., 5-meri, solitarii, axillarii; pedicellis 3-5 mm. longis, calycibus sepalis crassis subtriangularibus,
3X3 mm., margine minute ciliatis; petalis oblongis 15-20X5-8 mm., staminibus 20, connatis, usque ad
apicem cohaerentibus, polyadelphiis in fasciculis 3-5, 8-10 mm. longis, stylis 3-4X1^ mm., caducis; stig-
matibus 2-2 mm. longis, 3 mm. latis ovariis paullo minoribus, ovariis 3X3 mm. , 8-n-locularibus. Fructus
grandis, 7-10 cm. X9-10 cm., ovalis, ad basin tuberculato-sulcatus, apice cum papilla magna vix prominenti,
sulco circular i plus minusve 25 mm. diam. circumdata, cortice crasso 7-9 mm. diam. ; segmentis 8-1 1, pulpa
vesiculari acida, seminibus grandibus 15-20X10-14X7-11 mm. ovato-aeutis, polyembryonicis, 40-70 in
fructusingulo.

Oct. io, 1913

Citrus Ichangensis

Citrus ichangensis Swingle.

A spiny shrub or small tree usually 5 to 15 feet high. Leaves narrow, 4 to 6 times
longer than wide, mostly 80 to 115 by 18 to 30 mm., with very large broadly winged
obovate or oblong spatulate petioles evenly rounded at the tip and narrowed abruptly
at the base, usually 35 to 60 by 20 to 30 mm.; with ovate-acuminate laminae more or
less caudate, emarginate at the tip and evenly rounded or bluntly pointed at the
base, usually 30 to 60 by 18 to 30 mm., often not equaling the winged petiole in area.
Flowers about 25 mm. in diameter, 5-merous; stamens 20, at first all connate to the
tips, finally breaking up into several bundles, about 10 mm. long. Pistil about 10 mm.

Fig. 3.— Citrus ichangensis „ n. sp.t from paratypes in the National Herbarium: E. H. Wilson No. 4737:
A , Cross section of a large fruit; natural size. B, seeds; natural size. (Drawn by J. F, Brewer.)

long; stigma nearly as large as the ovary; style short, caducous; ovary 8 to 11 celled;
ovules numerous in each cell. Fruit large, slightly oval, 8 to n by 7 to 10 cm., with
a rough and furrowed base and a broad very low papilla at the tip, about 25 mm. in
diameter, circumscribed by a shallow furrow; peel rather rough, 6 to 10 mm. thick.
Segments 8 to 11, nearly half filled with seeds; pulp vesicles fusiform, 8 to 12 by
2 to 4 mm. on stalks 2 to 8 mm. long. Seeds very large, usually 16 to 18 by 11 to 12
by 7 to 10 mm., more or less angular from mutual pressure, 40 to 70 per fruit, polyem-
bryonic; cotyledons thick and fleshy. (See PI. I and figs 1 to 7.)

Journal of Agricultural Research

Vol. I, No. i

This species differs from its congeners in having very large thick seeds and slender
leaves four to six times longer than broad, with very large, winged petioles often as
large or larger than the blade. It differs from Citrus hisirix DC. in having oblong
rather than triangular winged petioles and much larger flowers with connate stamens.

Distribution: Central and Southwestern China, i. Hupeh Province1

Ich’ang Prefecture. — Vicinity of Ich’ang. A. Henry, No. 3423, 1887 (?),
“Thorny bush 4 ft., white flowers; 'in a wild jungly place; a wild plant. ”
Flowers, Kew, Paris (Museum), Dahlem, Harvard (Gray Herbarium), Washington,
D, C. (National Herbarium); A. Henry, “Bottle A,” “ fruit
from same shrub as 3423,” 1887 (?); twigs and fruits, Kew;
seeds (fig. 1, C), Washington, D. C. (National Herbarium).
Pingshan Pa (in Ich’ang George, 10 km. [6}i miles] north¬
west of Ich'ang), E. H. Wilson, No. 4736 (small fruit, see
fig. 2), No. 4737 (large fruit, fig. 3), November, 1907, fruits
only (in spirits) from cultivated trees. Harvard (Arnold
Arboretum), Washington, D. C. (National Herbarium).
Ch’angyang (25 km. [15K miles] south-southwest of Ich’ang),
A. Henry, No. 7695, no date. “Shrub 6 to 7 ft.,” fruits,
Kew, sterile twigs, Harvard (Gray Herbarium); Nanto (20
km. [i2}4 miles] northwest of Ich’ang), E. H. Wilson, No.
202, April 25, 1900, flowers, Kew, Dahlem, Harvard (Arnold
Arboretum), New York (Botanical Garden). San-Yu-Tung
Glen, 10 li [4 miles] from entrance (13 km. [8oJ^ miles]
northwest of Ich’ang), E. H. Wilson, No. 2230B, July,
1907, “bushy tree, 15 ft., cultivated.” Fruits (see fig. 4, B ),
Harvard (Arnold Arboretum). Also eight duplicate speci¬
mens for distribution. Hsingshan District (about 17 km.
[io)4 miles] southeast of Hsingshan), 10 li (5.8 km. or 4
miles) below “ Liang-Shan-Kou ” (altitude 1,500 to 2,000
ft.), E. H. Wilson, No. 2230, May 7, 1907, 2 “bush, 3 to 5ft.,
flowers white, ravine,” flowers, Harvard (Arnold Arboretum)

2 sheets.3 (Also 8 duplicate specimens for distribution.) Hsingshan District, about
14 km. north-northwest of Hsingshan, 8 li (4X km. or 3^ miles) beyond “ Li-Er-Kou”
(altitude 4,200ft.), E. H. Wilson, No. 2230A, May 15, 1907, “Citrus, bush or tree,

3 to 20 ft., flowers white, escaped from cultivation, roadside,” flowers (see figs. 1,
A and B , and 4, C), Harvard (Arnold Arboretum). Five duplicate specimens for
distribution.

Fig. 4. — Citrus ichangensis
from paratypes in the her¬
barium of the Arnold Ar¬
boretum: At Calyx of
dwarf wild form and pedi¬
cel with bracts, E. H.
Wilson No. 3307, natural
size; B , young fruit, E. H.
Wilson No. 2230B, natural
size; C, flower bud and
pedicel with bracts, E. H.
Wilson No. 2230A, nat¬
ural size. (Drawn by
Theo. Holm.)

1 The geographic names in China are in southern Mandarin in accordance with the spelling given in
L- Richard’s, 1908, Comprehensive Geography of the Chinese Empire . . . Revised and translated into
English by M. Kennelly. Shanghai, p. 558-639.

2 Mr. Wilson’s diary for this date reads as follows: "In ravine gathered specimens of a wild citrus from
bushes 3 to 5 ft. tall, growing on cliffs of hard limestone.” Photographs of this ravine taken by Mr. Wilson
Jiave been distributed as Nos. 025 and 032.

3 One twig with flowers on one of the sheets is the type (see PI. I and fig. 5). The other specimens of
this same number resemble the type very closely, and some of them very probably were cut from the same
plant, in which case they would be merotypes.

Oct. io, 1913

Citrus Ichangensis

II. Szechw’an Province.

Kw’eichow Prefecture* — Near Wu Shan (35 km. [22 miles] east of Kw’eichow),
A. Henry, No. 7130, no date, fruits, Kew, British Museum, Harvard (Gray Her¬
barium); Kw’eichow Gorges: Fang Hsiang Hsia (Wind Box Gorge), E. H.
Wilson, No. 3307, May 1903, “2 to 3 ft., spontaneous,” 1 flowers (see figs. 4, A, and 6)
Kew, Harvard (Arnold Arboretum).

Fig. 5. — Citrus ichangensis : Flowering branch from the type specimen; E. H. Wilson,
No. 2230; i natural size, (Drawn by J. M. Shull.)

Ch'ung K’ing Prefecture. — Nanchw’an District (about 75 km. [47 miles] south¬
east of Ch’ungk’ing), “HouTs’ao Kou,”2 A. v. Rosthorn, No. 175, July, 1891, “in
dense woods,” sterile twigs, Dahlem; “Huang Ai Shan,”2 A. v. Rosthorn, No. 1264,
sterile twigs, Dahlem.

Suiting Prefecture. — Ch’engk’ow Ping (about 135 km. [84 miles] northeast of
Suiting), R. P. Farges, no date, flowers, Paris (Museum).

1 Wilson, E. H., 1905, referring to this plant and locality, says: “Citrus japonica, ‘ Kumquat,' was common
on the cliffs and evidently spontaneous.” Gard. Chron., s. 3, v. 38, no. 969, p. 65, July 22, 1905.

2 Cf. Diels, Die Flora von Central-China, 1900, Bot. Jahrb. [Engler], Bd. 29, Heft. 3/4, p. 424-425, Dec.
4, 1900.

Journal of Agricultural Research

Vol. I, No. i

III. Kweichow Province.

Kweiyang Prefecture. — K*ai Chow (?) (60 km. [37 miles] north-northeast of
Kweiyang. Altitude 5,577 ft.), M. CavaeeriE, no date, young fruit, Paris (Museum).1

DETAILED DESCRIPTION OF CITRUS ICHANGENSIS

The typical Citrus ichangensis as it occurs in southwestern China is
a small tree or a large shrub, usually 5 to 15 feet high (1.5 to 5 meters),
but sometimes reaching 20 feet. It also occurs wild in fruiting condi¬
tion only 2 to 3 feet high on the cliffs of the Yangtze Gorges.

Fig. 6. — Citrus ichangensis-. Flowering branch of dwarf wild form; E. H. Wilson
No. 2230A; natural size. (Drawn by Theo. Holm.)

The twigs of the current growth are 2 to 4 mm. in diameter and con¬
spicuously angled, as is common in Citrus. The spines are straight,
usually 1 to 2 cm., sometimes 2 to 3 cm. long, and 2 to 3 mm. in diameter
at the base; they occur singly at one side of the axillary buds. (PI. I
and figs. 6 and 7.) Some specimens have very small spines or none at all.
A few nodes at the base of the twig are often spineless.

The leaves are long and slender and remarkable because of the size of
the winged petiole, which is sometimes larger than the blade. The leaves

1 All of the specimens in this list have been studied by the writer and most of them have been photo¬
graphed, so all are to be considered as truly paratypic.

Oct. io, 1913

Citrus Ichangensis

are from 60 to 135 mm., generally 80 to 1 15 mm. long, and from 12 to 32
mm., mostly 18 to 30 mm. wide, the length being usually four or five

Fig. 7- Ctirus tchangensts ; Flowering branch from a paratype in the herbarium of the Arnold Arbore¬
tum; E. H. Wilson No. 2230A; naturalsize. (Drawn by Theo. Holm.)

times the width. The winged petioles are obovate or spatula te oblong,
rather abruptly narrowed into a wingless but sometimes margined base,

Journal of Agricultural Research

Vol.I.No. i

evenly rounded at the tip or sometimes truncate or subcordate, 25 to
72 by 12 to 33 mm., usually 35 to 60 by 20 to 30 mm., the wingless
basal portion being 4 to 5 mm. long and 1^ to 2 mm. in diameter. The
blades are ovate acuminate or elliptical acuminate, evenly rounded or
very bluntly pointed at the base and narrowed into a more or less acumi¬
nate or caudate apex, which is, however, abruptly rounded and usually
emarginate at the very tip, 20 to 66 by 13 to 30 mm., usually 30 to 60
by 18 to 30 mm. (See PI. I and figs. 6 and 7.) The petioles and
laminae have rather numerous slender secondary veins that run nearly
parallel and rather straight almost to the margin, making an angle
with the midrib varying from about 45 0 to nearly 90°. (See fig. 5.) The
internodes are 12 to 30 mm., usually 15 to 20 mm., long.

The flowers are borne singly in the axils of the leaves (alongside of
the spine when present) ; they seldom occur at the end of the twigs.
The flower buds are cylindric or subcylindric, with a hemispherical tip and
a truncate base, all parts being very prominently glandular dotted. (See
figs. 4, 6, and 7.) The pedicels are short and slender, 4 to 6 mm. long,
1 to 2 mm. in diameter, prominently glandular dotted, with a few very
small bracts near the base. The calyx is fleshy, 4 to 6 mm. in diameter;
the sepals are subtriangular, 3 by 3 mm., thick and fleshy, margins
minutely ciliate. The corolla is white; when fully open it is about 25 to
30 mm. in diameter, with cylindric-oval petals 12 to 18 mm. by 8 to 10
mm. wide, and 20 stamens 8 to 10 mm. long cohering almost the whole
length but separating into a few bundles in fully developed flowers. The
anthers are 2 to 3 by 1 to 1^ mm. in size. The pistil is about 10 mm.
long, stout, on a cushionlike disk 2% mm. high and 4 mm. in diameter,
with a subglobose ovary 4 by 4 mm. The style is stout, 4 mm. long, 1 to
1 X mm. in diameter, caducous, leaving a very short basal portion attached
to the fruit. The stigma is large, subglobose, 2 to 3 mm. in diameter,
almost as large as the ovary, which is 8 to 11 celled, with numerous
ovules in each cell. (See fig. 1, A.)

The fruits are subglobose, slightly longer than wide, 8 to 11 cm. (3 % to
4^ inches) long, 7 to 10 cm. {2% to 4 inches) in diameter, with a wrinkled
and furrowed base and an inconspicuous, very low, and broad papilla at
the top, tipped with the persistent base of the style and delimited by a
shallow circular furrow, making a circle about 20 to 35 mm. in diameter,
usually 25 to 30 mm. (See fig. 2.) The fruits look like very large,
short and thick lemons.

The peel is rather rough, resembling that of a large lemon, 6 to 10 mm.
thick, usually 7 to 9 mm. There are from 8 to 11 segments. In a large
11-celled fruit (Wilson No. 4737) the segments are 72 mm. long, 25 to
35 mm. wide, and 20 mm. thick; in a small 8-celled fruit (Wilson No.
4736) from the same locality they are 60 mm. long, 25 mm. wide, and
18 to 22 mm. thick.

Oct. io, 1913

Citrus Ichangensis

The pulp vesicles are fusiform, pointed at both ends, 8 to 12 by 2 to 4
mm., rarely reaching 18 mm. in length, on a slender stalk 2 to 8 or rarely
10 mm. long, attached to the dorsal ovary wall and also along the periph¬
eral half of the membrane dividing the segments. The core is solid, 6 to
10 mm. in diameter, more or less stellate in cross section because of the
thickening of the membranes at their attachment. The center of the
core is less solid than the periphery, where there are small groups of fibro-
vascular bundles opposite the attachment of each membrane.

The seeds are very large, light brown in alcoholic material, very
numerous, from 40 to 70 in a single fruit and from 4 to 10 in a segment.
Usually from 4 to 6 large seeds and sometimes one or more small ones
occur in a segment. The seeds are cuneate ovate in outline seen from
above and oval or subquadrangular seen from the side, 15 to 20 mm.
long, 10 to 14 mm. wide, 7 to 11 mm. thick, mostly 16 to 18 by 11 to
12 by 7 to 1 o mm., with a straight edge 6 to 8 mm. long where attached
to the placenta. (See fig. 3, A and B.) They have a dark-brown cap
8 to 10 mm. in diameter at the base; the outer seed coat is thick, tough,
and cartaceous, while the inner coat is thin and silky. The seeds of the
wild form, collected in the vicinity of Ichang by Henry (No. 3423), are
more angular through mutual pressure than those of the cultivated speci¬
men and are also thicker. (See fig. 1, C.)

There are often two large embryos and usually several small ones in a
single seed. Frequently the cotyledons are greatly deformed by mutual
pressure of the several embryos. It is almost certain from the structure
of the seeds of Citrus ichangensis that the cotyledons remain buried in the
soil during germination, as in all the commonly cultivated species of
the genus.

The dwarfed wild form of the species, found near the eastern end of
the Windbox Gorge just below Kweichow (Wilson No. 3307), grows only
2 to 3 feet high and bears diminutive leaves scarcely over one-third
the size of those of the cultivated form, the petioles being 16 to 23 by
7 to 8 mm. and the blades 7 to 15 by 4 to 7 mm. in size. In striking con¬
trast to the diminutive leaves are the very numerous long spines which
are unusual in showing a slight upward curvature. (See fig. 3.) Doubt¬
less the habitat of this form on semiarid cliffs will serve to explain its
small size.

Fruits collected by Augustine Henry near Ichang, likewise from a wild
form, are remarkable for the fact that the numerous short, thick, £.nd
very large seeds occupy all the space in the segments, leaving room for
scarcely any juice. The seeds are rather narrower in the cultivated form,
but possibly this is in part due to their having an abundance of space in
which to develop.

Still, in all essential characters the cultivated and wild forms agree,
and doubtless the larger, juicier fruit of the cultivated form is due in part
to the better nourishment the tree receives and also in part to the selection

Journal of Agricultural Research

Vol. I, No. i

practiced by the Chinese gardeners, who would naturally have chosen the
most promising of the wild forms to propagate. Unlike many other culti¬
vated citrous fruits, this species shows no evidence of having been hybrid¬
ized; it is rather a selected form of a wild species.

Both the wild and cultivated forms of Citrus ichangensis will be secured
as soon as possible for trial in this country. Careful exploration at higher
altitudes near the northern limit of the species in China should bring to
light exceptionally hardy forms that would be invaluable to breeders of
hardy citrous fruits.

THE RELATIONSHIPS OF CITRUS ICHANGENSIS

Citrus ichangensis stands apart from all the other known members of
the genus. Its huge, thick seeds are unlike anything heretofore known
in Citrus, and its long, slender leaves with their very large, broadly
winged petioles, often exceeding the blade in area, distinguish it at once
from most of its congeners.

Citrus histrix DC., a curious and little-known East Indian species,
also has leaves with broadly winged petioles, often larger than the blades,
but differs greatly from Citrus ichangensis in having very small flowers,
often only 4-parted , with perfectly free stamens. Even the broadly winged
petioles of C. histrix are distinctly different, being more gradually nar¬
rowed toward the base and usually more abruptly truncate at the tip,
making them somewhat triangular in outline, whereas those of the
Chinese species are often oblong or elongate elliptical.

The other species of Citrus having very large, broadly winged petioles,
such as C. celebica Koord., C. papuana Bail., and C. macroptera Montr.,
native to the Malayo-Polynesian region, are apparently closely related to
C. histrix , if, indeed, they are not to be considered as forms of it. They
all agree with C. histrix in having winged petioles more or less triangular
in outline and show no close affinity with Citrus ichangensis .

The bulky seeds of Citrus ichangensis with their large brown caps and
thick deformed cotyledons are so much larger than those of its congeners
that they can not be mistaken for those of any other species of Citrus.
They are much more like those of the African species of hard-shelled
citrous fruits belonging to the genera Balsamocitrus and Aeglopsis.1

PREVIOUSLY PUBLISHED NOTICES OF THE SPECIES

In 1907 L. Diels2 referred to Citrus histrix DC., two numbers collected
by A. v. Rosthorn in Szechwan in 1891, noting that one (No. 1264) had
narrower leaves with inconspicuous venation and the other (No. 175)

1 Stapf, Otto, 1906. Plantae novae Daweanae in Uganda Iectae. Jour. linn. Soc. [London] Bot., v. 37,
p. 50 5, pi. 22.

Swingle, Walter T., 1912. I,e genre Balsamocitrus et un nouveau genre voisin, J3gIopsis. Soc. Bot.
France, t. 58 (s. 4, t. 11), (M&n. 8d.) p. 236 and 241, fig. B and pi. 3.

2 Diels, I,., 1900. Die Flora von Central-China. Bot. Jahrb. [Engler], Bd. 29, Heft. 3/4, p. 424.

Oct. io, 1913

Citrus Ichangensis

broader, distinctly veined leaves. Sterile specimens of both of these
numbers in the herbarium at Dahlem belong undoubtedly to Citrus
ichangensis and differ but slightly in shape and venation.

In 1 91 1 H. L£veill6 published a “Citrus Cavaleriei ” in an article by
Julien Cavalerie 1 without a recognizable description. A specimen collected
by P£re Julien Cavalerie in the Province of Kweichow, China, preserved
in the Museum d’Histoire Naturelle at Paris, is almost certainly Citrus
ichangensis. In his account of the Aurantiacese of Kweichow, he says of
this species :

Citrus Cavaleriei , L6vl. I found in the forest, remote from any habitation in the
vicinity of Ma-Jo and of Kai-Tch6ou [K’ai Chow] at about 1,700 meters [5,577 feet]
altitude, a kind of spiny orange tree, in the undergrowth of the forested slopes. The
tree is arched (voht6) and completely covered with moss. One tree had fruits of the
size of an apricot and flowers at the same time . The fruit is hard and rounded in shape ;
the winged petiole is so much developed that it constitutes half of the leaf. I did not
see this tree cultivated anywhere. It is the only wild species [of Citrus] in the high
regions.2

There is nothing in this description to distinguish this plant from
Citrus histrix DC., and upon applying to M. L6veill£ to see the type
specimen he declared this name to be “a true nomen nudum ” that had
been published by mistake, and a note to this effect was later published.3

A SUBSPECIES FROM THE KHASI HILLS

Several good specimens of a Citrus from the Khasi Hills in Assam,
collected by J. D. Hooker and T. Thomson in 1850 and preserved in the
Kew Herbarium, were at first supposed by the writer to be identical
with Citrus ichangensis , as they showed the same peculiar, very large
and broadly oval or oblong winged petioles. After careful study, how¬
ever, the Khasi specimens were found to differ from the typical Chinese
material in a number of points.

In the first place, all of the Khasi specimens show leaves with less
acuminate blades than those of the Chinese material; moreover, the
leaves of the Indian specimens are distinctly more variable both in size
and in shape. The immature fruits collected by Hooker and Thomson
in this locality are all slightly oblate instead of slightly prolate like the
Chinese fruits from Pingshan Pa (Wilson Nos. 4736, 4737)* The fact
that Hooker and Thomson call this plant a “wild orange” is additional
evidence that the fruits did not have the lemonlike appearance of the
Chinese form. Finally, the flowers in Clarke’s specimen preserved in the
British Museum occur in three to six flowered axillary panicles instead of
singly, as in all the Chinese material seen. The tree reaches a height of

1 Cavalerie, Julien, 1911. Tes Aurantiac^es du Kouy-Tch&ni. Bui. de Gdogr. Bot., t. 21 (ann. 20, s. 4),
no. 261, p. 2x1.

2 Translation from Cavalerie, Julien, 1911, loc. dt.

3 LJeveille], H., 1911. Tes Aurantiacees du Kouy-Tch£ou. Bui. de G6ogr. Bot., t. 21 (ann. 20, s. 4),
no. 262, p. 236.

Journal of Agricultural Research

Vol. I, No. x

30 feet in the Khasi region and has not been recorded over 20 feet in
China. This, however, might easily be due to differences in the expo¬
sure, orange trees growing in a forest often being much taller than those
in the open without shade.

More material and, above all, ripe fruits will be needed to decide defi¬
nitely whether the Khasi “wild orange” belongs to Citrus ichangensis .
It is certainly much more closely related to this latter species than to any
other. For the present it seems best to consider it as a subspecies of
the Ichang lemon. The technical diagnosis is as follows: 1

Citrus ichangensis latipes Swingle.

Differs from C. ichangensis in having the leaves more variable in size and shape
with the tips acute, not caudate, the flowers in few-flowered (3 to 5) panicles instead
of solitary, and the fruits oblate instead of prolate spheroidal in shape.

Distribution: Assam, Northeastern India. Khasi Hills

Living Bridge,2 Hooker and Thomson, September 2, 1850, "small orange, wild”
fruits, Kew; Myrung Wood (altitude 5,700 ft.), J. D. Hooker and T. Thomson, July 6,
1850, "Aurant. Tree 30 ped. alt. Frt size of a walnut,” fruits, Kew; Moflong
(altitude 6,000 ft.), 3 J. D. Hooker and T. Thomson, July, 1850, fruits, Kew; Moflong( ?),
J. D. Hooker and T. Thomson, “Citrus latipes H. f. and T. Regio temp, (indig.)
alt. 5,000-6,000 ped.,”4 no date, sterile twigs, Harvard (Gray Herbarium); Khasi
Hills, C. B. Clarke No. 21879 (Collector Rutton), 1873, flowers, British Museum.

DETAILED description of citrus ichangensis latipes

The leaves of Citrus ichangensis latipes vary greatly in size and shape,
ranging from 65 to 152 by 12 to 48 mm., the length varying from three
to seven times the width. The petioles in particular, though always
broadly winged, are distinctly more variable than in the Chinese material.
They vary from oblanceolate linear to spatulate oblong or elongate obcor-
date. The largest petioles occur in a fruiting branch from Living Bridge
(the type specimen of the subspecies in Kew Herbarium) ; they are spatu¬
late oblong, 75 to 92 by 44 to 48 mm., tapering rapidly into a marginate
base 4 to 6 mm. long. A specimen from Moflong (in Kew Herbarium)
has oblanceolate-linear petioles 30 to 45 by 10 to 16 mm. The other
material is intermediate between these two extremes, and one twig from
Myrung Wood (in Kew Herbarium) has elongate-obcordate petioles 35
to 45 by 16 to 20 mm. in size. The blades of the leaves vary from ovate
to lanceolate and are 35 to 65 by 14 to 40 mm.; in some specimens the
laminae are decidedly smaller than the winged petiole, while in others
the reverse is true.

1 Citrus Ichangensis latipes, subsp. nov. — Citrus ichangensis afhnis, foliis acutis haud caudatis, floribus

in paniculatis pauci-floribus (3-5) dispositis, fructibus oblatis.

3 This is the type of the subspecies.

3 Cf. Hooker, J. D., 1854, Himalayan Journals, London, v. 2, p. 2S8, 292, 323.

4 This specimen has only a lithographed label with manuscript additions. One of the twigs has extremely
ong and slender winged petioles like the specimen from Moflung in Kew Herbarium and probably was a
part of the same collection. The other specimens of Hooker and Thomson in Kew Herbarium have this
same label carrying in manuscript the name “Citrus latipes H. f. and T.,” but have in addition original
labels giving the exact locality and date of collection.

Oct. io, 1913

Citrus Ichangensis

PREVIOUSLY PUBLISHED NOTICES OF THE SUBSPECIES

Very little has been published concerning this plant. The first notice
seems to have been given it in 1874 by Edmund Goeze, who lists it as
“ Citrus laiipes Hook. fil. et Th. A very peculiar species from India/’ 1

In 1875 J. D. Hooker, in his Flora of British India,1 2 cited it under
the name “ C . laiipes Hook. f. and Thoms. Herb. Ind. Or.” as a synonym
of C. histrix DC., an erroneous determination doubtless due to the lack
of flowers and mature fruits in the Khasi material at his disposal. The
name “Citrus laiipes Hook. f. and Thoms.” is a nomen nudum without
standing in taxonomy, since no description has been published under it.

Efforts are being made to secure ripe fruits and viable seeds of this
interesting tree, which, like the Chinese form of the species, promises to
be very cold resistant.

POSSIBLE USES OF CITRUS ICHANGENSIS

Mr. E. H. Wilson informs the writer that the form of this species culti¬
vated in the Ichang region yields an excellent fruit known to foreign
residents of the Yangtze Valley as the “ Ichang lemon.” These fruits are
shipped down the river to Hankow and west well into Szechwan, and
are so much esteemed as to command good prices.

The large size of the seeds makes it probable that Citrus ichangensis will
produce very vigorous seedlings, and hence it is likely to be of value as a
stock on which to graft other citrous fruits. These numerous large seeds,
which promise to render this plant so valuable as a stock, have the draw¬
back of greatly reducing the proportion of juice, because of the space
they take up. However, experience has shown that it is relatively a
simple matter to breed nearly seedless varieties of citrous fruits by
selection or hybridization.

So far as is now known, Citrus ichangensis is native farther north than
any other evergreen species of Citrus, only the deciduous Citrus trifoliata
having a more northerly range. Besides having the northernmost range
of any known evergreen species of Citrus it occurs at the highest altitudes
reported for any wild species of the genus. In the Hsingshan District,
in latitude 31 0 10', Mr. Wilson collected this plant at an altitude of 4,200
feet, and P&re Cavalerie found it in central Kweichow at a height of
5,577 feet.

At Moflong in the Khasi Hills, Hooker and Thomson found the Khasi
subspecies growing wild at an altitude of 6,000 feet. As to the winter
climate of this part of Assam J. D. Hooker says:

In November the vegetation above 4,000 feet turns wintry and brown, the weather
becomes chilly, and though the cold is never great, hoarfrost forms at Churra, and
water freezes at Moflong.3

1 Translation from Goeze, Edmund, 1874. Ein Beitrag zur Kenntniss der Orangengewackse. Ham¬
burg, p. 19*

2 Hooker, J. D., 1875. Flora of British India, v. 1, London, p. 515.

3 Hooker, J. D., 1854. Himalayan Journals, v. 2, London, p. 323.

Journal of Agricultural Research

Vol. I, No. i

Around Ichang, which is situated at an altitude of about 2,000 feet, the
winters may be severe, as is proved by the meteorological record for the
year 1888, which showed an absolute minimum of 220 F. ( — 5.6° C.) in
February.1 It is highly probable that a series of observations extending
over a number of years would show that the minimum temperature
occasionally falls decidedly lower than this. It would undoubtedly be
colder at an altitude of 4,200 feet in the near-by Hsingshan District,
where this species grows wild.

Mr. Wilson, who knows the climate of this part of China well, is con¬
fident that the “Ichang lemon’ ’ will prove to be one of the hardiest
citrous fruits. Add to this the fact that the fruit is of a quality good
enough to cause it to be exported to cities several hundred miles distant
and it is obvious that this strikingly distinct new species of Citrus promises
to be of value as a hardy substitute for the lemon, as well as a vigorous
and hardy stock for other citrous fruits, and is eminently deserving of the
attention of experimenters for use in the breeding of new types of hardy
citrous fruits now so much in demand in this and other countries.

Its discovery in a part of China as accessible as Ichang is a further proof
of the rich harvest of new species of plants that awaits the botanist and
agriculturist in China.

DESCRIPTION OF PLATE

Plats I. Citrus ichangensis Swingle: The type specimen from Hsingshan District,
Hupeh Province, China, E. H. Wilson, No. 2230, May 7, 1907; in the herbarium
of Arnold Arboretum; natural size.

1 Doberck, William, 1889. Meteorological observations made at Ichang, China, and at South Cape
Formosa, in 1S88. Quart. Jour. Roy. Met. Soc. [London], v. is, no. 72, p. 242.

CYSTICERCUS OVIS, THE CAUSE OF TAPEWORM CYSTS

IN MUTTON

By B. H. Ransom,

Chief , Zoological Division, Bureau of Animal Industry
INTRODUCTION

It has been known for nearly half a century that cysticerci occur in
mutton, but they have generally been looked upon as zoological curi¬
osities rather than parasites of real economic importance; in fact, it
seems that this opinion has been so commonly accepted as an established
truth that a systematic examination of sheep for such cysticerci, or
measles, like that given cattle and hogs, has been considered unnecessary
by meat-inspection authorities. So far as this country is concerned,
however, the belief that sheep measles are rare has been lately discovered
to be quite erroneous. Instead of being rare, sheep measles have been
found to be of much the same order of frequency as beef measles and far
more common than pork measles, which are almost unknown in the
United States. Where the presence of measles has been carefully looked
for, the percentage of affected sheep has run 2 per cent and over, and
during the calendar year 1912 approximately 20,000 sheep carcasses were
retained under Federal inspection at various abattoirs on account of
measles, most of them during the last few months of the year.

In the light of these figures it is quite evident that the mutton cysti-
cercus is far from being the unimportant parasite it is commonly assumed
to be, and it is furthermore quite certain that as inspectors become gen¬
erally more familiar with this parasite and with the proper methods of
inspecting for its presence the percentage and gross number of cases
found will materially increase.

As yet sufficient data are not at hand to indicate the extent of direct
injury to sheep by the measles parasite, so that the chief practical impor¬
tance of sheep measles recognized at the present time is in their relation to
meat inspection and public health. Like beef and pork cysticerci, the
mutton cysticercus is of special interest in meat inspection because it
affects the musculature, that part of the animal which is at once the most
valuable for food purposes and the most difficult to inspect thoroughly.

The beef and pork cysticerci are well known to be the intermediate
stages of two species of tapeworms occurring in man. The question
naturally arises, Is the mutton cysticercus likewise the intermediate stage
of a human tapeworm ? The leading foreign meat-inspection authorities
have held that the mutton cysticercus is simply Cysticercus cellulosae ,
the pork cysticercus, in an unusual host, and have laid down identical

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(rs)

7954°— 13 - 2

Vol. I, No. 1
Oct. 10, 1913
A — 1.

Journal of Agricultural Research

Vol. I, No. r

regulations governing the disposal of affected hog and sheep carcasses.
The American meat-inspection regulations, which are similar to, though
necessarily somewhat more stringent than, the German regulations
because of the lack of a Preibank system in this country, require the con¬
demnation of carcasses heavily infested with C. cellulosae and permit
slightly infested carcasses to be rendered into edible fat. As a con¬
demned carcass is entirely destroyed for food purposes and as the value
of a sheep carcass rendered into edible tallow is scarcely greater than
that of one which has been condemned and made into fertilizer or other
inedible products, a carcass infested with C. cellulosae in any degree
whatsoever would be practically excluded from use as food under Ameri¬
can regulations. Accordingly, if the mutton cysticercus were actually
C. cellulosae , the 20,000 sheep carcasses in which muscle cysticerci were
found last year would have been eliminated from the meat supply of the
United States. Relatively this loss would not have been very great, and
in actual money value it would not have exceeded $100,000. In thefuture,
however, much greater losses would occur, because the more efficient
methods of inspection which would be developed by experience would
naturally lead to the detection of more nearly all the cases of sheep
measles than the earlier, less efficient methods. The number of sheep
affected with measles is probably considerably in excess of 1 per cent of
the entire number slaughtered, and accordingly the loss on this account
would be very large if anywhere near all the cases were found on inspec¬
tion and if they were disposed of under the assumption that the parasite
involved is C. cellulosae .

Shortly following the discovery of the first cases found last year, the
writer undertook an investigation of the question of sheep measles, with
the result that it was quickly proved that the parasite involved is cer¬
tainly not Cysticercus cellulosae , though closely resembling it in some re¬
spects, and in due course of time it was definitely established that the
mutton cysticercus is the larval stage of a dog tapeworm.

The question of sheep measles is therefore much less serious than it
would be if the parasite were one transmissible to man, particularly if it
were the rather dangerous Cysticercus cellulosae . So far as meat inspec¬
tion is concerned, however, sheep measles, though less important as a
public-health question, are almost as important as though the parasite
involved were transmissible to human beings, because meat containing
parasites of sufficient size to be noticeable is more or less objectionable
as food for esthetic reasons if on no other account.

HISTORICAL SUMMARY

Considering critically the various statements which have appeared
relative to muscle cysticerci in sheep prior to the recent investigations
by the present writer, it may be noted in the first place that excepting
one of Morot’s (1899c)1 cases (No. 3), which was quite evidently one of

1 Bibliographic references in parentheses refer to the "Bibliography,'’ pp. 54-57.

Oct. 10, 1913

Cysticercus Ovis

generalized coenurosis, there is no definite conclusive evidence that more
than one species of parasite is concerned in sheep measles; hence the
presumption is that the muscle cysticerc i reported from sheep *all belong
to a single species. Taking into account the fact that it has now been
proved by experiment that muscle cysticerci in sheep develop into tape¬
worms distinct from either Taenia solium or T. hydatigena , it is quite
clear that none of the observers reporting muscle cysticerci in sheep has
given sufficient evidence to show that the parasites in any instance were
Cysticercus cellulosae , as they were held to be by some, or C. ienuicollis , as
they were held to be by others, and not in all cases, C. ovis. Commonly
the only evidence to support the observer’s identification is a statement
that the parasite showed the
characters of C. cellulosae (01 1,

Armbriister, Colberg, Rick-
mann, Herter) or C. ienuicollis
(Chatin, Glage). In a few
cases measurements of the
hooks have been recorded, but
these apply equally as well or
better to C. ovis than to C .
cellulosae or C. ienuicollis .

Rongert’s report is of special
interest in this connection, as
he gives a photomicrograph of
the hooks (fig. 1), comparison
of which with the hooks of
C. cellulosae shows that the
hooks agree imperfectly, thus
demonstrating the incorrectness of Bongert’s positive opinion that the
parasite was C. cellulosae . The opinion formerly held by the present
writer (i9o8d) that certain partially grown muscle cysticerci with hooks
not yet fully developed which had been found in a sheep were C. cellu¬
losae on account of the presence of certain characters also found in C.
cellulosae is likewise seen now to be quite erroneous.

Railliet and Morot noticed that the hooks of a cysticercus resembling
Cysticercus cellulosae from a sheep heart, though agreeing fairly well in
size with C. cellulosae hooks, as shown by the measurements which they
give, corresponded closely in form to those of C. ienuicollis . They
accordingly so identified the cysticercus, at the same time, however,
calling attention to the fact that the hooks are fewer in number than is
usual in C. ienuicollis and that they are smaller, differences possibly to
be attributed, according to their view, to the location of the parasite in
the muscles instead of in the serous membranes. It is quite probable —
in fact, not to be doubted — that the parasite in this case was C. ovis.

Fig . 1 . — Cysticercus ovis: Hooks, X 2 75 . ( After a photomi¬
crograph by Bongert, 1899a, fig. 3.)

1 8 Journal of Agricultural Research voi.i,No.i

Not only have observers failed to give sufficient evidence that the
mutton cysticerd in any case exactly agreed in morphology with Cysti-
cercus cellulosae or C. ienuicollis , but they have also failed to produce
experimental proof to support their identifications. C. cellulosae has
never been produced experimentally in sheep by feeding Taenia solium
eggs (Teuckart, Kiichenmeister, Perrondto); nor, vice versa, has T. solium
been produced in man as a result of ingesting mutton cysticerci (Chatin,
Ransom1).

There is also no good evidence that Taenia hydatigena has ever been
obtained as a result of feeding the mutton cysticercus to dogs. It is
true that Chatin states that such is the case, but the evidence that the

tapeworms were identical with those belong¬
ing to Cysticercus tenuicollis consists simply
in ChatnTs affirmation that they were the
same, and there is no objective evidence at
all to support this view. It also should be
noted that no one has shown that segments
of T. hydatigena , when fed to sheep, will pro¬
duce muscle cysticerci. Leuckart, Kiich-
enmeister, and others have found only C.
tenuicollis as a result of such experiments.

Cobbold’s opinion that Cysticercus ovis
is the larva of a human tapeworm, the so-
called Taenia tenella , has never had any
supporting evidence and, of course, is now
entirely discredited. Cobbold, however, it
is interesting to note, was quite correct in
another opinion which he at one time held — namely, that it is probable
that the adult of C. ovis occurs in one of the carnivora.

Most of the records of muscle cysticerci in sheep are based upon iso¬
lated cases in which the parasites have usually been more or less degen¬
erate. Thus, Cobbold noted the presence of degenerated cysticerci in
mutton on several occasions and described Cysticercus ovis on the basis
of a single specimen (fig. 2) which had lost the caudal bladder before it
came into his hands. Maddox described C. ovipariens (figs. 3 and 4) on
the basis of one degenerated cysticercus. The number of cases seen by
Mobius, reported by Kiichenmeister, is not stated. Chatin apparently
saw muscle cysticerci on several occasions, and some of these evidently
were alive and undegenerated. Morot refers specifically to five cases
and refers to an indefinite number of others, in all of which the parasites
were degenerated and were recognized as cysticerci only from the char¬
acter of the cysts. Railliet and Morot reported one case of a single,
apparently undegenerated cysticercus in the heart of a sheep, and refer

Fig. 2. — Cysticercus ovis: Head and
neck, X 30. (After Cobbold, 1869a, p.
30, fig. 2.)

1 For an account of the present writer’s experiments, see pp. 21-26.

Oct. io, 1913

Cysticercus Ovis

Fig. 3. — Cysticercus ovipariens (= C. ovis): Fragment of head, X 85.
(After Maddox, 1873a, pi. 19, fig. 1.)

to a similar case of cysticercus in the heart of a kid. The case reported
by Olt and Bongert showed numerous cysticerci, some of which appar¬
ently were alive. In another case seen by Olt the parasites were all
degenerate. Armbriis ter found calcified cysticerci in 2 or 3 sheep out of a
shipment of 16 head. One case of muscle cysticerci was found by Colberg
in which numerous
degenerated para¬
sites were present.

In a case of cysti¬
cerci in a sheep heart
reported by Railliet
the parasites were
very young, without
hooks. Glage is the
only author thus far
who has given a
detailed statistical
record of the fre¬
quency of muscle
cysticerci in sheep.

His records, however,
are based entirely upon the presence of degenerated cysticerci, and it is
not improbable that he overlooked many cases of live cysticerci. He
found 32 cases (1.45 per cent) among 2,198 carcasses in which the head
muscles and hearts were examined and 16 cases (0.8
per cent) among 1,984 carcasses in which only the
hearts were examined. Rickmann fails to state the
number of cases observed. The cysticerci in the one
case reported by the present writer in 1 908 were un¬
degenerate but only partly grown. Herter mentions
one case and says that only nine cases of sheep
measles were recorded in the meat-inspection reports
of Prussia for the year 1909. Making a very liberal
allowance for the number of indefinitely reported
cases, the total number of individual cases of sheep
measles reported in the literature prior to the recent
investigations in this country is considerably less than
100, and in only a very few of these were the cysticerci at all numerous
or present in a living, fully developed, undegenerated condition. It is
accordingly not surprising that the identity of these parasites should
have remained so long undiscovered, particularly in view of the fact
that they have received but little attention from experienced parasitolo¬
gists, who, moreover, have had very unsatisfactory material for study.

Fig. 4. — Cysticercus ovi-
pariens (= C . ovis):
Hooks, X 160. (After
Maddox, 1873a. pi- 18,
fig- 5-)

Journal of Agricultural Research

Vol. I, No. i

Cobbold, for example, apparently studied only one specimen (imperfect),
and Railliet seems to have had only one fully developed undegenerated
specimen for critical examination.

Up to the present time sheep measles have been reported from the
following countries: England, Germany, France, Algeria, German South¬
west Africa, New Zealand, and the United States.

In completing this brief critical summary of the literature, only a few
words need be given concerning the morphology of the parasites. As
already noted, morphological details have been omitted from most of
the accounts given of the recorded cases. The measurements of the
hooks given by Railliet and Morot correspond to Cysticercus ovis , as do
Bongert’s measurements and photomicrograph. Maddox was the first to
observe the mammillated surface of the caudal bladder, which, however,
has not been recognized as a distinctive difference between C. ovis and
C. ienuicollis , except by the present writer (i9o8d), and apparently has
escaped attention from other observers.

LIFE-HISTORY INVESTIGATIONS

Under date of February 29, 1912, Dr. S. E. Bennett, inspector in
charge at Chicago, Ill., reported to the Bureau of Animal Industry that
a number of sheep carcasses had been found to be infested with measles,
and under date of March 1 Dr. O. B. Hess, inspector in charge at Seattle,
Wash., also reported the finding of measles in several sheep carcasses.
Specimens were forwarded to Washington from both stations for labora¬
tory examination. The cysts in the specimens were all degenerate, but
fragments of the caudal bladder of cysticerci were found, and in view of
the presence of cuticular papillae, which are likewise present on the
caudal bladder of Cysticercus cellulosae , and in accordance with the
opinion of German meat-inspection authorities as to the identity of
mutton cysticerci, the diagnosis of C. cellulosae was made. Shortly fol¬
lowing the first reports, information was received that out of 4,537 sheep
slaughtered at Seattle, Wash., 79 carcasses were retained on account of
measles, and that during a month at Chicago 224 carcasses were retained.

With this information at hand it was immediately apparent that the
diagnosis of Cysticercus cellulosae could not be correct, for the reason
that C. cellulosae and its tapeworm stage, Taenia solium , are exceedingly
rare in the United States. Probably not more than a dozen cases of
pork measles are found annually at any of the large stations, where the
number of hogs slaughtered amounts to hundreds of thousands. It was
unbelievable that a parasite so rare in its usual host should be so common
in sheep. A few days spent in studying numerous specimens obtained
at the abattoirs in Chicago developed the fact that the sheep-measle
parasite was certainly not C. cellulosae , though in certain characters
they were very similar. In some details of structure the muscle cysticerci

Oct. io, 1913

Cysiicercus Ovis

resembled C. tenuicollis , but in other respects the two forms did not
agree. Accordingly an experiment was undertaken to determine whether
the parasites would develop in dogs and whether the tapeworms, if any
developed, would prove to be T. hydatigena (the tapeworm corresponding
to C. tenuicollis; also known as T. marginata , the marginate tapeworm
of the dog), as affirmed by Chatin (1886a), who stated that he had
obtained T. marginata by feeding mutton cysticerci to dogs, or whether
they would prove to be some other species. Seven dogs were under
observation in 1912. Five of these were fed cysticerci from sheep
muscle, while two, as controls, were fed C. tenuicollis from the omentum
or mesenteries of sheep. With three exceptions, as noted below in the
records of the experiment, the dogs were given a dose of castor oil and
the feces examined for the presence of parasite eggs before the cysticerci
were fed. During the experiment the dogs were nourished on dog bis¬
cuits, corn-meal mush, and some cooked meat but no mutton and were
confined in separate kennels.

Dog No. 1. — A grayish brown young female. Fed muscle cysticerci from sheep.
Feces were not examined before feeding cysts.

March 25. Fed 1 cyst from heart muscle of sheep.

March 27. Fed 1 cyst from heart muscle of sheep— probably dead.

March 28. Fed 3 cysts from heart muscle of sheep.

March 29. Fed 3 cysts from heart muscle of sheep.

March 30. Fed 3 cysts from heart muscle of sheep — 1 probably dead.

April 1. Fed 1 cyst from diaphragm of sheep — probably dead.

April 2. Fed 1 cyst from body muscle of sheep.

April 3. Fed 2 cysts from heart muscle of sheep.

April 24. Fed 1 cyst from heart muscle of sheep.

April 29. Fed 2 cysts from heart muscle of sheep.

May 2. Fed 1 cyst from heart muscle of sheep.

May 21. Fed 1 cyst from heart muscle of sheep.

May 22. Fed 1 cyst from heart muscle of sheep.

May 24. Fed 2 cysts from heart muscle of sheep.

Total. . 23 cysts.

June 22. Eggs of Toxascaris and a tapeworm segment found.

June 27. Tapeworm segments found in feces.

July 24. Chloroformed. About 25 individuals of Toxascaris in upper half of
jejunum. Seven tapeworms, all with gravid segments, in ileum. Heads attached
near upper end of ileum, about 65 cm. from ileocecal valve. Length of tapeworms,
45 to 55 cm.

Dog No. 2. — A white-and-tan young female. Fed Cysiicercus tenuicollis from peri¬
toneum of sheep. Feces were not examined before feeding cysts.

May 10. Fed 1 cyst.
May 28. Fed 5 cysts.

Total.. 21 cysts.

April 5. Fed 3 cysts.

April 9. Fed 4 cysts.

April 11. Fed 1 cyst.

April 18. Fed 7 cysts.

June 22. Eggs of tapeworm and Toxascaris eggs found.

July 11. Two tapeworm segments found in feces.

July 26. Chloroformed. Numerous individuals of Toxascaris in jejunum and duo¬
denum. Nine tapeworms with gravid segments; one of the tapeworms was about
no cm. long. The tapeworms were attached about 8 cm. below the pylorus, 80 cm.

Journal of Agricultural Research

Vol. I. No. i

from the ileocecal valve, and the posterior ends of the worms extended to within 40
cm. of the ileocecal valve.

Dog No. 3. — A young black-and-white female. Fed muscle cysticerci from sheep.
March 29. Received one-half ounce of castor oil at 5 p. m. March 30. Feces were
examined with negative results. ,

April 5. Fed 1 cyst from myocardium.

April 6. Fed 3 cysts from myocardium.

April 10. Fed 3 cysts from myocardium.

April 11. Fed 6 cysts from myocardium.

April 13. Fed 1 cyst from myocardium.

April 29. Fed 1 cyst from myocardium.

May 2. Fed 1 cyst from myocardium.

May 21. Fed 1 cyst from myocardium.

May 23. Fed 1 cyst from myocardium.

May 24. Fed 3 cysts from myocardium.

Total... 21 cysts.

June 11. Feces examined but no eggs found.

No segments or eggs have been found (prior to July 22).

July 22. Chloroformed. Four tapeworms attached 25 to 35 cm. from the ileocecal
valve, one of them with gravid segments, about 45 cm. long when extended, other
three not over 2 to 5 cm. long. Three very short tapeworms in cecum. In large
intestine a string of about 10 gravid segments. Total number of tapeworms, seven.
Three individuals of Toxascaris in jejunum.

Dog No. 4. — A young red male. Fed muscle cysticerci from sheep. March 29.
Received one-half ounce of castor oil at 5 p. m. March 30. Feces were examined
and Toxascaris eggs found.

April 18. Fed 1 cyst from myocardium of sheep.

April 19. Fed 1 cyst from cheek muscle of sheep.

April 23. Fed 16 cysts — 3 from myocardium and 13 from muscles of sheep.

Hooks were well developed.

April 24. Fed 1 cyst from myocardium of sheep.

May 2. Fed 1 cyst from myocardium of sheep.

May 15. Fed 1 cyst from myocardium of sheep.

May 21. Fed 1 cyst from myocardium of sheep.

May 24. Fed 2 cysts from muscles of sheep.

Total .... 24 cysts.

June 11. Eggs of Toxascaris , but no tapeworm eggs found.

June 27. Three broken tapeworm segments found in feces.

July 24. Chloroformed. Two individuals of Toxascaris in upper part of jejunum.
Sixteen or seventeen tapeworms extending down into lower part of colon. None
attached more than 4 cm. above ileocecal valve. One attached in cecum. None
with gravid segments. Length, 20 to 50 cm.

Dog No. 5. — A medium-sized brindled male. Fed muscle cysticerci from sheep.
March 29. Received one-half ounce of castor oil at 5 p. m. March 30. Feces were
examined with negative results.

April 23. Fed 20 cysts from muscles of sheep. Hooks were well developed.

April 24. Fed 1 cyst from myocardium.

May 2. Fed 1 cyst from myocardium.

May 15. Fed 1 cyst from myocardium.

May 21. Fed 1 cyst from myocardium.

May 24. Fed 2 cysts from muscles of sheep.

Total. . 26 cysts.

Oct. 10, 1913

Cysticercus Ovis

June 11. Tapeworm eggs and eggs of Toxascaris were found in feces.

June 19. Two or three segments found in feces.

July 24. Chloroformed. No tapeworms found. Numerous d*ead fly larvae in colon
and small intestine. Numerous Toxascaris in upper part of jejunum and in duo¬
denum.

Dog No. 6.— A medium-sized white male. Fed muscle cysticerci from sheep.
March 29. Received one-half ounce of castor oil at 5 p. m. March 30. Feces were
examined and Toxascaris eggs found.

April 23. Fed 20 cysts from muscles of sheep. Hooks well developed.

April 24. Fed 1 cyst from myocardium.

May 2. Fed 1 cyst from myocardium.

May 15. Fed 1 cyst from myocardium.

May 21. Fed 1 cyst from myocardium.

May 24. Fed 2 cysts from muscles of sheep.

Total.. 26 cysts.

June 11. Tapeworm eggs and eggs of Toxascaris were found in feces.

June 19. Two tapeworm segments were found in feces.

July 26. Chloroformed. Eight or nine tapeworms with gravid segments, one of them
measuring 1 meter in length. Heads attached 135 cm. above the ileocecal valve, and
posterior ends of the worms extending to a distance of 55 cm. from the ileocecal valve.
Numerous individuals of Toxascaris in jejunum and in duodenum.

Dog No. 7. — A medium-sized black-and-white spotted female. Fed Cysticercus
tenuicollis from peritoneum of sheep. Feces were not examined before feeding cysts.

April 9. Fed 4 Cysticercus tenuicollis

April 18. Fed 7 Cysticercus tenuicollis .

May 28. Fed 7 Cysticercus tenuicollis.

Total.. 18.

June 22. Feces show a few young tapeworm segments.

July 11. Found several portions of tapeworms; each portion contained from 2 to 20
segments.

July 26, Chloroformed. Three or four individuals of Toxascaris in duodenum and
jejunum. Ten tapeworms with short strobila not over 10 mm. long in duodenum.

In continuation of the experiment with the dogs another experiment
was undertaken for the purpose of recovering the cystic stages of the
tapeworms. Ten lambs were purchased from a lot of thirty-nine, the
remainder of which were slaughtered at one of the packing houses in
Chicago and found to be free on post-mortem examination from both
muscle cysticerci and Cysticercus tenuicollis. One of the ten died shortly
after purchase and consequently was not used in the experiment. The
# sheep were kept during the experiment in floored and covered pens in
one of the sheep barns at the Union Stock Yards, Chicago, and were fed
dry hay and occasionally oats and received water piped from the water
mains. The identity of the various lambs was maintained by numbered
ear tags.

Lamb No. 1. — July 24. One half of a gravid segment from a tapeworm out of dog No.

1 (a dog which had been fed muscle cysts) was cut in pieces and given in a drench with
water.

August 7. Dr. Day reported that lambs Nos. 1,2,3, and 5 were more or less sick but
would probably recover.

Journal of Agricultural Research

Vol. I, No. i

August 21. Is very thin and has a diarrhea, but is feeding well.

October 15 (eighty-three days after feeding). Chloroformed. In poor flesh, very
little fat. Cysticerci were found in the panniculus camosus. Forty-two degenerate
cysts were counted in the diaphragm ; ten degenerate cysts in the wall of the esophagus.
Several cysts in anterior lobes of lungs, 2 to 3 mm. in diameter; contents caseous. One
contained a small dead cysticercus, 1 mm. in diameter; rudiment of head present.
Numerous small degenerate cysts in heart. Numerous cysticerci in muscles of masti¬
cation; some living, others degenerate. A few nodules in the wall of the rumen, and
one in the wall of the fourth stomach, 2 to 4 mm. in diameter, hard, shotlike, with
thick wall and cheesy contents. No cysticerci were found in these cysts. Nodules
present on wall of cecum, probably Oesophagostomum . No cysticerci found in these
nodules. Many degenerate cysts among those present in the musculature in various
parts of the body. The sizes of 13 live cysts measured in situ were as follows, in milli¬
meters: 9 by 3.5, 8 by 3, 7 by 4, 7 by 3, 6 by 3, 5 by 4, 4 by 2.5, 5 by 3, 7 by 4, 8 by 4,
8 by 3, 6 by 2.5, and 9 by 4. A cyst 5 or 6 mm. in diameter with thick leathery cap¬
sule contained a live cysticercus which was active under the microscope. This
cysticercus was not fully developed , only the blade of the hooks being formed . Other
cysticerci showed fully developed hooks, and cysticerci from degenerate cysts showed
in some cases hooks not yet fully formed.

Lamb No. 2. — July 26. A gravid segment from a tapeworm out of dog No. 6 (a dog
which had been fed muscle cysts) was given in a drench with water.

August 7. Dr. Day reported that lambs Nos. 1,2,3, and 5 were more or less sick but
would probably recover.

August 1 7 (twenty-two days after feeding) . This animal died , but its death was not
reported until two days later, when decomposition was so far advanced that Dr. Day
did not attempt a post-mortem examination.1

Lamb No. 3. — July 26. A gravid segment from a tapeworm out of dog No. 6 (a dog
which had been fed muscle cysts) was given in a drench with water.

August 7. Dr. Day reported that lambs Nos. 1,2,3, and 5 were more or less sick but
would probably recover.

August 18 (twenty-three days after feeding). This animal died. Decomposition was
far advanced the following day when a post-mortem examination was made, but some
of the masseter muscle and some of the muscle of a hind leg were obtained. Dr. Day
reports that cysts in the masseter muscle were quite well formed and contained a tiny
white spot just visible to the eye. Microscopic examination by Dr. Day showed that
the head was not well formed, but papillae were evident on the caudal bladder.

Lamb No. 4. — July 24. A gravid segment (cut in pieces) from dog No. 1 (a dog which
had been fed muscle cysticerci) was given in a drench with water,

August 7. In very bad condition; probably will die.

August 11 (eighteen days after feeding). Dead.

August 12. An incomplete post-mortem examination was made by Dr. Day. Ad¬
vanced decomposition. A number of cysts were obtained from the masseter muscles.

Lamb No. 5. — July 24 a gravid segment from a tapeworm out of dog No. 3 (a dog
which had been fed muscle cysticerci) and on July 26 two gravid segments from a tape-,
worm out of dog No. 6 (a dog which had been fed muscle cysticerci) were given in a
drench with water, a total of three segments.

August 7. Dr. Day reported that lambs Nos. 1,2,3, and 5 were more or less sick but
would probably recover.

1 In prior publications (Ransom, 1913, p. 78; 1913, p. 31) it was erroneously stated that all of the lambs
which had been fed eggs of the muscle cyst tapeworm showed tapeworm cysts in the muscles. The con¬
dition in lamb No. 2, of course, was not determined, as no autopsy was made on this animal. The state¬
ment (Ransom, 1913, p. 31) that the lambs died in 14 to 22 days after feeding is also inaccurate. It should
be 13 to 23 days.

Oct. io, 1913

Cysticercus Ovis

August 12 (ten days after feeding) . Dead. Post-mortem examination by Dr. Day the
following morning showed a large number of cystic parasites in the masseter muscles,
heart, tongue, and diaphragm. There were also numerous cystic parasites in the
skeletal muscles and a few hemorrhagic spots.

Lamb No. 6. — July 24, four segments from tapeworms out of dogs Nos. 1 and 3 (dogs
which had been fed muscle cysticerci), two segments from each dog, and July 26 six
segments from tapeworms out of dog No. 6 (a dog which had been fed muscle cysticerci)
were given in a drench with water, a total of ten segments.

August 5. Appears ill and out of condition.

August 6 (thirteen days after feeding). Dead. Post-mortem by Dr. Day showed
that the parasites had already migrated to the muscles, and were found as very minute
cysts, more numerous in the heart and masseter muscles than elsewhere. There vrere
about 25 c. c. of fluid in the pericardium. The heart was very thickly studded with
minute cysts. There were about 350c. c. of fluid in the peritoneal cavity. A careful
examination of the fluid was made, but no parasites were found. The liver appeared
normal.

Lamb No. 7. — July 26. A gravid segment from a tapeworm out of dog No. 2 (a dog
which had been fed Cysticercus tenuicollis from the peritoneum of sheep) was given
in a drench with water.

August 21. Reported by Dr. Day as doing well.

October 18 (eighty-four days after feeding). Chloroformed. Animal in poor flesh.
Twelve to fifteen cysts on omentum and mesenteries, two of which are alive, the others
degenerate. One degenerate cyst under peritoneum in pelvic cavity. Degenerate
cysts vary in size up to a maximum of 20 mm. in diameter. Contain dead cysticerci,
a small amount of colorless serous fluid and flocculent d£bris or a greenish, caseous
material. The live cysticerci measure 8 by 15 mm., and show the usual macroscopic
characters of Cysticercus tenuicollis. A few degenerate cysticerci of small size on the
surface of the liver. No cysticerci in the muscles, lungs, or other organs, except as
noted above. Oesophagostomum nodules on the intestine.

Lamb No. 8. — July 26. Ten gravid segments from a tapeworm out of dog No. 2
(a dog which had been fed Cysticercus tenuicollis from the peritoneum of sheep) given
in a drench with water.

August 21. Reported by Dr. Day as doing well.

October 17 (eighty -three days after feeding). Chloroformed. Animal in poor flesh.
A considerable number of small degenerate cysticerci on surface and in depths of
liver. About 25 degenerate cysts on omentum and mesenteries. One live cysti¬
cercus on omentum about 12 mm. in diameter shows the usual macroscopic char¬
acters of Cysticercus tenuicollis. One degenerate cyst on tendinous portion of diaphragm
(abdominal surface). Small nodules in lungs, one of which contained a young dead
cysticercus showing under the microscope transverse ridges on the cuticle of the
caudal bladder. Synthetocaulus nodules also present on the lungs. Several pockets
in the lungs with fibrous walls containing greenish pus. The contents of these pockets
were examined, but no cysticerci were found. Heart and muscles were free from
parasites. A cyst from the omentum, 8 mm. in diameter, with thick fibrous wall con¬
tains a dead cysticercus with evaginated head and bladder about 3 mm. in diameter.
Two cysts from the omentum or mesentery, 5 and 6 mm. in diameter, respectively,
contain each a dead cysticercus and a small amount of colorless serous fluid and
flocculent debris. The other degenerate cysts are similar, except the contents in some
are greenish, caseous. Their size varies from 2.5 to 10 mm,, and all have thickened
walls to % mm. thick. The degenerate cyst from the tendinous portion of the dia¬
phragm is flattened, 8 mm. in diameter. Its wall is thin, and it contains a dead
Cysticercus tenuicollis and a small amount of serous fluid and white flocculent matter.

Journal of Agricultural Research

Vol. I, No. i

Lamb No. 9. — A check animal, not fed with tapeworm segments.

October 18. Chloroformed. In poor flesh. Free from parasites except Oesopha-
gosiomum nodules on the intestines.

The following experiments relating to the possibility of the develop¬
ment of sheep -measle tapeworms in man have been caried out, the writer
being the subject.

On March 6, 1913, a cysticercus about 5 mm. in diameter, and March 14 another
cysticercus of similar size, both from sheep hearts, were swallowed. Both cysticerci
were alive and in good condition, exhibiting lively contractions of the caudal bladder
when viewed under the microscope. On March 28 eight fully developed cysticerci
were isolated from a sheep carcass heavily infested with Cysticercus ovis and swallowed.
These cysticerci were apparently in good condition and were undoubtedly alive,
as they showed active movements under the microscope. No signs of tapeworm
infestation have appeared in the case of the writer.

SUMMARY OF LIFE-HISTORY EXPERIMENTS

Five dogs were each fed from 21 to 26 muscle cysticerci from sheep on
various dates between March 25 and May 24. Subsequent to June 11,
tapeworm eggs or segments were demonstrated in the feces, or tape¬
worms were found post-mortem in the case of all five dogs. No tape¬
worms were found in one of the dogs (No. 5) post-mortem, but a month
earlier this dog had shown tapeworm eggs and segments in the feces.
In the case of two of the dogs (Nos. 5 and 6) it was evident that the tape¬
worms had reached egg-producing maturity within seven weeks, as the
earliest feeding of cysticerci was on April 23, eggs being demonstrated
in the feces on June 11. The number of tapeworms recovered varied
from 7 to 16.

Two dogs were fed Cysticercus tenuicollis , 18 and 21 cysticerci, respec¬
tively, between April 5 and May 28. The first tapeworm eggs were found
in the feces on June 22. On post-mortem examination 9 tapeworms
were found in one dog and 10 in the other.

Six lambs (Nos. 1 to 6) were fed with gravid segments of tapeworms
from the dogs which had been fed Cysticercus ovis , two (Nos. 7 and 8)
with gravid segments of tapeworms from one of the dogs which had been
fed C. tenuicollis , and one (No. 9) was retained under the same conditions
as the others but without receiving any cysticerci. Lambs Nos. 1 to 6
received yi to 10 segments, and lambs Nos. 7 and 8, 1 and 10 segments,
respectively. Of the former all but the one receiving half a segment died
in 13 to 23 days after feeding, the one receiving 10 segments being the
first to die, followed by one receiving 1 segment (death in 18 days), then
by one receiving 3 segments (death in 19 days), then by two more receiving
1 segment each (death in 22 and 23 days, respectively), leaving the lamb
which received half a segment to survive until killed— 83 days after feed¬
ing. Both of the lambs fed with segments of Taenia hydatigena (adults

Oct. io, 1913

Cysticercus Ovis

of C. tenuicollis) survived until killed at the close of the experiment.
All but one of the lambs (No. 2), which died 22 days after feeding, were
examined post-mortem.

Omitting this lamb from consideration, all of the lambs which received
segments from the tapeworms produced by feeding muscle cysticerci
showed cysticerci in their muscles when examined. Those found in the
lamb which died 13 days after feeding were very small; those in the
lamb which died 23 days after feeding were somewhat farther along
in development, the beginnings of the head being already evident.
Eighty-three days after feeding, the muscle cysticerci were found to
have reached full development ; some which had fully developed were
already more or less degenerated, and some were found which had begun
to degenerate before they reached their full development. In addition
there were present live cysticerci which had not yet fully developed.

The lambs which had been fed segments of Taenia hydatigena showed a
few Cysticercus tenuicollis , most of which were degenerate. In both
animals there were small degenerate cysticerci on the liver. There were
no visible lesions of the liver in the lambs fed segments of T. ovis. No
C. tenuicollis was found in any of the lambs fed segments of T. ovis , and
no C. ovis in the lambs fed segments of T. hydatigena. The check lamb
showed neither C. tenuicollis nor C. ovis, and neither of these parasites
was found at the post-mortem inspection of the remainder of the lot
from which the experiment sheep had been selected.

Since these experiments show that muscle cysticerci in sheep resembling
Cysticercus cellulosae and corresponding to the form described by Cob-
bold as C. ovis develop into tapeworms when swallowed by dogs, it has
been definitely proved that these cysticerci are not C. cellulosae. The
adult of C. cellulosae (Taenia solium) does not occur in dogs; moreover,
the tapeworms which were produced in the dogs are quite different from
T. solium. Furthermore, the experiments prove that the muscle cysti¬
cercus and its adult stage are specifically distinct from C. tenuicollis and
T . hydatigena . It appears that the ingestion of one or more gravid
segments of T. ovis is likely to prove fatal to sheep.

Attempts to produce tapeworms in man by feeding mutton cysticerci
failed. On three occasions live mutton cysticerci were swallowed by the
writer, a total of 10 cysticerci being ingested. No evidence of tapeworm
infestation has since appeared. This experiment tends to prove that
Cysticercus ovis is not transmissible to man.

SYNOPSIS OP hWH HISTORY

The adult of Cysticercus ovis is a tapeworm ( Taenia ovis) which occurs
in the intestine of dogs. Since the parasites which live on dogs as a rule
also thrive on wolves, and, since coyotes and other wolves frequently

Journal of Agricultural Research

Vol. I, No. i

devour sheep, it is quite likely that T. ovis also occurs in coyotes and other
wolves as well as in dogs. In view of the fact, however, that dogs come
in much closer relations with sheep it seems quite evident that dogs are
chiefly responsible for the transmission of the parasite to sheep. It is
possible though rather unlikely that the tapeworm occurs in other carni¬
vores than dogs and wolves. There is little likelihood that the parasite
is transmissible to man, and for all practical purposes its nontransmis-
sibility to man may be considered an established fact. No such tape¬
worm has been reported from man, and, moreover, there are no authentic
cases of the occurrence in man of any dog tapeworm belonging to the
genus Taenia. Furthermore, Chatin has noted that the swallowing of
muscle cysticerci from sheep failed to result in infestation in his case.
The present writer, as already noted, has likewise on three occasions
swallowed live and active muscle cysticerci from sheep without resulting
infestation (p. 26).

Following the ingestion of the eggs of the tapeworm by sheep, the
parasites reach the muscles in less than 13 days; they either do not pass
through the liver or, unlike Cysticercus tenuicollis , leave no trace of their
passage through this organ. In less than three months (83 days) the
cysticerci reach their full development. As early as seven weeks after
the ingestion of the cysticercus by a dog, its development to the mature
egg-producing tapeworm may be complete. The development therefore
appears to be somewhat more rapid than in the case of Taenia hydatigena ,
which was found by Teuckart (1856a) to require from 10 to 12 weeks.
No doubt, however, the period required for development is subject to
great variation, and though seven weeks is perhaps near the minimum
for T. ovis , the period very likely may be greatly prolonged, as has been
noted by Hall (1911, p. 510) in the case of the gid tapeworm.

ZOOLOGICAL DESCRIPTION OF THE SHEEP-MEASLE PARASITE
Taenia ovis (Cobbold, 1S69) Ransom, n. comb., 1913.

1869: Cysticercus ovis Cobbold, 1869a, p. 30, fig. 2 (in Ovis aries; England).

1873: Cysticercus ovipariens Maddox, 1873a, p. 245-253, pi. 18, figs. 1-15, 17-18, pi. 19, fig. 1 (in Ovis
aries ; England).

1878: Cysticercus cellulosae of Kiichenmeister, 1878, in Kiichenmeister and Ziim, i878-i88ia, p. 104
(apparent misdetermination of C. ovis; in Ovis aries; Germany).

1885: Cysticercus tenuicollis of Chatin in Railliet, 1885a, p. 234 (apparent misdetermination of C. ovis;
in Ovis aries; France).

1886: Cysticercus oviparus Leuckart i886d, p. 498 (for C. ovipariens).

1913: Taenia ovis (Cobbold) Ransom, 1913.

Specific Diagnosis of Taenia.

Larval stage . — An oval cysticercus (PI. II, fig. 1) 3.5 by 2 mm. to 9 by 4 mm. in
diameter. Head and neck invaginated from the wall of the caudal bladder not at
one end but about midway between the ends. Membrane of bladder very thin;
with small mammillate projections; not corrugated transversely (fig. 5 and fig. 6, a).
Neck transversely corrugated, coiled spirally when invaginated, 1 to 5 mm. long when
evaginated. Head 500 to 800 /* in width; suckers oval, 240 to 320/1 in diameter; rostel-

Oct. io, 1913

Cysticercus Ovis

lum prominent, 275 to 375/x in diameter. Hooks (fig. 6) 24 to 36 in number, commonly
28 to 32, arranged in a double crown of alternating large and small hooks. Hooks
rather slender (more slender and more lightly built than those of Cysticercus cellulosae) ;
dorsal root of large hooks longer than the blade ; in both
large and small hooks a more or less well-marked outward
curving of the dorsal border of the hook in the transitional
region between the blade and dorsal root; ventral root of
small hooks transversely enlarged, not bifid but sometimes
presenting a faint median groove. Large hooks 156 to 188/t
long, average 173/4; blade (from point of blade to tip of
ventral root measuring in a straight line) 68 to 84/4, average
78/i (based on measurements of 24 hooks, fully developed
or nearly so, from 10 cysticerci taken from various sheep
and 13 hooks from the heads of 4 adult worms). Small
hooks 96 to 128) <1 long, average 113/4; blade (from point of
blade to tip of ventral root measuring in a straight line) 48
to 60 jjl, average 57/4 (based on measurements of 26 hooks, fully developed or nearly so,
from 11 cysticerci taken from various sheep and 10 hooks from the heads of 4 adult
worms).1

Calcareous corpuscles numerous in the neck, less numerous in the head, and very
rare in the caudal bladder.

Adult stage (PI. II,fig.3; text figs. 7, 8, 9, and 10).— Length of living worms with gravid
segments, 45 to no cm. Length (preserved material), 14 to 53 cm. ; maximum width,
4 to 8.5 mm.; terminal segments, 2.5 to 15 mm. long by 4 to 6 mm. broad, usually
longer than broad (measurements of 17 specimens with gravid segments). Strobila
tends to twist in the form of a spiral. Head 0.8 to 1.25 mm. in breadth; neck, 0.65
to 0.9 mm. wide (measurements of 26 preserved specimens). Rostellum 375 to 430/4
in diameter (8 specimens). Suckers 2 70 to 320/4 in diameter (4 specimens). Number,
arrangement, shape, and size of hooks as in larva. Segments with convex lateral
borders, in consequence of which the edge of the strobila commonly presents a scal¬
loped outline whose regularity is broken by the protuberant genital papillae. The
genital papillae are irregularly alternate and are situated posterior of the middle of the
segment; in gravid segments they may attain a diameter of over 1 mm. and a height
of three-fourths of a mm. Genital sinus large, varying in depth and width up to a
maximum of about 400/1. Cirrus pouch 450 to 550/4 long; inner end near the outer
side of the ventral longitudinal excretory vessel. The testicles are distributed in an
area which extends anteriorly to the anterior limits of the segment and laterally to the
longitudinal excretory vessels. This area is bounded posteriorly by a curved line
which in sexually mature segments intersects the median line at a distance from the
anterior border of the segment varying from a little more than a third to a little less
than half the length of the segment and intersects the longitudinal excretory vessels
a short distance in front of the posterior border of the segment, 4hus leaving an approxi¬
mately semicircular space entirely free from testicles, most of which is occupied by
the ovary. Behind the latter is the so-called yolk gland. The ovary is bilobed, the

1 M easurements of 26 hooks.

Member.

Larva.

Adult.

Member.

Larva.

Adult.

Large hooks:

Entire .

Average .

Blade .

it. -

156 to 188

173

76 to 80
79

It.

160 to 184
i73
68 to 84
75 j

Small hooks:

Entire .

Average .

Blade .

H-

96 to 120
112

52 to 60
57

/*.

104 to 128
116
48 to 60
57

Average .

Fig. 5. — Cysticercus ovipariens
(=C. ovis): Papillae on caudal
bladder, X 160. (After Mad¬
dox, 1873a, pi. 18, fig. 1 5.)

Journal of Agricultural Research

Vol.I.No.i

7ae/?/a da/an/cepsfadu/f) 7ate/?/a 6ra66e/fa/va)

Fig. 6.— Hooks of Taenia ovis , T. hydatigena, T. solium, T. balanicePs, and T. krabbei. Large and small
hooks designated by the same letters are from the same heads. The hooks shown in v and ^ are from the
type material of T. krabbei (B. A. I. No. 19352). Enlarged . (Original.)

Oct. 10, 1913

Cysticercus Ovis

antiporai lobe being slightly larger than the other. Laterally the ovary extends to
the testicular field, but anteriorly is separated from it by a space which is greatest in

Fig. 7. — Sexually mature segments of Taenia ovis. Enlarged. (Original.)

the median line. Posteriorly the testicular field extends beyond the posterior limits
of the ovary but slightly, if at all, and falls short of a transverse line drawn through
the posterior border of the yolk gland.

Gravid uterus (figs. 9 and 10) with 20
to 25 lateral branches from the median
stem. Eggs (embryophores) oval, 30 by
24 to 34 by 2&jj, in diameter.

Hosts. — Larval stage: Sheep ( Ovis
dries); goat {Capra kircus).1 Adult stage:

Dog (Cants familiar is).

Location. — Larval stage : Muscles
(heart, voluntary muscles, esophagus),
more rarely lungs, wall of stomach (?),
and kidneys (?). Adult stage: Lumen
of small intestine.

Localities. — England, France, Germany, Algeria, German Southwest Africa, New
Zealand, and United States.

Type Specimens. — Probably not in existence.

imm.

Fig. 8. — Sexually mature segments of Taenia hydati *
gena. Enlarged. (Original.)

REMARKS ON MORPHOLOGY AND COMPARISON WITH OTHER SPECIES

The larval stage of the sheep-measle tapeworm somewhat resembles
Cysticercus cellulosae in its general morphology. The spirally disposed
neck and head and the mammillate surface of the caudal bladder suggest
the pork cysticercus. The smaller average size and more delicate struc¬
ture of the cysticercus and the shape and number of the hooks, however,
differentiate it quite clearly from C. cellulosae. The hooks are somewhat
slighter in build, have smaller blades, and are different in outline; the
number commonly exceeds the usual number found in C. cellulosae ,
though the limits of variation in number are such in the two forms (24
to 32 in C. cellulosae , according to various authors, and 24 to 36 in C.

1 This record is based on a specimen in the collection of the Bureau of Animal Industry collected in April,
1912, from the heart of a goat about 2 years old, origin unknown, slaughtered at one of the abattoirs in
Kansas City, Mo.

79540— 13 - 3

Journal of Agricultural Research

Vol. I, No. i

ovis) that a definite diagnosis can not be made in individual cases on the
basis of the number of hooks if this number happens to be 32 or less.

Apart from the fact that its normal location is in muscle and not on
serous membranes, Cysticercus ovis may be distinguished from C. tenui-
collis by its smaller size, the different relationship of the head and neck

Fig. 9.— Gravid segments of Taenia ovis. Enlarged. Fig. io.— Gravid segments of Taenia kydar

(Original.) tigena. Enlarged. (Original.)

to the caudal bladder, the presence of mammillate projections on the
surface of the caudal bladder instead of transverse corrugations, and the
different size of the hooks. In C. ienuicollis the head and neck are
invaginated from one end of the caudal bladder instead of from the side,
as in C. ovis (PI. II, figs. 1 and 5). The mammillate projections on the
surface of the caudal bladder of C. ovis (figs. 5 and 1 1 ) are very much in

Oct. io, 1913

Cysiicercus Ovis

10mm

Surface of caudal bladder of Cysiicercus ovis showing papillae.
Enlarged, (Original.)

contrast to the transverse rugae on the caudal bladder of C. ienuicollis
(%. 12).

As the nvmber of hooks of Cysiicercus ienuicollis has been found by
various observers to

vary from 26 to 44, an
accurate distinction be¬
tween this form and C.
ovis which would be
applicable in all cases
can not be drawn on
the basis of the number
of hooks, though, as a
rule, the number of
hooks found in C. ovis
is less than the num¬
ber commonly present
in C. ienuicollis . There
is also an overlapping
in the size of the hooks,
the recorded limits for the large hooks being 170 to 220 ja in C. ienuicollis
(larva and adult) and 156 to 1 88ju in C. ovis (larva and adult), and for
the small hooks no to i6o/r in C. ienuicollis (larva and adult) and 96 to

128/* in C. ovis (larva and
adult).

The hooks of Cysiicercus
ienuicollis , however, average
considerably larger than
those of C. ovis , both in total
length and in length of blade
(fig. 6). In 25 large hooks
from four adult and two
larval individuals of Taenia
hydatigena (C. ienuicollis)
ranging in length from 180
to 212 fit averaging 197/9 the
blade varied from 72 to 10 8/t
in length and averaged 93/*;
and 20 small hooks from the

Hmm.

same specimens ranging in
length from 116 to 136//,
average 12 9/9 had blades

Fig. is. — Surface of caudal bladder of Cysiicercus ienuicollis ranging in length from 60

showing transverse furrows. Enlarged. (Original.) £q 76^ average 68/1. The

average length of 37 large hooks of T. ovis (adult and larva) having a
range of 156 to 1 88g was 173/9 with the blade ranging from 68 to 84/1,

Journal of Agricultural Research

Vol. I, No. i

average 78/*. The average length of 36 small hooks of T, ovis (adult
and larva) having a range of 96 to 128// was 113/*, with the blade
ranging from 48 to 60 fi, average 57/1. In form the hooks of T. hydatigena
and T. ovis are very similar. The small hooks may be distinguished
from each other by the fact that the ventral root, though transversely
enlarged in both species, is rather deeply bifid in T. hydatigena (fig. 6, r),
a condition which is absent in T . ovis or at most only faintly indicated.

Of the more common tapeworms of the dog the one with which Taenia
ovis seems most likely to be confused is T. hydatigena (T. marginata), the
adult of Cysticercus tenuicollis . Apart from the differences exhibited by
the hooks as noted above, the segments of the strobila show certain
characters by which the two species may be differentiated. (PI. II,
figs. 3, 4, 5; and text fig. 6.) The strobila of T. hydatigena is thicker
(dorso-ventrally) relatively to its other dimensions than that of T. ovis
and the latter has a tendency to twist spirally. The segments of T,
hydatigena have a rather regular quadrilateral form, and the edge of the
strobila is comparatively straight, whereas in T. ovis the segments have
convex lateral borders, the convexity usually being well marked, and
the edge of the strobila presents a scalloped outline. The posterior
margin of the segment projects more prominently in the former than in
the latter species. In T . ovis the genital pore is in a large prominent
genital papilla, and there is a large and deep genital sinus; in T. hydatigena
the genital papilla is small and the genital sinus shallow and inconspicuous.
The testicles in T. ovis do not extend posterior of a line drawn through
the anterior border of the yolk gland parallel with the posterior border
of the segment ; in T, hydatigena they extend beyond the posterior limits
of the ovary and yolk gland practically to the posterior border of the
segment (figs. 7 and 8). With respect to the branching of the uterus,
T. ovis and T. hydatigena are quite different, the uterus of the former
having 20 to 25 lateral branches from each side of the median stem,
whereas the uterus of the latter has but 5 to 8 such branches (figs. 9 and 10).

' Other well-known tapeworms of the dog, such as Taenia pisiformis
(T. serrata)y Multiceps multiceps ( T . coenurus), Multiceps serialis (T,
serialis ), Echinococcus granulosus (T. echinococcus ), and Dipylidium
caninum , are less likely than T. hydatigena to be confused with T. ovis .
In addition to distinct morphological differences, the small size of E.
granulosus and D, caninum precludes any chance of mistaking them for
T. ovis, T, pisiformis may be distinguished by the large size of its hooks
(the large hooks being 225/4 or more in length) and the small number
of lateral branches of the uterus (8 to 10). M, serialis may be dis¬
tinguished from T. ovis by the fact that the hooks are considerably
smaller, the recorded limits of length of the large hooks being 135/4 and
1 57/4, that the ventral roots of the small hooks are distinctly bifid, and that
the genital papillae are small and inconspicuous. M. multiceps has large
hooks about the same in length as those of T. ovis but with blades longer

Oct. io, 1913

Cysticercus Ovis

than half the total length of the hook; and as the genital sinus and genital
papilla are very small, the two species may be readily distinguished from
each other.

Of the less common or less known tapeworms of the dog the species of
Dibothriocephalus and Mesocestoides are immediately to be distinguished
from Taenia ovis by the absence of cephalic hooks and rostellum and by
the location of the genital pores in the ventral median line of the segment.
Likewise, the absence of hooks and rostellum distinguishes Ophidioiaenia
punica ( Proteocephalus punicus) 1 from T . ovis.

The remaining species of tapeworms known to occur in the dog are
Taenia balaniceps , T. brauni, T. br achy soma , and T. krabbei , all of which,
with the exception of the last, may be readily distinguished from T. ovis
upon the basis of their published descriptions.

Taenia balaniceps Hall (1910, pp. 139-151, figs. i-8) differs from T. ovis
in various particulars, among which may be mentioned the following:
The worm is smaller, the length of the longest specimen being only 24 cm. ;
the head is smaller, not exceeding 75 2/4 in breadth, and the segments in
corresponding stages of development are smaller. The hooks are smaller,
93 to 9 8/4 being given as the limits of length of the small hooks and 145/*
as the length of the large hooks (fig. 2.) The testicles extend practically
to the posterior border of the segment, as in T. hydatigena . The lateral
branches of the uterus, instead of being slender and more or less separated
by intervening spaces as in T. ovis , are comparatively thick and are
pressed close together.

Taenia brauni Setti, 1897 (Setti, 1897b, pp. 2 10-2 14, pi. 8, figs. 9-14),
differs from T. ovis in that it is much smaller, its total length being from
15 to 18 cm., and the size of the posterior segments 5 or 6 mm. long by 3.5
mm. wide. T. brauni was described as lacking a true rostellum but as
possessing a double crown of 30 hooks, the large hooks measuring 130 to
140 /£, though in some cases only 95 to 100/4 in length, and the small hooks
usually 85 to 90/4, occasionally 70 to 75/1, in length. T. ovis, however, has
a well-developed rostellum and hooks considerably larger than the dimen¬
sions given for T. brauni and is thus clearly a different species from
the latter, though the two forms agree in possessing prominent genital
papillae and perhaps are similar in regard to the branches of the uterus,
as Setti states that the lateral branches are numerous, slender, and per¬
pendicular to the medium stem.

Taenia brachysoma Setti, 1899 (Setti, 1899c, pp. 11-20, pi. 1, figs. 1-9),
is also a much smaller species than T. ovis , specimens with gravid seg¬
ments being not over 10 cm. long and not over 3 mm. in maximum
width. The number of hooks is 30 to 32. The large hooks measure 135
to 145/4 and the small hooks 95 to 105/4 in length, the former thus being
considerably smaller than in T. ovis, and the latter averaging smaller.
The ventral roots of the small hooks are described as having a median
groove, thus presenting a condition intermediate between simple and

1 This species, as pointed out by Hall (1910, p. 146), is probably not a true parasite of the dog.

Journal of Agricultural Research

Vol. I, No. x

bifid, at the same time twisted so that the lateral axis tends to lie in the
plane of the blade and dorsal root.1

The genital papillae are small and inconspicuous in T. brachysoma and
the genital sinus measures not over 170 /4 in maximum depth. The lat¬
eral branches of the uterus number only 10 to 12 on each side of the
median stem.

Taenia krabbei Moniez (1879c, pp. 161-163; 1 880a, pp. 44-50, 56, pi. 1,
figs. 12-14, pl* 2> figs. 4-7) produced in a dog by feeding cysticerci from
the muscles of reindeer is described as much longer, wider, and thicker
than T. coenurus and T. serrata and has much wider segments propor¬
tional to their length, but its head is more delicate. It is also much
larger than T. marginata , the head is larger, and the segments are wider
in proportion to their length. The genital pores are located in large
papillae, often attaining a diameter of 1 millimeter, equal to the length
of the contracted segment. The cysticercus according to Moniez is
much smaller than the cysticercus of T. solium . The number of hooks
varies from 26 to 34. The caudal vesicle, compared to the size of the
head and neck, is very slightly developed and does not contain much
fluid. The orifice of invagination of the cysticercus may be either at
one pole or at one side. The invaginated head and neck commonly curve
spirally as in Cysticercus cellulosae , but to a less degree. The size of the
hooks is not given by Moniez.

If the stated magnification of a drawing by Moniez is correct, the
length of the large and small hooks would be about 215/4 and 160/4, respec¬
tively, but inasmuch as the large hooks of C. tenuicollis , shown in another
drawing, measure, according to the magnification given, about 350/4 in
length, whereas the maximum recorded length is less than 250/4, it is
not unlikely that there has been some error also in stating the magnifi¬
cation of the drawing of the hooks of T. krabbei , so that sizes calculated
from the magnifications of Moniez's drawings can not be considered at
all accurate. Cysticerci in the Bureau of Animal Industry Helmintho¬
logical Collection found in reindeer in Alaska by Dr. D. S. Neuman and
corresponding to T. krabbei , so far as may be determined from Moniez’s
description and figures, except as to the size of the hooks, have hooks
(fig. 6) of the following dimensions: Targe hooks 150 to 170/4 in length,
average 162/4, with blades 75 to 80/4 long, average 77/4; small hooks 85
to 120/4 in length, average 107/4, with blades 52 to 60/4 long, average 57/4
(measurements based on 34 large and 34 small hooks from 8 cysticerci).
The average size of the hooks is thus less than the average of the hooks
in C. ovis, but they show no remarkable difference in form from those
of the latter. Corresponding closely to Moniez's findings, the number
counted on eight heads varied from 26 to 32. The invaginated head and
neck of the cysticercus form a much larger structure than in C. ovis both
actually and relatively to the size of the caudal bladder. On account of

1 Setti does not make it clear whether this twisted condition is invariably present. The small hooks of

Taenia hydatigena commonly present a similar appearance after subjection to the pressure of a cover glass.

Oct. io, 1913

Cysticercus Ovis

their shriveled condition the size of the cysticerci could not be accurately
determined; apparently, however, they are somewhat smaller than C.
cellulosae , rather slender and considerably elongated. The cysticercus
of T. krabbei is readily distinguished from C. ovis by its elongated form,
by the fact that the orifice of invagination of the head and neck is com¬
monly at one end of the cysticercus instead of at the side, and by the
larger size of the body formed by the invaginated head and neck both
* actual and relatively to the size of the caudal bladder. On account of
certain evident similarities, such as the prominent genital papillae, and
on account of the lack of an accurate detailed description of T. krabbei ,
no clear distinctions can be drawn between T. krabbei and T. ovis, though,
no doubt, distinct differences could be found upon comparing specimens
of the two species.

Since the foregoing paragraph was written some of Moniez’s cotypes
have been received from Prof. R. Blanchard, one specimen of the adult
(B. A. I. No. 17351) and two specimens of the cysticercus (B. A. I. No.
17352). The cysticerci, considerably shrunken, measure about 2 by 3
mm. The surface of the caudal bladder is mammillated (as is also the
case in the Alaskan cysticercus), and the cysticercus in this character
thus resembles Cysticercus ovis. The number of hooks was not deter¬
mined, as most of them in the one specimen dissected were lost in
mounting. Two of the large hooks measured 148// in length and had
blades 70 ji long. A small hook measured 1 05/1 in length and had a blade
60 f± long (fig. 6, v, vf). It has thus been determined that the sizes
heretofore assigned to the hooks of Taenia krabbei , based on Moniez’s
drawings, are erroneous and the apparent discrepancy between T. krabbei
and the Alaskan cysticercus, noted in the preceding paragraph, has been
removed. The ventral root of the small hooks is transversely enlarged,
but is not distinctly bifid. A tendency toward the bifid condition, how¬
ever, has been observed in some instances in the Alaskan specimens.
The data thus far available do not indicate a specific difference between
Moniez's species and the Alaskan form, and the weight of evidence is
still in favor of the correctness of the presumption that the Alaskan
cysticercus and T. krabbei are identical. The adult specimen (B. A. I.
No. 1 7351) corresponds closely to the drawing given by Moniez (1880a).
The segments are remarkable for their great breadth, as compared with
their length, and the large genital papillae, about a millimeter in diameter,
are quite conspicuous. As the strobila may be abnormally contracted
in length, too much weight should not, perhaps, be placed upon the
extreme shortness of the segments relative to their width as a feature
by which T. krabbei may be distinguished from T. ovis. It seems prob¬
able, however, that there is a more or less marked difference in this respect
between the two forms. The two posterior segments in the specimen of
T. krabbei , which are gravid, are nearly as long as broad, measuring
about 4 mm. in length by 4.5 mm. in breadth. They are considerably
smaller than the gravid segments of T. ovis. A distinct difference

Journal of Agricultural Research

Vol. I, No i.

between T . ovis and 7\ krabbei is apparent in the gravid uterus. Instead
of the 20 to 30 lateral branches seen in T . ovis there are in T, krabbei
only about 10 lateral branches from each side of the median stem. It
is quite clear from the brief study which has been made of the type
material of T. krabbei that it is specifically distinct from T . ovis , although
the similarity between the two species is very close in many respects.

MACROSCOPIC APPEARANCE OF CYSTICERCUS OVIS

The cyst of the fully developed undegenerated cysticeicus as seen
embedded in the muscles of its host is oval and varies in size from 4 by 2.5
mm. to 9 by 4 mm. or slightly larger (PI. Ill, A and B ). It is whitish in
color and varies in transparency according to the thickness of its fibrous
capsule, which may be very thin and rather transparent or comparatively

thick and rather opaque. In transparent cysts
the head and neck of the cysticercus are apparent
as a small, bright, white spot showing through
the wall of the cyst. Removed from its cyst the
cysticercus (PL II, fig. 1) appears as a small
oval vesicle very transparent and delicate, filled
with a clear fluid, and varying in size when fully
developed from 3.5 by 2 mm. to 9 by 4 mm. On
one side may be seen the opaque white head and
neck invaginated into the vesicle or quite com¬
monly partially evaginated and then projecting
above the surface of the vesicle. Cysticercus
ovis is more delicate in appearance and averages
in size smaller than C. cellulosae. It is considerably smaller than a fully
developed C. ienuicollis.

Degenerate cysts (fig. 13, bt and Pis. Ill, fig. E, and IV, fig. 2) vary in
size, shape, thickness of capsule, and consistency and color of contents. The
sizes of 50 degenerate cysts taken at random varied from 3.5 to 15 mm.
in diameter; 7 by 4 mm. was a common size. Different authors have
observed cysts varying in size from that of a millet seed to that of a bean.
The shape is commonly oval or spheroidal, but may exhibit various
irregularities.

The fibrous capsule of the degenerate cyst may be quite thin or relatively
very thick. For example, the capsule of a cyst from the masseter muscle,
measuring 7 by 4 mm. , was about one-third of a millimeter thick ; another
cyst, 5 by 2.5 mm. in diameter, from the same muscle had a capsule about
three-fourths of a millimeter thick; a cyst 10 by 7 mm. from the heart had
a capsule 3 mm. thick; and the capsule of another cyst, 8 by 6 mm. in
diameter, also from the heart, measured one- third of a millimeter in
thickness. The cavity of the cyst is commonly irregular in shape and
contains besides the cysticercus a mass of caseous, caseo-calcareous, or
calcareous material, or sometimes an albuminous coagulum or a soft
purulent substance. The color of the contents may be white, yellowish,

Fig. 13. — Cysticercus ovipariens (=
C. Ovis): a, Hook,Xi6o; b, cyst
containing cysticercus cut across,
X2. (After Maddox, 1873a, pi.
18 fig. 1.)

Oct. 10, 1913

Cysticercus Ovis

greenish, orange, or brown, and several of these colors may be observed
in the contents of a single cyst. In some cases the cysticercus more or
less shriveled and commonly with evaginated head may be readily dis¬
tinguished upon close scrutiny, but generally is to be found only with
difficulty in degenerate cysts. The dead cysticercus found in degenerate
cysts usually has a bright-white color which makes it more readily
apparent when the contents of the cyst happen to be mostly of a con¬
trasting color. In some of the larger degenerate cysts it is noteworthy
that the cysticerci found have been no larger than those found in much
smaller cysts. For example, the cysticerci found in two degenerate cysts,
10 by 9 and 10 by 7 mm. in diameter, respectively, measured in their
shriveled condition 2 mm. in diameter in one case and 2 >2 mm. in diameter
in the other and thus were somewhat smaller than the shriveled cysti¬
cercus from a cyst 5 by 4 mm. in diameter, which measured 3 by 2 mm.

DISTRIBUTION IN BODY

The cysts of Cysticercus ovis as found in sheep carcasses are usually com¬
paratively few in number and are commonly limited to the heart or
diaphragm, though in many such cases if the muscular parts of the carcass
are cut into slices additional cysts are brought to view. Not uncommonly
cysts may be found in the muscles of mastication and in the tongue.
Sometimes they appear superficially on the muscles beneath the skin,
sometimes in the panniculus carnosus itself. The abdominal musculature
is not uncommonly affected. Degenerate cysts may be found in the
lungs, and in this location they can not be distinguished macroscopically
from the small degenerate cysts of C. ienuicollis . The parasites have
been found in a degenerate condition in the wall of the esophagus.
Degenerate cysts found in the wall of the rumen and fourth stomach in
a lamb which had been fed segments of tapeworm (pp. 23 and 24) were
propably C. ovis . Morot has found degenerate cysts in the kidney which
may have been C. ovis. Degenerate cysticerci in the liver are probably
not C. ovis, but are more likely C. tenuicollis, which frequently occurs in
this location. In the writer’s experiments none of several lambs fed
segments of the tapeworm stage of C. ovis showed any invasion of the
liver, whereas the liver was affected in each of two lambs fed segments of
Taenia hydatigena.

Cysticercus ovis is therefore essentially a parasite of the intermuscular
connective tissue and occurs but rarely in other locations. Except the
heart and diaphragm, the parasite appears to have no distinct preference
for any particular location in the carcass, and the parts named may appear
to be preferred by the parasite simply because these parts are the most
readily examined in post-mortem inspection, so that carcasses which have
these parts affected are likely to be picked out by inspection, whereas
other carcasses which may harbor cysts somewhere in the depths of the
musculature are passed by because they show no cysts in accessible parts.
The muscles of the head, particularly the muscles of mastication, are

Journal of Agricultural Research

Vol. I, No. i

frequently the seat of infestation, and these muscles may be considered
as perhaps a preferred location, though this is uncertain. That the
tongue is a common location has been established by Dr. W. J. Stewart
of the Bureau of Animal Industry, who has found that about one-half of
i per cent of the tongues of all sheep slaughtered at his station are infested.

LOCATION IN SHEEP CARCASSES EXAMINED IN UNITED STATES

In the cases given in Table I the carcasses were examined by slicing the
musculature. The number of cysts found in various locations is given.
The number found in the head in some instances includes cysts found in
the tongue. The columns designated “Superficial” and “Deep” refer,
respectively, to cysts elsewhere than in the heart, diaphragm, and head
which were either found on a superficial examination of the dressed
carcass (Superficial) or were embedded in muscle so that they were found
only on dissection (Deep). Cases Nos. i to 6 were examined by Dr. I. C.
Mattatall at National Stock Yards, Ill.; Nos. 7 to 12 and 13 to 16 by the
writer at Seattle, Wash., and Portland, Oreg., respectively; Nos. 17 and
18 by Dr. R. E. Holm at Wallace, Idaho; No. 19 by Dr. E. C. Joss at
Tacoma, Wash.; Nos. 20 to 25 by Dr. E. C. Joss at Seattle, Wash.;
Nos. 26 to 32 by Dr. E. C. Joss at Portland, Oreg.; Nos. 33 to 35 by the
writer at Chicago, Ill.; Nos. 36 to 38 by Dr. I. C. Mattatall at National
Stock Yards, Ill.; and No. 39 is lamb No. 1 in the experiments already
reported in this article (pp. 23 and 24).

Table I. — Location of Cysticercus ovis in sheep carcasses examined after dissection .

Location of cysts.

Case No.

Heart.

Dia¬

phragm.

Head.

Super¬

ficial.

Deep. )

Case No.

Heart.

[

Dia¬

phragm.

Head.

Super¬

ficial.

Deep.

I .

21 .

1 or 2

2 .

22 .

27 .

i or 2

4 .

24 .... .

1 or 2

c .

2? .

6 .

26 .

7 .

27 .

8 .

28 .

9 . . . ,

20 .

10 .

11 .

21 .

12 .

22 .... .

12 .

72 .

1 or

14 .

OO #

more.

ic .

l6 .

0*r .

2C .

17 .

O j

76 .

18 .

O . .

27 .

IQ .

78 .

-3

20 .

i or 2

30 2 .

X 1

1 X indicates numerous cysts.

2 This carcass also had degenerate Cysticercus ovis in the lungs and wall of esophagus and degenerate cysts
in the wall of the rumen and fourth stomach which were probably C. ovis.

Oct. ro, 1913

Cysticercus Ovis

A carcass examined by Dr. 0. B. Hess at Seattle, Wash., not recorded
above, showed 1 cyst in the heart, 3 in the masseter muscles, 15 in the
forequarters, 22 in the “rack/' 13 in the saddle, and 7 in one hind leg.
The number in the diaphragm or visible superficially was not stated.

Besides the carcasses referred to above there were examined in Chicago
in April, 1912, by Dr. W. C. Siegmund and the writer, 59 carcasses which
had been retained in the course of routine inspection on account of the
presence of cyst in the heart. The examination consisted in examining
carefully the diaphragm and the surface of other exposed muscles,
examining the internal and external muscles of mastication and tongue
after slicing them, and finally examining the cut surfaces after the carcass
had been cut into three to five market cuts.

Four carcasses for which the number of cysts in the heart was not
recorded showed no additional cysts. Fifty carcasses had one cyst in
the heart. Ten of these had additional cysts, three having one cyst
each in the diaphragm, two having one and two cysts, respectively, in
the muscles of mastication, two having one superficial cyst each in the
abdominal musculature and on the hind leg just below the patella,
respectively, three having one cyst each on the cut surface of a hind
quarter, “rack,” and forequarter, respectively, and one having a cyst
in the wall of the esophagus. Three carcasses which had two cysts in the
heart showed no additional cysts. Two carcasses which had three cysts
in the heart showed no additional cysts.

DEGENERATION OF CYSTICERCUS OVIS

The cysticerci observed in the course of the routine post-mortem
inspection of sheep are usually more or less degenerated, and are either
in a condition of caseation or calcification (Pis. Ill, fig. E , and IV, fig. 1).
This does not necessarily indicate that live cysticerci are relatively rare.
It may be accounted for in part by the fact that degenerate cysticerci
are much more conspicuous than the live parasites and, hence, less
likely to be overlooked. On the other hand, the validity of this explana¬
tion is somewhat offset by the possibility that the cysticerci remain alive
only for a short period compared with the length of time they persist in
the degenerated condition, in which event one would expect to find
degenerated cysticerci more often than living ones. How soon degenera¬
tion may begin or how rapidly it may proceed is uncertain, but it is
quite clear that in different instances the process varies considerably in
these respects and in its character as well. Degeneration as noted else¬
where may occur before the cysticerci have reached their full develop¬
ment. It is probably often influenced by the presence of bacteria
introduced by the parasite itself or carried to the cyst by the blood
stream, and bacterial action may perhaps have a great deal to do with
the large size commonly attained by the degenerate cysts of Cysticercus

ovis.

Journal of Agricultural Research

Vol. I, No. i

The results of the experiments described in another part of this paper
prove that degeneration may begin in less than three months after
infection, but no data are at hand to show how soon the process may be
completed; nor, on the other hand, is it known how long the cysticercus
may remain in the tissues of its host before it dies and degenerates.

The various degenerative processes occurring in Cysticercus ovis have
not been worked out in detail and, hence, will not be considered at length.
They are quite similar, at least in some of their variations, to the processes
of degeneration which affect C. bovis and C. cellulosae. A very common
occurrence in the case of C. ovis , as already alluded to, which seems to be
quite unusual in the case of the other two species, is the tendency of
degenerate cysts to reach a size which is very large in comparison with
the cysticercus itself. In some instances it appears that the increase in
size of the cyst may go on indefinitely, fresh calcareous material being
continually deposited in the cyst, associated with a breaking down of
the inner layers of the capsule and a new growth peripherally.

Tike the beef cysticercus, Cysticercus ovis tends to degenerate com¬
paratively early when located in the heart. For example, the cysts in the
heart of a lamb killed 83 days after infestation (p. 24), so far as observed,
were all degenerate. Some of the cysticerci in other locations, including
the muscles of mastication, were degenerate, but the great majority
were alive. Except in the case of the heart, no definite relation has been
observed between the location of the cysticerci and the liability to early
degeneration.

The association of live and degenerate cysticerci in the same carcass
is a matter of interest, though of less practical importance than in the
case of beef and pork measles. In beef measles the association of live
and degenerate cysticerci in the same carcass is fairly common. It is
often stated in regard to Cysticercus cellulosae that if any of the parasites
in an infested carcass are degenerated it is likely that all of those present
will also be in the same condition. This is by no means invariably true.
In a case of pork measles seen by the writer in October, 1912, at an
abattoir in Chicago, most of the cysticerci were undegenerated, but
degenerate cysticerci were quite common, particularly in the diaphragm
and superficial muscles. In the case of C. ovis, so far as the writer's
experience goes, if the cysticerci found in the heart, diaphragm, muscles
of mastication, and other parts of the carcass readily accessible for
examination are degenerated, the cysticerci in other parts of the body are
likewise, as a rule, in a similar condition. Nevertheless, if C. ovis were
transmissible to man, it would be unsafe, when only degenerated cysts
are found on inspection, to pass a carcass for food upon the assumption
that any that might be present in uninspected portions of the muscula¬
ture would also be degenerated. Live and degenerated cysticerci
occasionally, at least, occur together in the same carcass. As noted

Oct. 10, 1913

Cysticercus Ovis

above, a considerable number of degenerated cysticerci were found in a
sheep 83 days after infection, though most of the parasites were still
alive and undegenerated. One other case is recalled in which degenerated
and living cysticerci were associated. In this case the cysticerci in the
heart, diaphragm, and muscles of mastication were degenerated and
partially calcified, as were several found in various portions of the body
musculature, but deep in the muscles of one hind leg there was a live
cysticercus showing no signs of degeneration whatever.

This accords with what would naturally be expected. One would
expect live and degenerate cysticerci in the same carcass as the result,
first, of variations in the longevity of cysticerci, as in the case of the
experimental sheep mentioned above, or, second, as the result of infesta¬
tions occurring at different times. It seems that the latter must surely
occur often. In view of the close association which commonly exists
between sheep and dogs, the sheep in a flock attended by an infested
dog are exposed to the chance of repeated infestation, and, hence, sheep
must frequently harbor simultaneously cysticerci which have come from
eggs ingested on various occasions.

DIAGNOSIS OF SHEEP MEASLES

So far as known, the presence of Cysticercus ovis can not ordinarily be
determined in the living animal, and its diagnosis therefore depends upon
a post-mortem examination. It is not always possible to determine
definitely whether cysticerci found in sheep or goats are or are not C. ovis
without resorting to the use of the microscope, but usually microscopic
examination is not necessary.

The location of Cysticercus ovis in muscle tissue differentiates it clearly
from C. ienuicollis , which, so far as has yet been proved, is found only in
relation with serous membranes. Cases occur, however, in which this
rule can not be applied with certainty, as, for example, when the dia¬
phragm or abdominal muscles are involved it is sometimes practically
impossible to state on the basis of location alone whether the parasite
in question is C. ovis or C. ienuicollis — that is, the parasite may appear
to be in direct relation both with the musculature and the serous mem¬
brane which covers the musculature. Here the size of the cysticercus
may help to determine its identity; if over 10 mm. (two-fifths of an inch)
in diameter, it is C. ienuicollis; if less than this size, it is probably C. ovis,
but may be a young C. ienuicollis .

The relation of the head to the caudal bladder — midway between the
two ends in Cysticercus ovis (Pi. II, fig. 1) and at one end (Pi. II, fig. 5)
in C. ienuicollis — will indicate the species if the parasite happens to be
of a well-marked oval form. Even in very young cysticerci in which the
head is yet rudimentary, the relative position of the head is the same as
in the fully formed cysticercus. Cysticerci affecting the liver of sheep or

Journal of Agricultural Research

Vol. I, No. x

goats may be assumed to be C. tenuicollis. C. ovis has not as yet been
found in the liver. Even in carcasses exhibiting heavy infestation of
the musculature, the liver has not been involved. Small-sized cysticerd
in the lungs, however, may be C. ovis, as degenerate cysticerci of this
species have been found in this location in a case of heavy infestation of
the carcass.

More difficulty is likely to be experienced in the identification of
degenerate cysticerci than of the live parasites, and even more than in
the case of the live cysticerci the location must be chiefly depended upon
in distinguishing macroscopically between Cysticercus ovis and C.
tenuicollis.

The cysticercal nature of degenerate cysts can often be confirmed by
squeezing out the cysticercus, or fragments of it. It should be remem¬
bered that the degenerate cyst may be of a much larger size than the
contained cysticercus, so that the fact that a cyst is larger than the
maximum size of Cysticercus ovis does not necessarily exclude this species
from consideration. Degenerate cysts of C. tenuicollis on the diaphragm
or abdominal muscles commonly become more firmly calcified than those
of C. ovis and show a white, wrinkled surface not seen in the case of
the latter.

Excluding from consideration cases of invasion of the musculature by
the gid bladder worm, whose true nature will be revealed by examination
of the brain and the discovery of characteristic lesions in that location
there are two known conditions which may be mistaken for the degen¬
erate cysts of Cysticercus ovis: Namely, large Sarcocysiis nodules and
encysted foreign bodies, such as barbs from certain plants which work
through the tissues and finally come to rest somewhere in the muscles
and become encysted.

In the case of Sarcocysiis nodules shown in the accompanying illustra¬
tion (PI. Ill, figs. C and D) there were a considerable number of nodules
in the diaphragm and heart, 5 mm. and upward in diameter. The walls
of these cysts were firm and thick, their contents of a purulent nature.
No cysticerci or remains of cysticerci could be discovered. Instead, in
each cyst there were found one or more small, transparent vesicles not
visible except microscopically. These vesicles, with delicate mem¬
branous walls of homogeneous structure without nuclei, contained a
finely granular substance and numerous calciform spores about 15/z long,
which demonstrated conclusively that the cysts were Sarcocysiis cysts.
Usually Sarcocysiis cysts in sheep are so small as to be evident only
microscopically, and cysts large enough to be seen with the naked eye
are, so far as known, very rare. Knowledge of the characteristics of the
unusual forms of Sarcocysiis cysts such as that described above is too
limited to enable one to state definitely the points by which they may be
differentiated macroscopically from degenerate Cysticercus ovis cysts. In

Oct. io, 1913

Cysticercus Ovis

the case of the latter, however, it is frequently possible by opening the
cyst and squeezing out its contents to demonstrate the presence of a
cysticercus or the visible and recognizable fragments of one. Sarcocystis
cysts simulating degenerate C. ovis cysts are, so far as appears from our
present knowledge, of rare occurrence, and consequently cysts occurring
in the musculature of the size and general appearance of degenerate C.
ovis are presumably C. ovis unless there is evidence to show that they are
not, such as, for example, the discovery of Sarcocystis spores and the
total absence of any cysticercus or remnant thereof.

Illustrating the possibility of confusing encysted plant barbules with
degenerate Cysticercus ovis cysts is a case recently observed in which
there was a small nodule about 5 by 4 mm. in diameter in the diaphragm
in the muscle tissue just beneath the serosa. This nodule consisted of a
thin capsule and contents of a somewhat caseous consistency and might
have been taken on casual observation for a small degenerate C. ovis
cyst. Careful examination, however, failed to reveal any morphological
evidence of a cysticercus, instead of which there were found in the midst
of the caseous material three or four tiny barbules from some plant, very
finely pointed and tapering and spirally coiled. These were scarcely
evident to the unaided eye amid the caseous material, but their nature
became quite apparent on microscopic examination.

GEOGRAPHIC DISTRIBUTION

Abroad, cases of sheep measles have been found in England, France,
Germany, Algeria, German Southwest Africa, and New Zealand.

In this country relatively few of the numerous cases found at abattoirs
have been traced to the point of origin of the infested sheep. Cases
traced to the point of origin have been from Montana (10 counties1),
Idaho (5 counties2), Washington (4 counties3), Oregon (11 counties4),
California (3 counties5), Colorado (1 county6), and Nevada (middle and
western part).

The parasite is probably more or less generally distributed throughout
the western United States, and is likely present also in the East, though
as yet no cases have been definitely traced to eastern localities. It is
probable that it will be found to occur wherever sheep are attended by
dogs, particularly wherever dogs have frequent opportunities of devour¬
ing dead sheep.

1 Rosebud, Yellowstone, Meagher, Cascade, Choteau, Hill, Blaine, Lewis and Clark, Teton, and Beaver¬
head Counties.

* Fremont, Bonneville, Bingham, Washington, and Canyon Counties.

3 Adams, Walla Walla, Yakima, and Klickitat Counties.

4 Polk, Benton, Marion, Multnomah, Crook, Gilliam, Morrow, Umatilla, Union, Wallowa, and Baker
Counties.

5 Modoc, Tehama, and Butte Counties.

6 Conejos County.

Journal of Agricultural Research

Vol. I, No. i

PREVALENCE

Most of the published records of sheep measles refer to isolated cases
found by accident, and accordingly indicate little as to the prevalence of
the parasite. Glage (1905), however, in Germany, found degenerate
cysticerci in the muscles of 32 out of 2,198 (1.45 per cent) sheep carcasses
examined for these parasites by inspection of the head muscles and the
heart, and in 16 out of 1,984 (0.8 per cent) in which the heart only was
inspected.

Table II shows the total number of sheep slaughtered at 26 meat-
inspection stations in the United States during 11 months beginning
January, 1912, and the number of carcasses found affected with muscle
cysticerci.

Table II. — Carcasses of sheep found affected with muscle cysticerci during 11 months at
26 meat-inspection stations in United States,

Station.

D.
E
F.
G
H
I.

J-

Total kill.

Affected.

Station.

Total kill.

Affected.

Number.

Number .

Per cent.

Number.

Per cent.

898

O. OI +

0 .

31, 237 1

O. 05 +

262, 361

p .

6, 920

. 01 +

4, 335 > I53

4, 678

. II —

Q .

116, 912

564

. 48+

100, 382

R .

19, 708

109

• 55+

i57><>53

. 01 —

S .

59> 759

61, 905

IO7

• T7~

T .

23, 381

132

• 57“

55, 205

. 02 —

U .

l6l, OIO

1,469

.91 +

237, 799

V .

2, 106

* °5“

1, 488, 997

2,695

.18+

W .

1, 435, 594

5, 739

.4 -

272, 739

. 02 —

X .

526.713

. 01 —

36, 976

. 06+

Y .

166, 581

. 01 +

706, 584

1, 010

■ i5-

Z .

86, 238

74i

.85-

AT 'jSjl

91, 7°4

I, 429, 264

Total . .

II, 60I, 898

17, 446

• 15+

The foregoing table does not indicate accurately the prevalence of sheep
measles. In the first place, many cases would necessarily be missed under
methods of inspection as nearly perfect as practically possible; in the
second place, the methods of inspection for Cysticercus ovis have not been
developed to the same degree of perfection at the various stations; and
finally, at certain stations the methods of inspection for C. ovis have
reached a high degree of efficiency only in recent months, while the figures
given cover also earlier months during which the inspection was less
perfect and during which it may even have happened that no cases were
found at all. For example, it will be noted from Table II that, in the case
of Station R, 0.55 per cent of the sheep slaughtered during January to
November were found to be infested. As a matter of fact, however, the
great majority of the cases of measles at the station were found toward
the close of the period covered; that is, 105 cases, or 2.25 per cent of about
4,300 sheep slaughtered, were found during September, October, and
November.

Oct. io, 1913

Cysticercus Ovis

The actual percentage of sheep infested with measles, at least in those
sections of this country where a close relationship exists between sheep
and dogs, probably approximates 5 per cent much more nearly than it
does the very small percentage derived from the figures given in Table II.

AGE OF INFESTED SHEEP

Information as to the age at which sheep are most likely to be found
infested with measles is meager. A priori it would be expected that
rather young animals would most commonly show infestation. As a
rule, young animals are more liable to infestation with tissue parasites
than old animals, possibly because their tissues offer less resistance to
the migration of the parasites than those of older animals. This greater
susceptibility is offset to some extent by the fact that the longer an
animal lives the more opportunity he has for becoming infested, other
things being equal.

Among a total of 189 infested sheep whose ages (approximate) were
recorded by inspectors of the Bureau of Animal Industry at several
stations, the distribution of cases according to age was as follows :

6 months .

Number
of cases.

. 20

2 to 4 years .

Number
of cases.

8 months .

. 57

3 to 5 years .

. 14

10 months .

. 3

4 years .

. I

1 year .

. 4

5 years .

years .

. 3

6 years .

2 years . ... 63

Owing to the lack of data as to the relative numbers of sheep of these
various ages which are slaughtered, the figures in the above table do
not prove anything. They seem to indicate, however, that Cysticercus
ovis is more commonly met with in young than in old sheep. As one
possible explanation of the apparent rarity of C. ovis in old sheep it is
reasonable to suppose that as the animals grow older any parasites which
they may have picked up in earlier life tend to disappear more or less
completely as a result of degeneration and absorption by the surrounding
tissues. Meanwhile with increasing age the susceptibility to infestation
diminishes, and this, combined with the death and disappearance of
the parasites acquired during youth, tends to result in freedom from
infestation.

ECONOMIC IMPORTANCE

Sheep measles, instead of being as formerly considered a sort of zoo¬
logical or pathological curiosity, is a matter of great importance to the
sheep grower, the butcher, and the consumer of mutton. Although the
tapeworm cysts are not transmissible to man, mutton infested with them
is not a desirable article of food, and modern ideas in meat inspection
require that mutton infested with these parasites to any considerable
79540— 13 - 4

Journal of Agricultural Research

Vol. I, No. i

extent shall either be condemned or rendered into tallow, according to
the degree of infestation, although theoretically there is no objection
from the hygienic standpoint to passing affected mutton for food after
the parasites have been removed. Practically, however, it is impossible
in many cases to remove the parasites, because such extensive dissection
would be required that there would be but little left of the meat when
the parasites had been removed. Consequently, therefore, a large
number of sheep carcasses which are retained by inspectors on account
of measles go either to the tallow tank or to the condemned tank, because
the character of the infestation is such that it is impracticable to remove
the parasites.

At first thought it might seem that the loss on account of these con¬
demnations would fall on the butcher, as the sheep are already bought
and paid for before they are passed upon by the meat inspector, but as
a matter of fact the producer is made to bear at least a part of the loss.
When a condition involving losses on account of condemnations exists
among live stock and continues to prevail, the butchers naturally and
invariably make ample allowances in the prices paid for the probable
loss from condemnations based upon their experience as to losses in
the past, so that the producer, although he may not realize it, is made
to bear more or less of the burden, sharing it, perhaps, with the con¬
sumer, to whom it is likely the butcher will pass on a portion of his loss.

The Federal meat-inspection records, as already noted, indicate that
tapeworm cysts in the muscles of sheep are common throughout the
West, and furthermore, it is safe to say that the proportionate number
of cases of sheep measles found on post-mortem inspection, already
representing a high percentage, will continue to increase as meat inspec¬
tors become more expert in detecting the presence of the parasites.
The natural consequence will be that sooner or later, if this is not already
the case, the sheep raiser will suffer a reduction in the selling price of
his product below that which he would receive were it not for the losses
from condemnations experienced by the butcher.

This indirect loss is in all probability not the only loss experienced
by the sheep raiser. It has already been noted that in the experiments
five of the lambs died in from 13 to 23 days after infestation. These died
approximately in the order of the size of the doses of tapeworm eggs
given, those receiving the smallest doses surviving the longest. Three
of them received only the eggs contained in a single tapeworm segment,
the other two receiving 3 and 10 segments, respectively. The sixth
sheep, which survived, received only one-half of a segment, yet the num¬
ber of eggs was sufficient to make the animal sick for a time, corre¬
sponding probably to the period during which the embryonic worms were
invading and establishing themselves in the muscles. Quite clearly,
therefore, the sheep-measle parasite is deadly in its effects upon sheep,
provided a sufficient number of tapeworm eggs are swallowed, and even

Oct. io, 1913

Cysticercus Ovis

if not enough are swallowed to kill the animal, it may be made sick by
the invasion of the parasites. Accordingly it is quite probable that
many of the cases of death and sickness, which are more or less con¬
stantly occurring among sheep without apparent cause, are the result
of infestation with the measle parasite.

It has been suggested by Dr. S. W. McClure that sheep measles may
be responsible for the many stiff lambs found during spring and summer
on the western sheep ranges.

SIGNIFICANCE IN MEAT INSPECTION

As Cysticercus ovis affects the very part of the carcass which is the
most valuable as food — namely, the musculature — it is of great interest
in meat inspection and of special importance on account of its prevalence.

Under a system of meat inspection which recognizes but one class of
meats as fit for food, such as the system provided for by Federal law in
this country, it is proper to pass for food sheep carcasses which show a
slight infestation with Cysticercus ovis after the removal of the parasites
and any lesions caused by them. Carcasses showing more than a slight
infestation may be rendered into edible tallow, but if heavily infested
should be condemned. As a rule, packers do not take advantage of the
provision which permits moderately infested carcasses to be rendered
into tallow, but prefer to treat such carcasses the same as condemned
carcasses and to manufacture them into inedible products. Though it
is possible that all the parasites may not be found and removed from
slightly infested carcasses, since it is manifestly impracticable to inspect
the depths of the musculature throughout the carcass, it has been deter¬
mined by experience that there is little likelihood that more than one or
two, if any, cysts will be present in the depths of the muscles if only a
few are found in the heart, diaphragm, head muscles, tongue, and other
superficial or readily accessible parts. Accordingly, if only a limited
number of the parasites are found in these locations, there is no reason¬
able objection to passing such a carcass after their removal.

Fven if carcasses are occasionally passed which contain a few cysts
that have escaped observation because hidden in the musculature, no
great harm is done, since the parasites are not transmissible to man and
at most can only offend the esthetic sense of the consumer. Certainly
the consequences of passing such carcasses do not balance the great
waste which would result if all sheep carcasses infested in any degree
whatsoever (amounting to 1, 2, 3, perhaps even 5, per cent of the total
number slaughtered) were excluded from use as food. In the light of
our present knowledge the German regulations are unnecessarily strin¬
gent in placing sheep measles in the same category as pork measles, the
basis of these regulations, of course, being the unproved and apparently
altogether false assumption that the parasite concerned is Cysticercus

Journal of Agricultural Research

Vol. I, No. i

cellulosae, and hence transmissible to man. Under American regula¬
tions concerning Cysticercus cellulosae, necessarily more stringent than
the German regulations because of the absence of a Freibank in our
system of handling meats, no sheep carcass affected with measles even
in the slightest degree could be passed for food if the sheep parasite were
Cysticercus cellulosae . The demonstration of the fact that the muscle
cysticercus of sheep is not Cysticercus cellulosae and that it is not trans¬
missible to man therefore means that many thousands of sheep carcasses
which would otherwise go unnecessarily to the tallow or condemned tank
are saved for food, and thus fortunately one of the factors involved in
diminishing our already too slender meat supply has been eliminated.
Even during the year 1912, when the prevalence of sheep measles was
first recognized and before the inspection for Cysticercus ovis had been
developed to an efficient stage, the money value of sheep carcasses
retained on account of measles amounted to nearly $ 100,000.

The person who kills mutton for his own use need not be so critical
nor so strict with reference to sheep measles as the official meat inspector.
The latter, in the absence of legal provision for a Freibank where meat
not dangerous to human health but of inferior grade can be sold, has to
exclude a great deal of meat from the market which is fit for food under
certain conditions, though it can not properly be passed on the same
basis as meat unconditionally fit for food. Home-dressed sheep car¬
casses, therefore, even though infested in a higher degree than would be
permitted in carcasses which may pass for food under the Federal meat-
inspection regulations may better be utilized for food than wasted,
although here the individual will largely be governed by his own feelings
in the matter, by his squeamishness or lack of it. Such carcasses, how¬
ever, should not be sold, at least not without declaration of their nature,
as they are obviously of less value than carcasses which are free from
infestation.

So far as its detrimental effect on account of the presence of Cysticercus
ovis is concerned, measly mutton may be eaten with impunity unless
the parasites are very numerous or have produced a watery condition or
discoloration of the meat, in which case the carcass should be discarded
even though the prospective consumer may have no objections to the
meat from an esthetic standpoint. In order that further propagation of
the parasites may be avoided, a measly sheep carcass discarded from use
as human food should not be fed to dogs unless it is first cut into small
pieces not exceeding 2 or 3 pounds each and thoroughly boiled.

SURVIVAL OF CYSTICERCUS OVIS AFTER DEATH OF HOST

The length of time Cysticercus ovis may survive after the death of its
host has not been determined. It will, however, live at least six days.
Cysticerci in portions of a carcass shipped from Chicago on March 25,
1913, presumably the day of slaughter, and received in Washington on

Oct. xo, 2913

Cysticercus Ovis

March 28, were still alive on March 31. After its arrival in Washington
the meat was kept in an ice box, at a temperature not lower than 40° F.

As to the period of survival when frozen it was found in one experiment
that the cysticerci in a sheep slaughtered on October 15, 1912, were
dead on November 7, 23 days after slaughter, the mutton meanwhile
having been kept in a frozen condition. Through an oversight no exami¬
nation of the mutton was made at intervening dates, so that no informa¬
tion was obtained as to how long the parasites actually retained their
vitality. The cysticerci were observed by Dr. L. E. Day, who took
charge of this experiment on November 7, to be slightly shriveled after
thawing. On November 7, about half a pound of the infested mutton
was fed to a dog and similar amounts on November 8, 9, 10, and 1 1. On
the last date another dog was also fed. Autopsy on the former dog on
December 2, 24 days after feeding, showed no parasites of any kind in
the alimentary tract. The other dog when examined post-mortem on
January 4, 53 days after feeding, showed a few Dipylidium caninum ,
but no other parasites.

From this experiment it appears probable that a period of three weeks
is sufficient, as in the case of Cysticercus bovis , to insure the death of
cysticerci in mutton. Since, however, Cysticercus ovis is not transmissible
to man, the same necessity of holding slightly affected carcasses in cold
storage for a sufficient period of time to destroy the vitality of any
cysticerci which may have been overlooked does not exist. In this
respect it is accordingly not so important as in the case of Cysticercus
bovis to know how long the cysticerci may survive after the slaughter of
its host.

PROPHYLAXIS

In addition to the highly important preventive measure of destroying
the carcasses of all dead sheep by burning, the simplest, most feasible,
and most effective means of eradication is to keep the dogs of the ranch
or farm free from tapeworms by systematic medicinal treatment. As
the sheep-measle tapeworm in dogs begins to produce eggs about two
months after infection, judging from the results obtained in the experi¬
ments, it is evident that dogs should be treated about every two months
in order to remove any tapeworms acquired since the preceding treat¬
ment before they have developed sufficiently to produce eggs. In prac¬
tice, however, such frequent treatment seems scarcely necessary, and it
is fairly certain that effective control of tapeworm infestation can be
maintained if dogs are submitted to treatment four times a year — that
is, every three months. The following method of treatment is employed
by Dr. E. T. Davison at the Federal Quarantine Station at Athenia,
N. J., and has proved very satisfactory in the case of imported sheep
dogs quarantined and treated on account of tapeworm infestation :

Allow the dog to have the usual feed and drink about 3 or 4 p. m. on the day
preceding treatment, but give nothing further in the form of food or drink, with

Journal of Agricultural Research

Vol. I, No. i

the exceptions noted, until after the medicine has acted. About io a. m., to a dog
of ordinary size, give four io-grain capsules filled with ethereal extract of male fern
(Oleoresina aspidii, U. S. P.), administering at the same time about an ounce of water
or milk, preferably the latter. By a io-grain capsule is meant one which will hold
io grains of quinine. Forty-five minutes later give a second dose, consisting of four
capsules (io-grain) filled with freshly ground areca nut, and with this give as before
about an ounce of water or milk. It is important that the areca nut be freshly
ground. This treatment is usually followed by profuse defecation and the expulsion
of the tapeworm, if any is present, in 30 minutes to an hour after giving the areca nut.
No untoward aftereffects have been noted in any case among several hundred dogs
treated with this remedy. The patient is usually ready for his evening meal.

In administering the medicine an assistant stands the dog up on his haunches and
holds the dog's mouth open by firmly grasping the upper jaw in one hand, the lower
jaw in the other. The capsules are dropped on the back portion of the tongue, and
enough water or milk is thrown in the animal *s mouth to make him swallow. After
administering each of the two doses the dog's head should be tied up as high as he
can hold it and not choke. If this detail is omitted, the patient will almost invariably
throw up the dose. During the remainder of the day the dog should be kept in
confinement and the fecal discharges gathered up and burned, buried, or otherwise
disposed of in such a manner as to prevent the possibility of contaminating the feed
or water of sheep or other live stock.

Incidentally it may be remarked that treating dogs for tapeworm will
rid them not only of the sheep-measle tapeworm but also of other species
of tapeworm whose intermediate stages are found in live stock, one of
which, the gid parasite, fortunately as yet not widespread in the United
States, affects the brain of sheep with almost invariably fatal results.
Though in certain localities coyotes harboring tapeworms are undoubt¬
edly responsible for some of the infestation of sheep with tapeworm cysts,
yet it is the dogs accompanying the sheep more or less constantly day and
night and depositing their feces laden with tapeworm eggs in the imme¬
diate neighborhood of the sheep which are the chief source of infestation,
and if this source is removed by keeping the dogs free from worms, the
sheep can be kept practically free from tapeworm cysts of all kinds.
In addition, it is important that the carcasses of all dead sheep be
destroyed by burning them in order to remove this source of infection of
dogs and coyotes.

SUMMARY

Sheep measles, or tapeworm cysts in mutton, were first recorded in
England in 1866 and the parasite named Cysticercus ovis in 1869 by
Cobbold. C. ovis usually has been considered identical with Cysticercus
cellulosae , the pork-measle parasite, and also has been confused with C.
tenuicollis. It has been considered an intermediate stage of a human
tapeworm, Taenia tenella or T, solium , and also of a dog tapeworm,
T. hydatigena or T . marginata .

Cysticercus ovis is the intermediate stage of a dog tapeworm, Taenia
ovis (Cobbold) Ransom. It may attain its full development in sheep in
less than three months after infection, and in the dog the tapeworm may

Oct. 10, 1913

Cysticercus Ovis

reach the egg-producing maturity in seven weeks after the ingestion of
the cysticercus.

Cysticercus ovis is commonly limited to the heart or diaphragm, but
not infrequently occurs in the muscles of mastication and tongue and
sometimes in various locations in the musculature. It may occur in the
lungs, the wall of the esophagus, or the wall of the stomach. Doubtful
locations are the kidney and liver. It is essentially a parasite of the
intermuscular connective tissue and is evidently rare in other locations.

The cysticerci seen by meat inspectors are usually degenerated.
Those located in the heart tend to degenerate early. Degeneration may
be well established in less than three months after infection. Either
partially grown or fully developed cysticerci may degenerate and may
be associated with living cysticerci in the same carcass as a result of
variations in longevity of the parasites or of repeated infections.

There is no known method of diagnosing the presence of Cysticercus
ovis in the living animal. The parasites are to be recognised in the
sheep carcass by their location in the musculature, by their small size,
and by the lateral position of the head of the cysticercus, C. tenuicollis
being found in relation with serous membranes, being of larger size
when fully developed than C. ovis , and having its head in an apical
position with reference to the caudal bladder. In some cases microscopic
examination may be required to differentiate between these two species.
The possibility exists of confusing degenerate cysticercus cysts with
Sarcocystis cysts and with encysted foreign bodies, such as plant barbules.

Sheep measles have been reported from England, France, Germany,
Algeria, German Southwest Africa, and New Zealand and have been
found in sheep from seven Western States of this country. It probably
occurs wherever sheep are attended by dogs, but has not yet been found
in sheep known to have originated in the eastern United States (p. 45).

Over 17,000 of the sheep slaughtered under Federal supervision during
the year 1912, prior to December 1, were found to be affected with
measles. With the development of more efficient methods of inspection
for Cysticercus ovis the number of cases detected will be relatively much
more numerous. The number of infested sheep in the Western States
probably exceeds, on the average, 2 per cent of the total number. Young
sheep, not over 2 years of age, apparently are more likely to show infesta¬
tion than old sheep.

Cysticercus ovis is of economic importance, first, because of the losses
resulting from the condemnation of carcasses found by the meat inspector
to be more or less heavily infested, and, second, because of the direct
losses which probably occur among sheep as a result of the invasion of
the parasites. The extent of these losses can not be estimated at present.

Cysticercus ovis is of special interest in meat inspection because it
affects the musculature and because it is so prevalent. Carcasses which

Journal of Agricultural Research

Vol. I, No. i

are only slightly infested may properly be passed for food after the
removal of the parasites, but carcasses showing a heavy infestation should
be condemned. Moderately infested carcasses may be rendered into
edible tallow, but are usually treated the same as condemned carcasses
and are manufactured into fertilizer and other inedible products. As
C. ovis is not transmissible to man, meat-inspection regulations concern¬
ing it need not be so stringent in certain respects as those governing
beef measles or pork measles.

The length of time Cysticercus ovis may survive after the death of its
host has not been determined.

The most important preventive measures against the infestation of
sheep with Cysticercus ovis consist, first, in destroying by fire the car¬
casses of dead sheep on the farm or range so that they may not be
devoured by dogs or wolves, and, second, in keeping dogs free from
tapeworms by systematic medicinal treatment. These measures will
also protect sheep from infestation with tapeworm cysts of various
other kinds, which they acquire from dogs.

BIBLIOGRAPHY

ArmbrOster.

1899. Cysticercus celluloses beim Schaf. Ztschr. f. Fleisch- u. Milchhyg., Berlin,

v. 10 (2), Nov., p. 34. [Wa.]

1900. Cysticercus celluloses beim Schafe. [Note.] Ibidem (n), Aug., p 254

[Wa.]

Bongert.

1899 a. Ein Fall von Cysticercus celluloses in der Muskulatur des Schafes. Ztschr.
f. Fleisch- u. Milchhyg., Berlin, v. 9 (5), Feb., p. 86-89, fig- 1-4. [Wa.]

Chatin, Joannes.

1885. See Railliet, Alcide, 1885 a, p. 234.

1886 a. Etudes sur la ladrerie du mouton. M6m. Soc. nation, d'agric. de France,
Paris, v. 130, p. 191-199. [Wc.j

1886 c. Nouvelles recherches sur la ladrerie du mouton. Bull. Acad, de m£d.,
Paris, v. 50, s. 2, v. 16 (39), 3 oct., p. 242-249. [Wm.]

Cobboed, Thomas Spencer.

1866 a. Tapeworms (human Entozoa), their sources, nature, and treatment.
London, vi+83 p., 15 fig. 120. [Wa.]

1866 r. Specimens of cystic Entozoa from veal and mutton. Tr. Path. Soc.
London (1865-66), v. 17, p. 462-463. [MS. dated Apr. 3.] [Wm.]

1869 a. Entozoa; being a supplement to the introduction to the study of
helminthology. London, viii-i-124 p., 3 fig. 40. [Wa.]

1873 c* The internal parasites of our domesticated animals; a manual of the
Entozoa of the ox, sheep, dog, horse, pig, and cat. London, ix+144 P*>
28 fig. 120. [Wa, Wm.]

1873 h. Maddox on an entozoon from the sheep. Lond. M, Rec., v. 1, Aug. 6,
p. 487-488. [Wm.]

1875 a. Revised list of Entozoa, with notes and references. Veterinarian, London
(566), v. 48, s. 4 (242), v. 21, Feb., p. 102-106, (567), s. 4 (243), Mar., p.
169-172. [Wa.]

1879 b. Parasites; a treatise on the Entozoa of man and animals, including some
account of the Ectozoa. London, xi+508 p., 85 fig. 8°, [Wa.]

Oct. io, 1913

Cysticercus Ovis

COLBERG.

1900. Fin Fall von Finnen beim Schafe. [Abstract from Verwaltnngsber. ii. d.
Magdeburg. Schlachthof f. 1898-99.] Ztschr. f. Fleisch- u. Milchhyg.,
Berlin, v. 10 (4), Jan., p. 71. [Wa, Wm.]

Gilruth, J. A.

1908. Report of ... on the Veterinary Division for 1907-8. 16. Rep. Dept. Agric.

New Zealand, Wellington, p. 163-214, 3 pi. [Wa.]

Glage, Friedrich.

1905. Beitrag zur Kenntnis der Kalkkonkremente beim Schafe. Ztschr. f.
Fleisch- u. Milchhyg., Berlin, v. 15 (7), Apr., p. 204-209, [Wa, Wm.]
Ha el, Maurice Crowther.

1910. A new species of cestode parasite ( Tania balaniceps) of the dog and of the

lynx, with a note on Proteocephalus punicus. Washington, p. 1 39-1 51,
9 fig. 8°. [Published Oct. 25.] [Wa.]

1911. Idem. Proc. U. S. Nat. Mus., Washington, D. C. (1780), v. 39, p. 139-151,

fig. 1-9. [Wa, Wc, Ws.]

1911. [A case of unusual delay in development of the adult tapeworm, Multiceps
multiceps.] [Secretary’s abstract of note read before Helminthol. Soc.,
Washington, D. C., Jan. 6.] Science, New York, n. s. (848), v. 33, Mar.
31, p. 510. [Wa.]

Herter.

1910. Vorkommen von Rinder- und Schweinefinnen beim Schafe. [Abstract
from Landwirtsch. Presse.] Deutsche Schlacht- u. Viehhof-Ztg. , Ber¬
lin, v. 10 (9), 27. Feb., p. 117. [Wa.]

K#chenmeister, [Gottlob] Friedrich [Heinrich].

1855 a. Die in und an dem Korper des lebenden Menschen vorkommenden Para-

si ten. Ein Lehr- und Handbuch der Diagnose und Behandlung der
thierischen und pflanzlichen Parasiten des Menschen. Zum Gebrauche
fiir Studirende der Medicin und der Naturwissenschaften, fur Lehrer
der Zoologie, Botanik, Physiologie, pathologischen Anatomie und fiir
praktische Aerzte. Abt. 1: Die thierischen Parasiten. Leipzig, xvi-f
486 p., figs., 9 pi. 8°. [Wa, Wm.]

1878. See Kiichenmeister and Ziim, 1878-81 a, p. 104.

KOchenmeister, [Gottlob] Friedrich [Heinrich]; and ZttRN, Friedrich Anton.
[1878-81 a]. Die Parasiten des Menschen. Leipzig, Aufl. 2, x+iv+ 5-582 p., figs.,
15 pi. 8°. [Wa.]

Leuckart, Karl Georg Friedrich Rudolph.

1856 a. Die Blasenbandwurmer und ihre Entwicklung. Zugleich ein Beitrag

zur Kenntniss der Cysticerctis- Leber. Giessen, 1 p. 1., 162 p., 3 pi. 40.
[Wm.]

1880 b. Die Parasiten des Menschen und die von ihnen herriihrenden Krank-
heiten. Ein Hand- und Lehrbuch fiir Naturforscher und Aerzte. Leip¬
zig und Heidelberg, Aufl. 2, v. 1, Lief. 2, Abt. 1, xii-f 337-856 p., fig.
*3*-353- 8°. [Wa, Wm.]

1886 d. The parasites of man, and the diseases which proceed from them. A
textbook for students and practitioners. Natural history of parasites in
general. Systematic account of the parasites infesting man. Protozoa-
Cestoda. Transl. from the German, with the cooperation of the author,
by William E. Hoyle. Edinburgh, xxvi p., x 1., 771 p., 404 fig. 8°.
[Wa, Wm.]

Maddox, R. L.

1873 a. On an entozoon with ova, found encysted in the muscles of a sheep.
[Read before Roy. Micr. Soc., May 7.] Month. Micr. J., London (54),
v. 9, June 1, p. 245-253, pi. 18-19. [Wa, Wm, Wc.]

Journal of Agricultural Research

Vol. I, No. i

Moniez, R[omain-Louis].

1879 c. Note sur le Tcenia krabbei , esp&ce nouvelle de Taenia arm£. Bull, scient.

d£p. du nord [etc.], Lille, s. 2, v. 2 (5), mai, p. 161-163. [Wm, Wc.]

1880 a. Essai monographique sur les cysticerques. Th&se. Lille, 190 p., 1 1.,

3 pi. 40. [Wm.]

Morot, Charges.

1897. Observation de nodules intramusculaires de nature ind6termin6e, a texture
fibro-cas6euse chez l'agneau et a texture fibro-calcaire chez le cheval.
[Read 10 juin.] Bull. Soc. centr. de mdd. v£t., Paris, v. 51, n. s., v. 15,
30 juin, p. 327-329. [Wa.]

1899 e. Ladrerie et pseudo-ladrerie musculaires chez le mouton. Ibidem, v. 53,

n. s., v. 17, 30 d£c., p. 495-500. [Wa.]

Olt.

1898 b. Cysticercus celluloses in den Muskeln eines Schafes. Deutsche thierarztl.
Wchnschr., Karlsruhe, v. 6 (50), 10. Dec., p. 439-440. [Wm.]

Perroncito, Edoardo.

1900 d. Esiste una Tcenia tenella diversa dalla T. solium ? [Read 13 luglio.]

Gior. r. Accad. di med. di Torino, an. 63, s. 4., v. 6 (8), agosto, p. 814.
[Wm, Wc.]

1900 e. Idem. Gior. r. Soc. ed Accad. vet. ital., Torino, v. 49 (47), 24 nov., p.
1109-1110. [Wa, Wm.]

Railuet, Alcide.

1885 a. Elements de zoologie m6dicale et agricole. Paris, [Fasc. 1], 800 p., 586
fig. 8°. [Published oct.] [Wa.]

(1902). Cysticercose cardiaque chez un mouton. Bull. Soc. v6t. de la Marne, 20
juillet, p. 38.

1904. Idem. [Abstract.] Progr&s v6t., Agen, an. 17, v. 22 (8), 21 f£v., p. 161.
[Wa.]

Railuet, Alcide; and Morot, Charles.

1898 a. Cysticercus tenuicollis dans la paroi du coeur d’un mouton. [Read 2 avril.]
Compt. rend. Soc. de biol., Paris, v. 50, s. 10, v. 5 (13), 8 avril, p. 402-404.
[Wa, Wm, Wc.]

Ransom, Brayton Howard.

1908 d. Occurrence of the Cysticercus of Tcenia solium in sheep. [Read before
Am. Ass. Adv. Sc.] Science, New York, n. s. (703), v. 27, June 19, p.
950-951. [Wa, Wm, Wc.]

1912. Cysticerci in American sheep, reindeer, and cattle. [Note read before Hel-

minthol. Soc., Washington, D. C., Mar. 14.] Ibidem, n. s. (903), v. 35,
Apr. 19, p. 636. [Wa, Wm, Wc.]

1913. An important newly-recognized parasitic disease of sheep. National Wool

Grower, Salt Lake City, v. 3 (1), Jan., p. 30-33. [Wa.]

1913. An important newly recognized parasitic disease of sheep. [Secretary's
abstract of paper read before 12. Meet. Helminthol. Soc. Washington,
D. C., Nov. 21, 1912.] Science, New York, n. s. (941), v. 37, Jan. 10,
p. 78. [Wa, Wm, Wc.]

Rickmann, W.

1908 a. Tierzucht und Tierkrankheiten in Deutsch-Siidwestafrika. Berlin,
xi+364 p., fig. A-D. 8°. [Wa.]

Oct. io, 1913

Cysticercus Ovis

Setti, Ernesto.

1897 a. Nuovi elminti dell'Eritrea. Boll. mus. di zool. [etc.], Genova (57), 50 p.,
pi. 8-9, 41 fig. [Wm.]

1897 b. Idem. Atti Soc. Eigust. di sc. nat. e geogr., Genova, v. 8 (2), giugno,
p. 198-247, pi. 8-9, fig. 1-41. [Wc.]

1899 b. Una nuova tenia nel cane ( Tania brachysoma n. sp.). Boll. mus. di zool.

[etc.], Genova (71), 10 p., pi. 1, 9 fig. [Wm.]

1899 c. Idem. Atti Soc. Eigust. di sc. nat. e geogr., Genova, v. 10 (1), mar., p.
11-20, pi. 1, fig. 1-9. [Wc.]

DESCRIPTION OF PLATES

Plate: II. Fig. i. — Cysticercus ovis from lamb which had been fed eggs of Taenia ovis
(lamb No. i, p. 23).

Fig. 2. — Cysticercus cellulosae. The cysticerci have been extracted from
their cysts. Natural size. (From photographs.)

Fig. 3. — Taenia ovis. This tapeworm was developed by feeding Cysti¬
cercus ovis to a dog (dog No. 6, p. 23). One-half natural size. (From
a photograph.)

Fig. 4. — Taenia hydatigena (T. marginata) from an imported sheep dog.

Fig. — T. hydatigena (T. marginata) from a dog (dog No. 2, p. 21) which
had been fed Cysticercus tenuicollis . In figure 5, diagonally above and
below, are shown two small specimens of C. tenuicollis developed in a
lamb (lamb No. 7, p. 25) by feeding segments of T. hydatigena. One-
half natural size except the two cysticerci, which are shown natural
size. (From photographs.)

Ill (colored). Figs. A and B. — Portions of muscle of sheep showing Cysticercus
ovis (undegenerated) in situ.

Fig. A. — Section of hind leg showing two “deep” cysticerci. Fig. B —
Hind leg showing three “superficial” cysticerci. (Two- thirds natural
size. Original.)

Figs. C and D. — Heart and portion of diaphragm of sheep showing Sarco-
cystis nodules likely to be mistaken for degenerate cysticerci. (Two-
thirds natural size. Original.)

Fig. E. — Sheep heart showing numerous small degenerate cysticerci
( Cysticercus ovis . ) (Two-thirds natural size . Original . )

IV. Fig. 1. — Carcass of sheep showing a degenerate cyst of Cysticercus ovis at
the point indicated by the penknife. (From a photograph by Dr. T.
White and Dr. A. English.)

Fig. 2. — Degenerate cysts of Cysticercus ovis in muscle of sheep; portion
of carcass shown in Plate III, figs. A and B. About natural size. (From
a photograph by Dr. T. White and Dr. A. English.)

Cysticercus Ovis

Plate II

Cysticercus Ovis

Plate III

Journal of Agricultural Research

THE SERPENTINE LEAF-MINER

By F. M. Webster, In Charge , and T. H. Parks, Assistant , Cereal and Forage Insect
Investigations , Bureau of Entomology

INTRODUCTION

The serpentine leaf -miner (Agromyza pusilla Meig., fig. i, a) was
described in 1830 from central Europe1 without definite locality or
host plant. The family to which this insect belongs consists of a group
of small flies the larvae of which are largely leaf-miners. Some, however,
are known to feed upon scale insects,
while Agromyza tiliae Couden2 and
A. magnicornis Loew 3 are known
to make galls on twigs of linden
(Tilia americana) and on leaves of
blue flag (Iris versicolor ), respec¬
tively. Of the species of economic
interest in America Agromyza sim¬
plex Loew occasionally becomes in¬
jurious to asparagus4 by mining
the stems. In Australia A . phaseoli
Coq. seriously injures stems of grow¬
ing beans, 5 while in India stems of
young peas are similarly injured by
a species of Agromyza.6

The habits of Agromyza pusilla as
a leaf -miner of clovers have long been

known, both in Europe and America, and its injuries have been recorded
by some of the earliest students of entomology. With the rapid increase
of alfalfa culture in the United States, especially in the irrigated sections
of the West, the work of this leaf-miner as an enemy of forage crops has
been more and more frequently called to the attention of the Bureau
of Entomology. During the past three years this insect has been the

Fig. i. — The serpentine leaf-miner {Agromyza
pusilla) : a, Adult; b, side view of head ; c, side view
of thorax, showing characteristic color pattern;
d, dorsal view of abdomen, melanic phase; e, out¬
line of thorax, showing location of characteristic
bristles. Much enlarged. (Original.)

1Meigen, J. W. Systematische Beschreibung der Bekannten Europaischen Zweiflugeligen Insekten.
T. 6, Hamm, 1830, p. 185.

2 Couden, F. D. A gall-maker of the family Agromyzidae. (Agromyza tiliae, n. sp.) Proc. Ent. Soc.
Wash., v. 9, p. 34-36, fig. 1, 1907. 1908.

3 Thompson, M. T. Three galls made by cyclorrhaphous flies. Psyche, v. 14, no. 4. P- 74. fig- 3, Aug.,
1907.

4 Chittenden, F. H. The asparagus miner. ( Agromyza simplex Loew.) U. S. Dept. Agr., Bur. Ent.
Circ. 135, 5 p., 2 fig., 1911.

6 Froggatt, W. W. The French bean fly. ( Agromyza phaseoli , Coquillett.) Agr. Gaz. N. S. Wales, v. 22,
pt. 2, p. 151--154, Feb., 1911. Also pub. as N. S. Wales Dept. Agr. Misc. Pub. No. 1399.

6 Maxwell-Lefroy, Harold. Indian Insect Life. Calcutta and Simla, 1909, p. 622-623.

Vol. I, No. 1
Oct. 10, 1913
K— 1

(59)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Journal of Agricultural Research

Vol. I, No. i

subject of investigations and observations made by several members
of the Section of Cereal and Forage Crop Insect Investigations, and the
following results are herein set forth regarding this leaf-miner as an
enemy of alfalfa (Medicago sativa) and other forage crops in America.

SYNONYMY

Mr. J. R. Malloch, recently of the Bureau of Entomology, after making
a careful study of specimens from Europe and also of a large amount of
material from widely separated localities in the United States, includes as
synonyms of Agromyza pusilla the following names heretofore supposed
to apply to valid species :

A. pusilla Meig., A. pumila Meig., A. strigata Meig., A. exilis Meig., A. amoena
Meig., A. puella Meig., A. pusio Meig., A. orbona Meig., A. blanda Meig. (?), A. dimi-
nuta Walker (?), Oscinis trifolii Burg., Oscinis brassicae Riley.

HISTORY OF THE SPECIES IN EUROPE

According to Schiner, “ the larvae mine the leaves of Euphorbia cyparis-
sias ,” the cypress spurge, also called “ quacksalver's spurge,” which
according to Britton and Brown has escaped from gardens to the road¬
sides and waste places in the Atlantic States.

The same authority quotes Bouch£ as stating of Agromyza amoena Meig.
that “the larvae mine leaves of Sambucus nigra, the common European
elder.”

Kaltenbach records observing the larvae of Agromyza trifolii mining in
the leaves of Trifolium medium in June and in those of T. repens (white
clover) in September. He also says of A. strigata: “The mining larva
lives in leaves of Campanula trachelium (bellflower).”

Goureau,1 in 1861, records Agromyza nigripes, a related European
species, as mining in the leaves of Medicago sativa (lucem), in Europe,
and his description of the habits and injury caused by these miners is
very similar to that which might be given of A . pusilla and its injury to
alfalfa in America.

Decaux,2 in 1890, records A. nigripes as mining the leaves of lucem
in France, and in the infested area estimates a loss of from 20 to 25
per cent of the crop due to the injury to the lucem leaves by this miner.

Groult,3 in writing of A. nigripes in France, records the mines during
August and September in fields of lucem and states that where large
numbers of the mines were present the devastation became noticeable
and the injured lucem made poor forage.

1 Goureau, Charles. Les insectes nuisibles aux arbres fruitiers, aux plantes potageres, aux cerdales et aux
plantes fourrageres. Bui. Soc. Sci. Hist, et Nat. de l’Yonne, v. 15, p. 76-454, juill., 1861. “ Agromyza
nigripes ,” p. 385-386.

2 Decaux, Francois. [ Agromyza nigripes Meig.] Ann. Soc. Ent. France, ser. 6, t. 10 [Bui.], p. ccvi-
ccvii, nov. 26, 1890.

s Groult, Paul. U Agromyza nigripes. Ee Naturaliste [Paris], an. 30 (ser. 2, an. 22), no. 517. P- 219-220,
sept. 15, 1908.

Oct. 10, 1913

Serpentine Leaf-Miner

Fig. 2. — Map showing known distribution of the serpentine leaf-miner throughout the world.

Journal of Agricultural Research

Vol. I, No. i

Mr. H. S. Smith, formerly of the Bureau of Entomology, noticed
dipterous larvae mining leaves of lucem in fields in Sicily, Italy, and
France during the spring of 1912, and from a pupa taken in one of these
mines, collected in Sicily during the last week of December, 1911, reared
A gromyza nigripes . He reports the work of this species in Europe as
similar to that of the alfalfa leaf-miner in America with which he is
familiar. Apparently the larva can be found mining in the lucern leaves
in the latitude of Sicily during the entire winter.

DISTRIBUTION OUTSIDE OF THE UNITED STATES

Outside of the United States this species has been found in middle,
central, and northern Europe — Italy, Sicily, Egypt, England, Scot¬
land, and Ireland. Its general distribution is shown in the map of the
world (fig. 2).

DISTRIBUTION WITHIN THE UNITED STATES

The general distribution of the species in the United States, excluding

Alaska and the insular
possessions, extends
from the coast region
of central New Jersey
southward to southern
Florida and westward
to southern California
and northwestern
Washington. It also
occurs about Hono¬
lulu, Hawaiian Islands.
(See map of the United
States, fig. 3.)

Specimens are in the collection of the United States National Museum
from the following localities :

Washington, D. C. (Coquillett and Pergande); Foristell, Mo. (Riley); Los Angeles,
Cal. (Coquillett); Las Cruces, N. Mex.; Douglas County, Kans.; Flagstaff, Ariz.; Wil¬
liams, Ariz. (H. S. Barber); Honolulu, H. I.; Iowa; Whittier, Cal. (P. H. Timberlake);
Biscayne Bay, Fla.; Texas (Belfrage); Plano, Tex. (E. S. Tucker); Cotulla, Tex. (F. C.
Pratt); Victoria, Tex. (Hunter).

Specimens in other collections are from the following localities :

Ocean County, N. J. (Dr. John B. Smith); Portland, Greg. (Melander); Moscow Mt.
(Melander); Mt. Constitution, Winlock, Port Gamble, Woodland, Palouse, Monroe,
and Olga, Wash. (Melander); Pullman, Wash. (Melander and Hyslop); Philadelphia,
Pa. (Henry ICraemer); Danbury, Conn.; Blue Hills, Woods Hole, Aubumdale, and
Chatham, Mass. (C. W. Johnson).

Fig. 3. — Map showing distribution of the serpentine leaf-miner within
the United States.

Oct. io, 1913

Serpentine Leaf-Miner

FOOD PLANTS IN EUROPE

According to Brischke, Brauer, and Kaltenbach the following host
plants in Europe are given for Agromyza pusilla and its synonyms:

Agromyza pusilla Meig.:

Spiraea ulmaria (meadow queen).
Solarium tuberosum (potato).
Hyoscyamus niger (henbane, hog bean).
Galeopsis telrahit (hemp nettle).
Stachys sylvantrica (hedge nettle).
Euphorbia cyparissias (cy¬
press spurge).

Agromyza strigata Meig.:

Campanula trachelium (bellflower).
Taraxacum geniculata (dandelion).
Sonchus oleraceus (sow thistle).

Agromyza strigata Meig. — Continued.

Beilis perennis (garden daisy).
Agromyza trifolii Burg.:

Trifolium repens (white clover).
Trifolium medium (zigzag clover).
Agromyza orbona Meig.:

Ononis spinosa (rest-harrow).

Ononis repens (rest-harrow).
Agromyza variegata Meig.:

Colutea arborescens (bladder senna).
Agromyza amoena Meig.:

Sambucus nigra (European elder).

FOOD PLANTS IN AMERICA

Besides alfalfa, this species has been reared in the United States from
the following plants, given here with the locality, date, and collector:
Cabbage ( Brassica oleracea):

St. Louis, Mo., June 17, 1876 (C. V. Riley); Georgetown, D. C., July, 1882 (Theo.
Pergande); Los Angeles, Cal., September, 1887 (D. W. Coquillett); Ames, Iowa,
date unknown (Herbert Osborn), reared from stems; Washington, D. C., May and
June, 1900 (Theo. Pergande); Athens, Ga., June 7, 1900 (Theo. Pergande); Browns¬
ville, Tex., February, 1908 (D. K. McMillan); Orlando, Fla., March 24, 1908 (H. M.
Russell); Honolulu, H. I., September, 1910 (H. O. Marsh), abundant and destruc¬
tive; La Fayette, Ind., May, 1912 (W. J. Phillips and Philip Luginbill).
Nasturtium :

Washington, D. C., July, 1897 (D. W. Coquillett), Arlington, Va., June 30, 1906
(I. J. Condit).

Radish ( Raphanus sativus) :

Honolulu, H. I., July, 1906 (Jacob Kotinsky); Washington, D. C., July, 1907
(C. H. Popenoe).

Potato {Solanum tuberosum) :

Foristell, Mo., June 3, 1876 (C. V. Riley).

Turnip ( Brassica rapa) :

Washington, D. C., July 30, 1906 (I. J. Condit); Corpus Christi, Tex., January 22,
1908 (D. K. McMillan); Arlington, Va., August, 1909 (E. G. Smyth).

Spinach ( Spinacia oleracea ):

San Francisco, Cal., 1907 (E. M. Ehrhom).

Watermelon ( Citrullus vulgaris ):

Orlando, Fla., June 13, 1907 (H. M. Russell).

Garden beet ( Beta vulgaris) :

Honolulu, H. I., 1906 (Jacob Kotinsky).

Sugar beet ( Beta vulgaris ):

Compton, Cal., April 13, 1910 (H. M. Russell) (adults reared from pupae collected
on leaves); Elsinore, Utah, August 5, 1910 (E. G. Titus) (adults collected on sugar
beets).

Pepper ( Capsicum sp.):

Brownsville, Tex., February, 1909 (D. K. McMillan).

7954°— 13 - S

Journal of Agricultural Research

Vol. I, No. i

Vetch ( Vida sp . ) :

Columbia, S. C., June 15, 1913 (Philip Luginbill).

Sweet pea ( Lathyrus odoratus) :

Tempe, Ariz., May 24, 1912 (V. L. Wildermuth); Sacaton, Ariz., May 25, 1912
(R. N. Wilson); Salt Lake City, Utah, June, 1911 (C. N. Ainslie).

Fenugreek ( T rigonella foenum-graecum) :

Salt Lake City, Utah, July 22, 1911 (T. H. Parks).

White clover ( Trifolium repens , PI. V, fig. 2):

Washington, D. C., June, 1879 (Theo. Pergande); Oxford, Ind., 1884 (F. M. Web¬
ster); Washington, D. C., September n, 1907 (C. N. Ainslie); Salt Lake City, Utah,
1911-12 (C. N. Ainslie and T. H. Parks); Lyman, Wyo., July 14, 1911 (T. H. Parks).
Red clover ( Trifolium pratense):

Salt Lake City, Utah, June to September, 1911 (T. H. Parks); Twin Falls,
Idaho, July, 1912 (T. H. Parks).

Sweet clover (Melilotus officinalis ):

Tempe, Ariz., May 14, 1912 (V. L. Wildermuth).

Rape ( Brassica napus , PI. V, fig. 1):

La Fayette, Ind., 1909 (W. J. Phillips); La Fayette, Ind., 1911 and 1912 (W. J.
Phillips and Philip Luginbill).

Cowpea ( Vigna unguiculata ):

Batesburg, S. C., July 12, 1904 (E. G. Titus); Lakeland, Fla., May 8, 1912 (G. G.
Ainslie; La Fayette, Ind., July and August, 1912 (Philip Luginbill); Columbia,
S. C., July 10, 1908 (G. G. Ainslie), September n, 1912 (Philip Luginbill), Como,
Miss., August, 1912 (T. H. Parks).

Cotton ( Gossypium barbadense):

Batesburg, S. C., 1912 (E. A. McGregor); Dallas, Tex., 1912 (A. Rutherford).
Tobacco ( Nicotiana sp.):

Chatham, Va., July, 1906 (W. W. Green).

Hedge mustard ( Sisymbrium officinale) :

Washington, D. C., June, 1900 (F. H. Chittenden and Theo. Pergande); Welling¬
ton, Kans., May, 1912 (E. O. G. Kelly).

Smooth rock cress (Arabis laevigata ):

Washington, D. C., June, 1900 (F. H. Chittenden and Theo. Pergande).

Plantain ( Plantago sp.):

Salt Lake City, Utah, July, 1912 (C. N. Ainslie).

Common mallow ( Malva rotundifolia ):

Tempe, Ariz., October, 1911 (V. L. Wildermuth).

The great variety in the food plants of the larvae, together with the
fact that the peculiar shaped but rather inconspicuous larval mines in
the leaves (PL V, figs, i and 3) do not readily attract attention except
when excessively abundant, leads to the suspicion that the insect may
occur unobserved in many localities not indicated on the map (fig. 3).
This is perhaps especially true throughout the West wherever it becomes
sufficiently abundant in alfalfa fields to be a pest. Therefore, in this
paper, it is considered with special reference to alfalfa culture.

RECORDS OF THE BUREAU OF ENTOMOLOGY

The earliest published record of this insect was by the late Dr. C. V.
Riley, who appears to have first reared the fly from larval mines in the
lower leaves of potato received from Foristell, Mo., June 3, 1876, other
individuals issuing later. At that time it was supposed to be an Oscinis.

Oct. io, 1913

Serpentine Leaf-Miner

On June 17, 1876, Dr. Riley noted that cabbage leaves in the vicinity
of St. Louis, Mo., were infested by some leaf-mining larvae, and from
these mines a single female fly was reared June 30, the larva pupating
underground. Several years later, when apparently the same insect was
found mining the leaves of cabbage, June 25, 1882, in Georgetown, D. C.,
by Mr. Theo. Pergande, interest in Dr. Riley's previous rearing from
cabbage leaves in St. Louis, Mo., appears to have been revived. In
1884 1 2 Dr. Riley described the species as Oscinis brassicae , evidently
failing to recognize as identical his former rearing from mines in potato
leaves, but calling attention to the similarity between his species and
Oscinis trifolii Burgess, which had been described five years before. This
same year (1884) the senior author found the same species in large num¬
bers attacking the leaves of white clover (Trifolium repens) at Oxford, Ind.

Three years after its first discovery in Missouri by Dr. Riley and during
June, 1879, the insect was observed to be very abundant about Washing¬
ton, D. C., attacking the leaves of white clover, and was carefully studied
by Mr. Theo. Pergande. It must be borne in mind that at that time (1879)
it was not positively known to attack clover or other plants elsewhere,
and as a result of Mr. Pergande’s labors adult flies were secured which
were afterwards described by Mr. Edward Burgess as Oscinis trifolii ?

In 1898 the late Mr. D. W. Coquillett, after examining the types of
both Oscinis brassicae Riley and O. trifolii Burgess, decided that both
were synonyms of Agromyza diminuta Walk.3 Further results are shown
by Mr. MalloclTs studies.

Its wide distribution in the alfalfa-growing section west of the Rocky
Mountains was especially noted by the junior author during the summers
of 1911 and 1912, when, during the months of June, July, and August, the
larvae were found mining in the leaves of alfalfa at almost every point
visited in connection with the investigation of the alfalfa leaf-weevil
(Phytonomus posticus Gyll.). The territory covered by these observa¬
tions comprises most of the alfalfa-growing section of Utah, southern and
western Idaho, and southwestern Wyoming. In fact, the mines were
present in limited numbers wherever alfalfa was found growing and in
places widely separated by the uncultivated desert. This may be illus¬
trated by quoting from field notes made at Lucin, Utah, August 20, 191 1 :

In a small field of alfalfa irrigated from a spring and in the midst of a desert west
of Great Salt Lake these leaf-miners were of common occurrence. There is no alfalfa
to the east for fully 90 miles and to the west for a distance of about 60 miles, this field
being just 6 miles from the Utah-Nevada State line. Both larvae and pupae were
observed.

1 Riley, C. V. The cabbage Oscinis (Oscinis brassicce n. sp.). TJ. S- Comr, Agr. Rpt. 1884, p. 322, pi. 8,

fig- 5.

2 Riley, C. V. The clover Oscinis. (Oscinis trifolii Burgess [n. sp.]). U. S. Comr. Agr. Rpt. 1879,
p. 200-201, 1880.

3 Coquillett, D . W. On the habits of the Oscinidae and Agromyzidae, reared at the United States Depart¬
ment of Agriculture. U. S. Dept. Agr., Bur, Ent., Bui., n. s., no. 10, p. 78, 1898.

Journal of Agricultural Research

Vol.I.No. i

Adults and pupae were collected at Boise, Idaho, by Mr. H. T. Osborn,
of the Bureau of Entomology, August 22, 191 1 ; and from mined leaves of
alfalfa received from Sarah A. Armstrong, July 3, 1905, from Fort Collins,
Colo., adult flies of this species developed en route.

Its distribution extends westward to the Pacific coast, and throughout
the irrigated sections of Washington, Oregon, and California. In a
communication dated January 25, 1912, from Mr. Wyatt W. Jones, of
Redding, Cal., the winter states that his attention has frequently been
called to a minute leaf-miner in alfalfa, very common in that region during
August and September. His attempts to rear adults resulted in securing

only parasites. On May 14, 1912,
Mr. Jones collected larvae and pupae
from young alfalfa plants grown from
seed sown in March of that year.

Mr. V. L. Wildermuth, who has
made a careful study of this insect in
the Imperial Valley of southern Cali¬
fornia and in Arizona, finds it very
generally distributed over the alfalfa¬
growing section of the Southwest,
where its injury to the hay crop is
probably greatest. It has also been
swept from alfalfa at Glendale, Cal.,
by Mr. T. D. Urbahns.

These flies were reared from larvae
mining alfalfa leaves at Wellington,
Kans., by the junior author in July,
1910, and again by Mr. E. O. G. Kelly,
of the Bureau of Entomology, at the
same place during the summer of
1912. While the injury was not
severe, Mr. Kelly reported plants with
from 12 to 20 mined leaves common
during June.

Two adults and numerous parasites
were reared from alfalfa leaves col¬
lected at Manhattan, Kans., by Mr. C. N. Ainslie in July, 1907. Mr.
Ainslie also reared adults and parasites from infested leaves of alfalfa
collected at Mesilla Park, N. Mex., May 21, 1909, and reported two or
three mines in one leaflet not uncommon in the lower leaves of plants in
a field of very young alfalfa.

Specimens have been collected from altitudes varying from below sea
level in southern California to 7,000 feet above sea level elsewhere.

Fig. 4.— Alfalfa leaf with eggs of the serpentine
leaf-miner in situ, somewhat enlarged, a,
Bgg, greatly enlarged; b, same, in situ, with
parenchyma of leaf partly dissected away,
much enlarged. (Original.)

Oct. io, 1913

Serpentine Leaf-Miner

Throughout the entire West the mines were found in limited numbers
wherever alfalfa is grown.

From the occurrence of the larvae and pupae in such widely scattered
points we are led to believe that the insect has long been established
throughout the alfalfa-growing sections of the West.

While this leaf-miner does not constitute a widespread menace to the
alfalfa crop, it works considerable damage in New Mexico, Arizona, and
southern California, because leaves mined by the larvae are unfit for
fodder; besides, the changed color of the hay reduces its market value,
especially if grown mixed with timothy.

DESCRIPTION OF THE LEAF-MINER, AGROMYZA PUSILLA.

THE ADULT (FIG. i)

In view of the great number of synonyms and the impossibility of
giving descriptions of all of these in this article, Mr. Malloch has drawn
up the following description, based on a large number of specimens
in the collections of the Bureau of Entomology and the United States
National Museum, the better to facilitate the recognition of the insect as
it occurs in America.

Male and Female. — Black, shining, marked in most variable degree with yellow.
Frons, except ocellar region and sometimes a narrow side stripe posteriorly, yellow;
remainder of head parts, except behind vertex, yellow. Mesonotum with a more or
less broad yellow margin which never extends distinctly around the anterior or the
posterior margin; four pairs of dorsocentral bristles present, as well as numerous short
hairs on disk; humeri with a black spot. Pleurae sometimes yellow, with a brownish
spot above and shortly behind the coxae, another large one covering the space between
the fore and mid coxae, and another one between the mid and hind coxae; at other
times almost entirely black, with the sutures and upper margin yellow\ Scutellum
entirely yellow, or yellow with black basal side spots, which in some cases extend
almost around the entire margin and on to the disk. Postnotum black. Abdomen
yellowish, with dark brownish bases to segments; or black, with pale apices to seg¬
ments; or entirely shining black, with the apical segments whitish or yellowish at
apex. Legs varying from almost entirely yellow, with only the tarsi brownish, to
almost entirely black, with knee joints yellow; the femora generally less intensely
black than other parts of legs. Mid tibiae without distinct posterior bristles. Wings
clear; second division of costa about two and one-half times as long as first section,
third and fourth veins divergent at extremities; outer cross vein as long as or slightly
shorter than the section of fourth anterior to it; basal two sections of fourth subequal
or the second slightly the shorter; last section of fifth vein about three times as long
as preceding section. Hal teres yellow.

Length, 1 to 1.75 mm.

This is a most variable species in color and is very widely distributed.

THE EGG (FIG. 4)

The eggs are pale, white, oval, about 0.25 mm. long, and can be frequently seen
through the epidermis from above. Figure 4, b, shows the egg partly dissected out
of one of these pits.

Journal of Agricultural Research

Vol. I, No. i

The: LARVA (FIG. 5)

Larva, newly hatched, about 0.12 mm. in length, nearly white, but soon turning
yellowish. When fully developed , it averages nearly 3 mm . in length , fully extended ,

and is bright translucent yellow, the black, chitin-
ized mouth parts, tracheal system, and dark con¬
tents of the posterior alimentary canal being plainly
visible through the body walls. Form subacute
anteriorly, increasing rapidly in diameter caudad
Fig. s • I#arva of the serpentine leaf- for about one-third of its length, then gradually

miner, lateral view. Enlarged, diminishing posteriorly to the bases of the anal

(Origiiial.) spiracles, where the body becomes rather suddenly

truncate, terminating abruptly. Anal spiracles large, porrect, extending beyond end
of cauda. Body segments visible and each encircled by a band, granular in appear¬
ance, which is sprinkled with minute setaceous tubercles. Anterior
spiracles much smaller than posterior, somewhat chitinized at tips,
knobbed, and situated in a slight depression.

Upon the ventroanal surface there occurs a tubercular, suckerlike
organ, in addition to which is a pair of rather large ventrolateral
tubercles placed between the anal spiracles and the organ mentioned
above. (Description by W. R. Walton.)

THE PUPARIUM (FIG. 6)

Puparium slightly less than 2 mm. in length, greatest width about 0.8
mm. Oblong oval in form, slightly flattened, the sides sinuate or
fluted in outline. Segments strongly marked. Bright yellow in
color when freshly pupated, gradually darkening to brown as the
development of the pupa progresses. Surfaces slightly shining, but
without sculpture. Anterior and posterior spiracles prominent, as
shown in figure 6. (Description by W. R. Walton.)

HIBERNATION

Mr. George G. Ainslie finds that at Lakeland, Fla., the larvae of the
serpentine leaf -miner may continue feeding throughout the entire winter.
They were observed by him mining in cowpeas in January, 1913. In the
Salt River Valley of Arizona Mr. V. L. Wildermuth finds that during
mild winters the larvae may mine in the leaves until after Christmas.
Ordinarily, however, in that locality, the larvae go into hibernation late in
November. At Brownsville, Tex., although we have no information
relative to this species, Mr. R. A. Vickery finds that other insects do not
hibernate at all, which agrees with what Mr. Ainslie observes to be true
of this species in Florida.

It would seem, therefore, that the species hibernates north of Florida
and extreme southern Texas and that, so far as known, hibernation takes
place only as pupas on or beneath the surface of the soil. In the North
only a small percentage of the last generation in the fall lives to enter
hibernation at all, owing to the fact that the larvae continue feeding in
their mines until late in the autumn, large numbers in this way being
killed annually by the early freezes of October and November. In the
Salt Lake Basin in Utah this insect begins to enter hibernation during

Fig. 6.— Pupa-
rium of the
serpen tine
lea f-miner,
ventral
view. En¬
large d .
(Original.)

Oct. 10, 1913

Serpentine Leaf-Miner

October, although many larvae continue mining until killed by frosts.
Moreover, a very large percentage of the larvae in the mines are parasitized
at this time, which greatly reduces the number of healthy pupae that
would otherwise enter hibernation. The junior author, in an effort to
secure hibernating puparia at Salt Lake City in January, 1912, gathered
old alfalfa leaves and loose soil from irrigation-ditch banks where the
mines had been common during the summer of 1911, but only parasites
issued from this material.

Healthy puparia formed late in October pass the winter in that stage
in the latitude of northern Indiana.

Hibernation takes place largely in waste places where volunteer alfalfa
is found growing. In the arid country of the West such patches of alfalfa
can be found everywhere along irrigation-ditch banks, fence rows, and
railway right of ways, where it escapes from cultivation.

BEGINNING OF ACTIVITY IN SPRING

Adults emerging from hibernation are abroad in April in southern
California and Arizona and during the month of May in the intermountain
region farther north. Evidently they do not travel far before oviposition
takes place. As an indication of this it was noticed, both in Utah and
again in Arizona and California, that the first mines observed in spring
were usually either confined to the foliage of a single plant or scattered
more or less sparingly over two or three adjoining plants. The occupants
of these mines, whether larvae or pupae, were all in nearly the same stage
of their development, thus indicating that the eggs were either deposited
by a single female, or, if by more than one, at about the same date.
It was noticed, also, that the female confines her oviposition to a small
area, usually placing only one egg in a leaf. In the Salt Lake Ba^in the
first mines in spring were usually found clustered on volunteer plants
along irrigation-ditch banks, where the insect probably had hibernated.

OVIPOSITION AND THE EGG PERIOD

The eggs are deposited in the cellular tissue of the leaf, and the process
of oviposition has been observed by several members of the Section of
Cereal and Forage Insect Investigations of the Bureau of Entomology.
The female deposits the egg from the underside of the leaf, frequently
near the margin, where she can anchor herself by hooking the tarsal
joints over the edge during oviposition. The fly inserts the ovipositor
into the tissues, thrusting the tip of the abdomen against the leaf and
puncturing the tissues with her ovipositor. She enlarges the opening
thus made by a rotary motion of the abdomen and places the egg well
up into the cellular tissue against the epidermis on the upper surface.

After the female has finished enlarging the opening she turns around
and sucks up the sap from the aperture, after which she is soon engaged

Journal of Agricultural Research

Vol. I, No. r

in making another incision in the leaf, where she repeats the feeding
operation. When several females are confined on one plant the under¬
side of the leaves soon becomes pitted with these feeding punctures made
with the ovipositor. Only a small percentage of the punctures contain
eggs, as the main function of the punctures seems to be to furnish food
for the adults. The larval mine always commences at this little hole
or pit.

The females in confinement readily feed on sugar water, and, no doubt,
nectar furnishes a part of their food, although no field observations
prove this.

The egg period lasts from three to eight days, varying with the seasons
of the year, but the average period of incubation can be considered as
four days.

HABITS OF LARVA AND LENGTH OF LARVAL PERIOD

The larva (fig. 5) commences feeding immediately after hatching and
starts mining through the tissues just beneath the upper surface. The

mine at first is very small and threadlike,
gradually widening with the growth of the
larva. Often the miner encircles the entire
leaf at first and then works into the uneaten
center, and frequently the mine crosses like
a figure 8. (See PI. V, fig. 3.) If the leaf is
small, the entire cellular tissue may be con¬
sumed, leaving only the epidermis; in such
cases the larvae have been observed to enter the leaf petioles and burrow
a short way downward in an effort to secure enough food to bring
them to maturity. The larva is not able to enter a fresh leaf in search
of food, but perishes when the food supply in one leaf is insufficient to
bring it to maturity.

The larva is provided with an oral appendage, or rasping organ (fig. 7),
with which it breaks down the cellular tissue and conveys it to the mouth.
This feeding “rake” is swung rapidly from side to side, twice a second or
oftener, while the body moves in an arc as far as can be easily reached,
when it is quickly brought back to the other end of the “swath” and
the body moved up a minute distance to reach new cells. The larva
continues thus feeding incessantly within its mine from the time of
incubation until maturity. Mr. C. N. Ainslie observed that feeding took
place at night as well as by day and that strong transmitted light thrown
upon the larva had no effect upon it. It is indifferent to all external
happenings, and the epidermis of the leaf may be stripped from the back
of the feeding larva without disturbing it, provided the head is not
uncovered. When the leaf epidermis is removed from the head, feeding
ceases, and the larva can not be induced to resume it.

Fig. 7.— Mouth armature of larva of
the serpentine leaf-miner, greatly
enlarged. (Original.)

Oct. 10, 1913

Serpentine Leaf-Miner

The larval period covers from 3 to 12 days; during the summer months
it is passed in 4 or 5 days, the time increasing as the days get cooler.
Many individuals are killed by the autumn frosts while they are yet
partially grown. They will, however, continue feeding under remarkably
low temperature conditions in an effort to survive; Mr. Wildermuth
reared larvae from the time of hatching till they were full grown in
from 10 to 12 days under a mean daily temperature of 46.8° F., and
where upon one occasion a minimum of 250 F. was reached.

PUPATION AND THE PUPAL PERIOD

The pupa (fig. 6) , when found within the leaf, is always at the enlarged
end of the mine where the larva stops feeding, and frequently in a cavity
next to the lower surface, so that there is no indication that the pupa-
rium is present until the leaf is turned over to view it from beneath.
The color is light yellow at first and gradually turns darker as transfor¬
mation progresses, becoming a deep-brown color before the adult emerges.
In the more humid section of the country the fully developed larva
invariably forsakes its mine and descends into the ground from one-fourth
to one-half inch below the surface, or crawls beneath some litter and
there pupates. This is apparently true over the entire country with
respect to the hibernating generation, but in the arid and semiarid regions
of the West it has been observed that during spring and summer much
of the transformation takes place within the larval mines in the leaves.

In the Salt Lake Basin and alfalfa-growing sections of southern Idaho
and Wyoming pupation occurs almost entirely within the larval mines
during the summer months. The junior author, who first studied the
species at Wellington, Kans., and afterwards at Salt Lake City, Utah,
at once noticed this difference in pupation habits in the two localities.
This same thing was noticed at Salt Lake City, Utah, by Mr. C. N.
Ainslie, who was rarely able to find mines from which the larvae had
emerged to pupate.

Mr. Wildermuth found that in the Imperial Valley of California during
the month of April about 50 per cent of the larvae pupate in the mines,
but in the Salt River Valley of Arizona only a small percentage trans¬
forms within the mines, the majority forsaking the leaf and pupating in
the soil. In Indiana, where this insect attacks cabbage, rape, and cow-
peas, this transformation takes place entirely within the soil. This is
also true in the region of the Southeastern States, where the mines are
found in the leaves of cowpeas and, as observed by Mr. McGregor, to
some extent in those of cotton.

No reason can be advanced to explain this difference in habit of pupa¬
tion, a careful study of the humidity in these widely separated localities
failing to offer any explanation therefor.

The pupal period during the summer months is about 10 days, but
ranges from 8 to 28 days from April to December.

Journal of Agricultural Research

Vol. I, No. i

THE ADULT PERIOD

The fly (fig. i , a) emerges through a slit cut in one end of the pupa-
rium and can be taken at almost all hours of the day in sweeping the
foliage with a net. Adults put in confinement have lived io days after
emerging, and the time elapsing between emergence and oviposition
has varied from 4 to 10 days. The eggs are deposited soon after copula¬
tion and in the manner previously described.

LENGTH OF LIFE CYCLE

The following may be taken as the average period elapsing for the
different stages of development during the months of June and July,
at a latitude of 40° :

Days.

Time elapsing between the emergence of the adult and oviposition. . 5

Egg period . 4

Larval period . 4

Pupal period . IO

Average time for one generation . 23

This period is considerably lengthened under existing low temperatures,
and a maximum period of 35 or 40 days may be required in the cool
weather of late autumn.

NUMBER OF GENERATIONS ANNUALLY

Since the larvae continue developing late into the autumn and many of
them are killed by the frosts of winter, the number of generations depends
entirely upon the latitude, altitude, and length of the growing season.
In northern Indiana during the season of 1912 Messrs. Phillips and Lugin-
bill recorded six generations in a series of experiments carried on from
the time the first larvae were found in May until November.

From field observations and generation experiments conducted by the
junior author and Mr. E. J. Vosler at Salt Lake City, Utah, there were
found to be at least five generations from August 1, 1911, to August 1,
1912. The generation experiments in 1912 were started with adults
swept from the fields in May, assumed to have issued from hibernating
pupae. The first generation in the spring is rather well defined and
occupies about one month. As the season progresses, the generations
so overlap that all stages of the insect can be found in the fields at the
same time, and the life cycle was found to be shortened to a minimum
of 18 days.

During the latter half of July and the month of August in the Salt Lake
Basin it was noticed that the injured leaves of alfalfa in open fields were
much more difficult to find than at any other time during the season.
Moreover, alfalfa and white clover found growing in the shade were more
generally infested than those growing in the open field. This was espe¬
cially noted at Laketown, Utah, August 4, 19x1, where a severe infesta-

Oct. 10, 1913

Serpentine Leaf-Miner

tion was noticed on alfalfa plants growing in the bottom of a dry irrigation
ditch where the vertical banks on each side kept the plants well shaded.
At the same time very few mined leaves could be found in the open fields.
There was, however, no interruption to the generation experiments car¬
ried on out of doors and in the shade at Salt Lake City, the adults con¬
tinuing to emerge and larvae to develop during this time.

Mr. Wildermuth, at Tempe, Ariz., during the season of 1912, remarked
the almost total disappearance of all stages during the months of July
and August, followed by their reappearance in September. He recorded
three generations from the last of April to the last week in June and two
more and a partial third generation between September and December
of the same year. At Tempe adults did not emerge from the puparia
in the generation experiments during July and August.

/oo

'so

’LAAM f&B. AfA*. AfAV kASAA T iJiSL.Y' OC7T S/Ott 0£C.

Fig. S. — Diagram showing the range in temperature throughout the year at three widely separated
localities at which observations were made on the serpentine leaf-miner.

In Arizona this disappearance of the insect apparently takes the form
of a period of aestivation during the hot weather of midsummer, when
the temperature in the open fields is too high for the successful propaga¬
tion of the species. This is less noticeable in the cooler alfalfa-growing
valleys farther north, where the summers are milder. Its presence in
Utah alfalfa fields in much reduced numbers during August indicates that
an attempt at aestivation is made there, but over a period of much shorter
duration than is found farther south, in Arizona.

In this connection we here present (fig. 8) curves representing the
normal mean temperatures recorded by the United States Weather
Bureau at Salt Lake City, Utah, and Phoenix, Ariz. As will be seen by
these curves the normal temperature at Phoenix, Ariz., from the first of

Journal of Agricultural Research

Vol. I, No. i

June until early September exceeds the highest mean temperature during
the summer at Salt Lake City, Utah. This may in part explain the
difference in habits of this insect at the two localities during midsummer.

INJURY TO FIELD CROPS OTHER THAN ALFALFA
MINING IN LEAVES OF COWPEA

This leaf -miner has been found burrowing in the leaves of the cowpea
in widely separated localities by several agents of the Bureau of Ento¬
mology.

Dr. E. G. Titus, formerly an agent of the bureau, on July 12, 1904, found
the leaves of the cowpea at Batesburg, S. C., generally attacked by leaf¬
mining larvae, most of which had already escaped from the mines. He
was able to rear two adults of this species and one hymenopterous para¬
site. Messrs. G. G. Ainslie and Philip Luginbill have observed mined
leaves at Columbia, S. C., the former in July, 1908, and the latter in
September, 1912. Mr. Luginbill also reared adults and parasites of this
insect from their mines in cowpea leaves on the plats of the experiment
station at Purdue University, La Fayette, Ind., in connection with studies
made at that point extending from July 6 to August 7, 1911. These
miners were attended by great numbers of parasitic Hymenoptera, Euihri -
chopsis agromyzae Vier.

The junior author observed larval mines in cowpeas at several points
in Mississippi during August and September, 1912, but in every case
the larvae were parasitized or had escaped from the end of the mine
through a slitlike opening and gone into the ground for transformation.

Mr. George G. Ainslie observed considerable injury to the cotyledons
of young cowpeas at Lakeland, Fla., May 8, 1912, there being from 2 to 12
mines in each cotyledon — enough to make the leaves appear sickly and
white. As many as 10 puparia were secured from moderately infested
leaves. The larvae left the mine to pupate.

The injury to cowpeas is seldom severe, because of the larger size of
the leaf, but may become so when the larvae are present in sufficient
numbers in the cotyledon of very young plants before there is sufficient
foliage to withstand their attack.

MINING IN LEAVES OF RAPE

The larvae in large numbers were observed by Mr. W. J. Phillips to be
mining in rape leaves at La Fayette, Ind., on July 6, 1909, and from the
material collected adults of this species emerged July 9. Plate V, figure
1 , shows one of these leaves containing several larval mines. The larvae
were observed to leave the mines and pupate on or beneath the surface of
the soil, and the complete life cycle was found to be passed in from 25
to 28 days.

More extended studies were made of this species as infesting rape at
La Fayette, Ind., during the season of 1912 by Messrs. Phillips and

Oct. io, 1913

Serpentine Leaf-Miner

Luginbill. Mines were also found in leaves of cabbage on May 9. They
were first noticed in the leaves of rape on July 12, about the time the
mines were noticed in this plant by Mr. Phillips three years before. A
series of experiments was carried on from May until November with
cabbage and rape as host plants, and a maximum of six generations was
found to occur in that latitude.

Here again, as is the case wherever these mines are found, a very large
percentage of the larvae in them were found to be parasitized, and a large
number of parasites were reared. Oviposition was observed, both in the
field and in confinement, to take place precisely as in the leaves of alfalfa.
The mines usually start from near the edge of the leaf, where the eggs are
deposited, and extend part way around the leaf on the upper side, being
visible only from above.

The extent of the damage to the crop under observation was not severe
and, perhaps, could be reduced by destroying all the old plants at the end
of the season and plowing deeply in the autumn to bury the hibernating
pupae.

Moreover, since cabbage seems to be a favorite food plant during the
spring, it is readily seen that this crop should not be succeeded by or
planted near rape, where trouble from this leaf -miner is anticipated.

MINING IN nBAVES OF COTTON

While primarily an enemy of forage crops, this miner has been found
feeding in leaves of cotton in the Southern States. In 1906, adults were
collected in cotton fields at Cotulla, Tex., by the late Mr. F. C. Pratt, and
a year later taken in a cotton field by Mr. E. S. Tucker, of the Section of
Southern Field-Crop Insect Investigations, Bureau of Entomology. Dur¬
ing the summer of 1912, adults determined as this species were reared
from cotton leaves at Batesburg, S. C., and Dallas, Tex., by Mr. E. A.
McGregor, of the Section of Southern Field-Crop Insect Investigations,
and by Mr. A. Rutherford.

The mines were observed at Batesburg by Mr. McGregor from the time
of the first appearance of the cotton seedlings until July. Table I, pre¬
pared by him, shows the percentage of infestation which existed on July
12, 1912.

Table I. — Infestation of cotton by the serpentine leaf-miner at Batesburg , C., July 12,

IQ12.

Plants in row.

Plants in¬
fested.

Percentage of
infestation.

107

I56

136

1 344

1 289

1 Total.

2 Average.

Journal of Agricultural Research

Vol. I, No. i

Mr. McGregor’s notes are as follows:

Data have not been accumulated from which to compute the percentage of leaves
affected. It is quite evident, however, that at this season the plants outgrow the
infestation and the rapidly forming leaves tend to reduce the percentage of infested
leaves. This phenomenon easily leads to the erroneous inference that the pest prefers
the seedling leaves and becomes less troublesome as the plants develop. On the
contrary, later on in the season freshly formed leaves appear to be just as desirable to
the leaf-miner as did the seedling leaves. The tortuous courses of the burrows often
sever the main veins of the leaves, causing the death of more or less of the leaf, which
may harbor several individuals.

The habits of the leaf-miner, as observed in cotton leaves by Mr.
McGregor, are here quoted :

The duration of the larval stage, while not fully established, approximates a week.
Feeding takes place and the tunnel is formed in the palisade tissue nearer the upper
surface * * *, as the grub increases in size the caliber of the burrow expands until
full development is attained at its cavernous end, when the larva escapes through a
valvelike incision and pupates in the soil. In the laboratory adults issued six days
after pupation.

Three hymenopterous parasites were reared by Mr. Rutherford from
the pupae of the host.

NATURAL ENEMIES OF THE SERPENTINE LEAF-MINER

Throughout its entire area of distribution this insect is severely para¬
sitized. Excessive parasitism was noted in the earliest studies of the
species about Washington, D. C., and the senior author reared numerous
parasites from the larvae mining in the leaves of white clover at Oxford,
Ind., in 1884. In connection with the studies made during the last
three years there have been reared at least 28 species of hymenopterous
parasites from the mines of this insect in the foliage of alfalfa and other
forage crops in the United States. At times these minute enemies have
become so numerous as to render even a careful study of the pest itself a
matter of some difficulty. But for their presence these leaf -miners would
beyond a doubt work much more serious ravages in the alfalfa fields of
the West than they do at present. Indeed, one is inclined to wonder
what the actual financial effects would be were some condition to arise
suddenly whereby the numbers and efficiency of these natural checks
were radically diminished.

The first generation of the leaf-miner to appear in the spring is not
severely parasitized, and from larvae and puparia collected at this time
numerous flies usually emerge. The following generation is more severely
parasitized, and thereafter the parasites increase rapidly, infestation becom¬
ing more and more severe, so that mined alfalfa leaves collected during
the summer and fall will usually yield parasites instead of adult leaf-
miners. To illustrate this point, the junior author, near Salt Lake City,
Utah, on September 16, 1911, selected in the field 45 mined alfalfa leaves,
43 of which contained 1 mine each, while 2 had 2 mines. Of the 47 mines.

Oct. io, 1913

Serpentine Leaf-Miner

3 contained healthy larvae and 2 healthy pupae of Agromyza, while the
remaining 42 mines, or 89.7 per cent of those examined, contained para¬
sites. Of these 42 mines, 25 contained parasitized larvae, 14 parasitized
pupae, and 3 were doubtful. Of the 25 parasitized larvae, 20 carried 1,
and 5 carried 2 external parasites, making 30 parasites on the 25 larvae
of the leaf -miner; these, with the 14 parasitized pupae, make a total of
44 individual parasites within the 45 mined leaves. In the Salt Lake
Basin from June to October, 1911, 75 to 90 per cent of the mines in
alfalfa leaves were found to be parasitized.

At Sacaton, Ariz., as early as May 25, 1912, Mr. R. N. Wilson, of the
Bureau of Entomology, found 89 per cent of the insects issuing from
mines of Agromyza pusilla to be parasites, while from material collected
there in June and July parasites alone emerged.

Mr. Wildermuth, at Tempe, Ariz., from experiments conducted during
the season of 1912, found that much the same degrees of parasitism
existed in that locality ; and while no record was kept to show the number
of parasites found in occupied mines, Table II shows the number of adults
and parasites which issued from large numbers of leaves containing Agro¬
myza larvae, collected in the field and kept in jars in the laboratory.

Table II. — Emergence of Agromyza pusilla and its larval parasites in Arizona and

California in IQI2 .

Date leaves
were
collected.

Locality.

Experiment

No.

Number of
Agromyza
issued.

Number of
parasites
issued.

Percentage
of parasites to
total insects
issuing.

May

Tempe, Ariz .

May

. do . . .

May

. do . . .

May

. do . . . .

IOO

May

. do .

June

. do .

IOO

Sept.

. do .

Oct.

. do . . . .

. do . . .

1 3

Oct.

. do . . .

Oct.

. do . .

Oct.

. do .

Do .

. . . do .

QEJ

. . do .

Nov.

. do .

Apr.

Total .

HI Centro, Cal .

478

18 .

86. 2

Apr.

Brawley , Cal .

/ J

7 c

Apr.

Bard, Cal .

As will be noted in Table II, the high percentage of parasitism falls
off rapidly upon the approach of cool weather, thus enabling the insect
to enter hibernation with a much reduced degree of parasitism. At
Lakeland, Ela., where no hibernation occurs, Mr. G. G. Ainslie records

Journal of Agricultural Research

Vol. I, No. i

no parasites present during January, 1913, among the larvae feeding in
cowpeas. From this fact it naturally follows that the season of greatest
injury to forage crops from leaf-miners will be during a period of pro¬
longed cool weather, when the temperature will naturally be unfavorable
to the rapid multiplication of the parasites. This is precisely the con¬
dition that exists where there are destructive outbreaks of the green
bug ( Toxoptera graminum Rond.) as then the native parasites are unable
to keep the pest in check. Of the life history of most of the parasites
reared in connection with this leaf -miner comparatively little is known.

Diaulinus begin! Ashm. — The parasite most thoroughly studied, as
well as the most abundant, widely distributed, and hence most important in

the control of the host
is a small chalcidoid,
Diaulinus begini
Ashm. (fig. 9) ; the larva
of which feeds exter¬
nally upon the body of
the Agromyza larva.
This parasite has been
reared from mines in
leaves of alfalfa, clover,
cowpeas, and rape in
Indiana, Kansas, Ari¬
zona, New Mexico,

Fig. 9. — Diaulinus begini , a parasite of the serpentine leaf-miner. At California, Utah, Wy-
left, hind leg of Diaulinus websteri. Greatly enlarged. (Original.) • i xi i i

ommg, and Idaho by

different members of the Bureau of Entomology and from mines of
Agromyza parvicornis in corn leaves at Salt Lake City, Utah.

The junior author was able to observe all stages of its development at
Salt Lake City. The female parasite wanders about over the leaf until
she locates the Agromyza larva in its mine below; then, pushing the ovi¬
positor through the membranous tissue of the leaf which constitutes the
roof of the mine, she places the egg upon the body of the host larva.
The egg, as observed upon the surface of the host larva, is smooth,
translucent, oblong, but rounded about equally at each end, and is about
0.5 mm. in length. The egg period is short, probably not lasting more
than one or two days. The young larva feeds externally upon the body
of its host, which dies while the parasitic larva is yet very young. Often
the presence of the parasitic larva can not be detected on the body of
the host without the aid of a microscope. The host larva is invariably
dead whenever one of these larvae, even though apparently just hatched,
can be found on its body. Occasionally two larvae feed on the body of
a single host larva, and in one case both parasitic larvae were observed to
complete their transformations and emerge. The larval period is seven
days. Figure 10 shows the full-grown larva. Pupation takes place

Oct. io, 1913

Serpentine Leaf-Miner

within the mines of the host and usually some distance away from the
remains of its victim. Figure 11 represents the pupa of this species.
The pupal period is seven or eight days, and thus the life cycle of the
parasite is considerably less than that of the leaf -miner.

Diaulinus web steri Cwfd.* 1 — Diaulinus websieri (fig. 9, a) is very closely
related to D . begini and, like the latter, it feeds externally upon the larva
of its host. In the life-history studies
made by the junior author at Salt Lake
City, its habits were in no way distinguish¬
able from those of Diaulinus begini , the
two species being reared together from
larvae found attached to the same host. Fig- IO-— Larva of Dmuiinm begini.

Greatly enlarged. (Original.)

Diaulinus webs ten has been reared from

Agromyzafrom Kansas, Utah, Arizona, and California, being the most
abundant parasite reared in southern California and Arizona. Of the
two species of Diaulinus reared by Mr. Wildermuth at Tempe, Ariz., this
species constituted 66 per cent of the material, while D. begini comprised
34 per cent. Of the Diaulinus reared at Salt Lake City D. websteri com¬
prised only 18 per cent, while 82 per cent were D. begini .

This species was reared from mines of Agromyza pusilla in hedge mus¬
tard at Wellington, Kans., in 1912, by Mr. E. O. G. Kelly. Mr. C. N.
Ainslie reared it from mines of Cerodontha dorsalis Loew in timothy
leaves at Ely, Nev. It is also an enemy of Agromyza parvicornis Loew.

Chrysocharis ainsiiei Cwfd* and C. parksi Cwfd * — These parasites
(fig. 12) are very important in the control of Agromyza pusilla
in the West. They feed internally and emerge from the
puparia of the host. Their life history is imperfectly known.
From hibernation material collected at Salt Lake City during
the winter of 1911-12 adults emerged from April 18 to 20,
which was 34 days before Agromyza pusilla was captured in
the fields.

From studies made by the junior author at Salt Lake City,
Utah, in 1 91 1 it was noticed that larvae of Agromyza collected
in the field, which pupated under observation in the labora¬
tory, would often yield adults of Chrysocharis exclusively in¬
stead of those of Agromyza. Only one parasite issues from
each puparium of the host, and dissections made of the
puparia often revealed this to be entirely occupied by the
larva or pupa of the single parasite, which had entirely con¬
sumed its host. But in some instances the puparium of Agromyza when
dissected revealed two embryo parasitic larvae within the body of the
host larva. As only one adult is known to emerge from each puparium
of the host, it is highly probable that when two internal parasitic larvae

1 The species of parasites marked with asterisks have been recently described in the Proceedings of the
United States National Museum, v. 43, p. 163-188 (1912) by Mr. J. C. Crawford, Associate Curator, Division
of Insects.

Fig. 11. — Pu¬
pa of Diau~
linus begini.
Greatly en-
larged.
(Original.)

7954°— 13 - 6

Journal of Agricultural Research

Vol.I.No.i

start to develop in one host, one kills and consumes the other. During
September in the Salt Lake Basin 88 per cent of the puparia collected
in the mined leaves of alfalfa yielded adults of Chrysocharis, and the
two species were about equally represented. Both species have been
collected in northern and central Utah, southern Idaho, the Imperial
Valley of California, and in southern Arizona. C. parksi has also been
reared from mined alfalfa leaves collected at Redding, Cal., in the Sacra¬
mento Valley. It was also reared from Agromyza mines in leaves of
nasturtium and narrow-leaved plantain at Salt Lake City.

Derostenus arizonensis Cwfd. — This parasite of the larva of Agro¬
myza constitutes a new species and is apparently confined to the South¬

west. It was reared
in large numbers by
Mr. Wildermuth from
mined alfalfa leaves
collected in the Salt
River Valley in Ari¬
zona, where it com¬
prised 36 per cent of
the larval parasites so
reared. Three speci¬
mens were reared by
Mr. Urbahns from
mined alfalfa leaves
collected at El Cen¬
tro, Cal.

Fig. 12. — Chrysocharis parksi , a parasite of the serpentine leaf-miner.
a. Middle and hind legs of Chrysocharis ainsliei. Greatly enlarged.
(Original.)

A single specimen

obtained from the large number of parasites reared at Salt Lake City,
Utah, was reared from an Agromyza larva in a leaf of fenugreek ( Trigo -
nella foenum-graecum). It was described by Mr. J. C. Crawford in the
Proceedings of the United States National Museum, volume 45, page
3i5» I9I3*

Derostenus diastatae How. — This species has been reared from
mines of Agromyza pusilla in cowpeas at La Fayette, Ind., by Mr. Philip
Luginbill. In the Eastern States it is an important parasite of Agromyza
parvicornis and A. angulata. It has not been recorded west of Kansas.

Derostenus punctiventris Cwfd.* — This insect was reared from puparia
of Agromyza in mines in leaves of alfalfa at Salt Lake City, by Mr. C. N.
Ainslie, and by the junior author, from alfalfa and white clover at Salt
Lake City, Utah, and Lyman, Wyo. It was reared only occasionally and
is of minor importance as an enemy of this leaf -miner. It also attacks
Agromyza parvicornis.

Derostenus pictipes Cwfd * — This parasite was reared from mines of
Agromyza pusilla in cowpeas at Columbia, S. C., by Mr. G. G. Ainslie in
1908 and at La Favette, Ind., by Mr. Philip Luginbill in 1911. It was

Oct. 10, 1913

Serpentine Leaf-Miner

also reared by Mr. C. N. Ainslie from mines of A. coquilletti Malloch in
leaves of Hordeum jubatum collected at Fort Collins, Colo.

Derostenus varipes Cwfd. — A single specimen of this parasite was
reared from Agromyza pusilla at Iya Fayette, Ind., by Mr. Luginbill.
Nothing is known of its life his¬
tory. It is a new species and
was described by Mr. Crawford
in the Proceedings of the
United States National Mu¬
seum, volume 45, page 315,

1913-

Diaulinopsis callichroma

Cwfd.* — This species was
reared from mines in leaves of
cowpea at La Fayette, Ind., by
Mr. Luginbill and from alfalfa
leaves at Tempe, Ariz., by Mr.

Wildermuth Verv few speci- ^flG* I3* — Zagrammosoma multilineata, a parasite of the ser-

, pentine leaf-miner. Greatly enlarged. (Original.)

mens were secured, and it

seems of little importance as a parasite of Agromyza pusilla.

Cirrospilus flavoviridis Cwfd. — Two specimens were reared from
mines in alfalfa leaves at Salt Lake City, Utah, by Mr. C. N. Ainslie, who
also reared it from mines of Cerodontha dorsalis Loew in timothy leaves
at Ely, Nev. It is also recorded as a parasite of Agromyza parvicornis.

It was described by
Mr. Crawford in the
Proceedings of the
United States Na-
tionalMuseum , volume
45, page 317, 1913.

Zagr ammo soma
multilineata Ashm. —
This species (fig. 13),
described in 1888, has
long been known as a
parasite of a lepidop-
terous leaf -miner (. Lith -
ocolletis sp . ) , from
which it was reared
by the senior author in Ohio in 1893. Only three specimens were reared
from Agromyza pusilla , two being reared at Wellington, Kans., by the
junior author in 1910 and one by Mr. Luginbill at La Fayette, Ind.

Closterocerus utahensis Cwfd.* — A few specimens of this parasite
were reared from mined alfalfa leaves at Salt Lake City, Utah, by Mr.

Fig. 14. — Pleuroiropis rugosithorax, a parasite of the serpentine leaf-
miner. Greatly enlarged. (Original.)

Journal of Agricultural Research

Vol. I, No. i

C. N. Ainslie and at Tempe, Ariz., by Mr. Wildermuth. Nothing is known
of its life history. It is also recorded as a parasite of A gromyza parvicornis .

Pleurotropis rugosithorax Cwfd.* — This species (fig. 14) was reared
sparingly from a puparium of A gromyza pusilla by both Mr. C. N. Ainslie
and the junior author at Salt Lake City, Utah. It is an internal parasite,
having been reared from the immature stages dissected from the puparia
of the host. Only one parasite issues from each puparium of Agromyza.

Eucoila hunteri Cwfd. — This species was not previously known. Two
specimens have been reared from puparia of Agromyza pusilla by Mr.
A. Rutherford at Dallas, Tex. These issued 16 and 17 days, respectively,
after the pupation of the host.

Sympiesis sp. ( ?) — One specimen of this species was reared by Mr.
Kelly from mines in alfalfa leaves at Wellington, Kans., in 1912. It was
also reared from mines in corn leaves at the same locality by the junior
author in 1909. This is probably a new species and is not confined to
one host.

MISCELLANEOUS UNDETERMINED PARASITES

The following miscellaneous Hymenoptera belonging to the super¬
family Chalcidoidea 1 were reared from mines of Agromyza pusilla , the
species being yet undetermined and their life history unknown.

Pteromalus sp. — (a) One specimen bearing Webster No. 6639 and
reared from mines in alfalfa leaves at Salt Lake City, Utah.

(1 b ) Three specimens bearing Webster No. 7492 and reared at the
foregoing locality from mined leaves of white clover.

Cirrospilus sp. — One specimen reared from mines in alfalfa at Tempe,
Ariz., and bearing Webster No. 7215.

Diaulinopsis sp. — Two specimens reared from mines in leaves of cowpea
and bearing Webster No. 6395.

Entedoninae. — One specimen from mined alfalfa leaves reared at Salt
Lake City, Utah, and bearing Webster No. 6639.

BRACONID PARASITES

The following species of parasites belonging to the family Braconidae
were reared from Agromyza pusilla in accordance with the data given
below.2

Opius agromyzae Vier. — La Fayette, Ind. (W. J. Phillips), Nos. 5170
and 6395.

Opius aridus Gahan. — Tempe, Ariz., May, 1912 (V. L. Wildermuth),
No. 7215.

Opius brunneipes Gahan. — Lakeland, Fla. (G. G. Ainslie), No. 9489.

Opius suturalis Gahan. — Tempe, Ariz., May, 1912 (V. L. Wilder¬
muth), No. 7215.

1 Specimens determined to genus or subfamily by Mr. J. C, Crawford.
3 The determinations are by Mr. A. B. Gahan.

Oct. io, 1913

Serpentine Leaf-Miner

PREDACEOUS ENEMIES OP THE SERPENTINE LEAF-MINER

Very few predaceous species are known to feed upon the serpentine
leaf-miner. This is largely due to the fact that the larvae feed well con¬
cealed within the leaf tissue and are thus not open prey. The following
predatory insects are known to feed on some stage of the leaf-miner:

Triphleps sp. — These adults are recorded by Mr. E. G. Smyth, recently
of the Bureau of Entomology, at Tempe, Ariz., to pierce with their
beaks the Agromyza larvae in their burrows.

Erythraeus sp. — These red mites are recorded by Mr. Wildermuth at
Tempe, Ariz., to attack and kill the Agromyza larvae in their tunnels.
Mr. Nathan Banks determines this as probably a new species.

REMEDIAL AND PREVENTIVE MEASURES

The excessive parasitism under which this species exists has so far pre¬
vented it from becoming destructively abundant or doing any widespread
serious injury. In case through any cause it should become more injuri¬
ous to alfalfa, doubtless cutting the crop for hay at once as soon as the
depredations were observed would prevent a recurrence. Its greater
abundance along ditches, roadsides, and other neglected places indicates
that frequent cutting of the alfalfa acts as a permanent check upon the
increase of the insect. East of the arid regions deep fall plowing would
bury the pupae so deep in the ground as to put them beyond the possibility
of emerging as adults. This is especially recommended for the annuals,
such as cowpeas and rape. Throughout the remaining western country
keeping down volunteer growth along ditch banks and in waste lands
would greatly diminish the number of pupae which yearly enter hiberna¬
tion. Of course, pasturing either clover or alfalfa would destroy all
larvae mining in the leaves eaten off by the grazing.

OTHER SPECIES OF THE GENUS AGROMYZA LIKELY TO BE MIS¬
TAKEN FOR THE SERPENTINE LEAF-MINER

The species of Agromyza are for the most part very similar to one
another in appearance. As a consequence there has been much con¬
fusion in their proper classification, and as a further result of this con¬
fusion articles have been published relating to one species which in the
light of our present knowledge clearly belong to another. It is with the
hope of preventing further errors of this nature that the following species
of Agromyza — the first of which has in the last year or two been confused
with the serpentine leaf-miner — are briefly treated in this paper:

Agromyza angulata Loew. — This leaf-miner (fig. 15) attacks leaves of timothy,
mining between the membranes in the same manner as the serpentine leaf-miner.

It was reared from puparia (fig. 16) in leaves of timothy found July 4, 1895, near
Bladensburg Road, D. C,, by Mr. Theo. Pergande.

Journal of Agricultural Research

Vol.I.No. I

During July, 1912, Mr. Philip Luginbill at La Fayette, Ind., reared these adults
from mines in leaves of volunteer timothy growing in protected places and was able
to secure all stages of the insect.

The eggs are deposited in the cellular tissue just above the epidermis on the ventral

side of the leaf, and in
punctures similar to
those made by Agromyza
pusilla and A. parvicor -
nis.

The egg stage is four to
five days.

The larvae feed in one
leaf until mature and
pupate in the mine . The
larval period is 8 to 10
days, the pupal period,
13 days. This makes a
total of 27 days elapsing

Fig. 15, — Agromyza angulata. Greatly enlarged, (Original.) from egg to adult.

Mr. Luginbill and Mr.

Phillips were also able to transfer these miners from timothy to wheat, rearing one
generation from wheat, using as parents flies reared from timothy mines.

The number of generations is not known. The following species of parasites were
reared by Mr. Luginbill in connection with his studies in Indiana:

Polycystus foersteri Cwfd.; Derostenus diastatae How.; Derostenus agromyzae Cwfd.;
Pleurotropis rugosithorax Cwfd.; Entedon thomsoni Cwfd.; N otanisomorpha ainsliei
Cwfd.

A single specimen was collected at Plummers Island, Md., July 28, 1912, by Mr. H.
L. Viereck, and specimens collected at Niagara Falls, N. Y., and Aubumdale, Mass.,

are present in the private collection of Mr. C. W. John¬
son, curator of the Boston Society of Natural History.
The species has never become sufficiently abundant to
attract attention.

Agromyza coquilletti Malloch. — This species (fig. 17)
was reared from a puparium found among the basal
leaves of volunteer wheat at Bucklin, Kans., November
6, 1909, by Mr. C. N. Ainslie. It was also reared at
Fort Collins, Colo., by Mr. Ainslie from a larva mining
a leaf of oats, June 30, 1910.

From three larvae mining leaves of Hordeum jubaium 1
in the same locality on July 16, 1910, one adult of this
species and seven hymenopterous parasites were reared.
These were determined by Mr. J. C. Crawford as Dero¬
stenus pictipes Cwfd.

Larvae were observed mining leaves of wheat at
Roosevelt, Utah, June 25, 1912, by Mr. C. N. Ainslie,
but from this material only parasites of the genus Pteromalus issued.

One specimen was reared from a blade of wheat at La Fayette, Ind., July 2, 1912,
by Mr. Philip Luginbill, and the junior author reared one adult of this species from a
larva mining a leaf of oats taken at Shoshone, Idaho, July 17, 1912.

Fig. 16.— Puparium of Agromyza
angulata , with lateral view of
anal appendages at left. Greatly
enlarged. (Original.)

1 In this connection we note that Mr. Ainslie reared from various-shaped mines in Hordeum collected at
Myton, Utah, June 27, 1912, two flies determined by Mr. Walton, of the Bureau of Entomology, as Hydrellia
scapularis Eoew. So far as can be ascertained, this is the first instance of the rearing of this species and the
first report that it affects vegetation.

Oct. io, 1913

Serpentine Leaf-Miner

Three specimens have been swept from growing wheat at Manhattan, Kans., by
Mr. C. N. Ainslie and one specimen from wheat at Lincoln, Nebr., by Mr. Geo. I.
Reeves, of the Bureau of Entomology.

The following localities are represented in the collection of Mr. C. W. Johnson:
Twin Rock, Pa. (Johnson); Nantucket, Mass. (J. A. Cushman); Norwich, Vt. (John¬
son); Hanover, N. H. (Johnson).

The species has never become a serious enemy of wheat or oats.

Agromyza virens Loew. — This species was reared from larvae taken in root stems of
white clover at La Fayette, Ind. , by the senior author in August, 1886. The maggots
were found singly in
the stem, sometimes
just under the epider¬
mis, and sometimes in
the center. In either
case parallel channels
were excavated, the
larvae working from the
point where the stem
originated. These flies
were determined ten¬
tatively as Oscinis sp.,
and a report1 of the
rearing describing the
larva and pupa was
published at that time. On October 19, 1898 , these flies were reared from larvae taken
in the pith of the garden sunflower ( Helianthus annuus) at Wooster, Ohio.

Mr. Theo. Pergande reared adults of this species from stems of Mulgedium acumina¬
tum collected by the senior author at La Fayette, Ind., in November, 1885. Several
undetermined hymenopterous parasites were reared from this material. These bear
No. 3640. Mr. Pergande also reared one adult miner on April 18, 1883, from stems of a
weed collected by Mr. Albert Koebele at Holdemess, N. H., in October, 1882, and
containing at that time mostly pupae. He also reared an adult from a stem of Ambrosia
artemisiaefolia (ragweed) received January 6, 1890, from A. M. Sharp at Gladbrook,
Iowa.

It has also been reared from heads of Rudbeckia sp. at Dallas, Tex.

There are in the collection of the United States National Museum two specimens
from Cambridge, Mass., marked "mining in stems of weed'’ (H. G. Hubbard); two
“from stems of Ambrosia,” March, 1895, District of Columbia; one "from Nabalus
albus May 14, 1883; two from California (Alameda and Los Angeles) collected by
Mr. Coquillett; one from Flagstaff, Ariz. (H. S. Barber); thirteen from Toronto,
Canada (William Brodie); one from Plummers Island, Md., and four from Washington,
D. C., collected by Mr. W. L. McAtee.

Agromyza melampyga Loew, var. marginalia Malloch. — Three adults were reared
from larvae mining in leaves of grass (Paspalum dilatatum) by Mr. Philip Luginbill at
Columbia, S. C., October 4, 1912.

Fig. 17.— Agromyza coquilletti. Greatly enlarged. (Original.)

SUMMARY

The serpentine leaf -miner is the larva of a minute yellow and black
fly which is common in alfalfaTields during the summer.

It is generally distributed over the United States, having a wide range
of food plants.

1 Riley, C. V. The clover-stem maggot (.Oscinis sp.). U. S. Comr. Agr, Rpt. 1886, p. 582, 1887.

Journal of Agricultural Research

Vol. I, No. i

The larvae injure the foliage of the plant by burrowing between the
membranes of the leaf and devouring the parenchyma.

The injury takes the form of a serpentine “mine” which encircles the
leaf, gradually widening as the larva increases in size.

Leaves of white clover and frequently of young alfalfa often have the
entire cellular tissue devoured, leaving only the two membranes.

There is usually only one larva present in each leaf.

The injury from this insect is greatest in the Southwest, where the dis¬
colored leaves, which in severe cases become brown, are sometimes present
in sufficient numbers to lower the quality and grade of the hay.

The injured leaves can be found in the fields from May until Novem¬
ber, the larvae continuing to feed until killed by frosts. In Florida the
larvae continue feeding throughout the winter.

The insect hibernates in the puparia beneath the surface of the soil at
the base of the plants.

There are five or six generations in latitude 41 °, the number varying
with the length of the growing season.

The generations overlap to such an extent that all stages can be found
in the fields during most of the season.

During the period of highest temperature in summer the larvae are
found usually infesting plants protected from the direct rays of the sun.
During this period in the arid Southwest the insect almost completely
disappears from the fields, reappearing in September.

The eggs are deposited in the leaf tissue and inserted in punctures
identical with those made by the adult in feeding. The egg stage during
June is 4 da}^s.

The larvae feed continuously day and night and confine their work to
a single leaf. The larval period during June is 4 days.

In the Eastern States pupation occurs entirely in the soil. It takes
place commonly in the larval chambers in the leaf in the arid Western
States. The pupal period during June is 10 days.

The average period of the complete life cycle is 23 days.

Besides alfalfa the following field crops are subject to attack: Clover,
cowpeas, rape, and cotton.

A few nearly related and very similar leaf-miners are known to attack
timothy, wheat, oats, and grasses. When these crops are affected, the
mine usually extends the entire width of the leaf, and may kill the plant
if it is very young.

Numerous parasitic insects attack and consume the larvae and pupae
within their mines. These are highly efficient and serve to keep the
insect in control.

The efficiency of the parasites decreases upon the approach of cool
weather.

Many of these parasites are functional in the control of more than one
species of leaf-miner, and are very widely distributed.

Oct. io, 1913

Serpentine Leaf-Miner

Frequent cutting of alfalfa kills the larvae in the leaves and does much
to protect this crop. This method should be followed where the injury
becomes serious.

Deep fall or winter plowing is advocated for annual forage crops and
cereals in order to bury deeply the hibernating puparia located near the
surface of the ground.

BIBLIOGRAPHY

Bouch£, P. F. Beitrage zur Kenntniss der Insekten-Larven. Stettin. Bnt. Ztg.,
Jahrg. 8, No. 5, p. 142-146, Mai, 1847.

** Agromyza amoena Meig.,” p. 142.

Brauer, Friedrich. Die Zweiflugler des Kaiserlichen Museums zu Wien. III.
Wien, 1883.

“Agromyza” p. 91-92.

Brischke, C. G. A. Die Blattminirer in Danzig’s Umgebung Schr. Naturf. Cesell.
Danzig, n. F., Bd. 5, p. 233-290, 1881.

u Agromyza trifolii Burgess,” p. 247; “Agromyza pusilla Meig.,” p. 249, 270, 273, 274, 28S; “Agromyza
strigata Meig.,” p. 260, 266.

Chittenden, F. H. The native clover leaf-miner ( Agromyza diminuta Walk.).
U. S. Dept. Agr., Bur. Ent., Bui., n. s., no. 33, p. 77, 1902.

“Agromyza diminuta Walk.”

Comstock, J. H. The clover Oscinis (Oscinis trifolii , Burgess [n. sp.]). U. S. Comr.
Agr. Rpt. 1879, p. 200-201, 1880.

“Agromyza ( Oscinis ) trifolii Burgess.”

COQUiUvETT, D. W. On the habits of the Oscinidse and Agromyzidae reared at the
United States Department of Agriculture. U. S. Dept. Agr., Bur. Ent., Bui.,
n. s., no. 10, p. 70-79, 189S.

“Agromyza diminuta Walk.,” p. 78.

Kaltenbach, J. H. Die Pflanzenfeinde aus der IClasse der Insekten. Stuttgart, 1874.

“Agromyza orbona Meig.,” p. riS; '' Agromyza strigata Meig.,” p. 408; “ Agromyza amoena Meig.,”
p. 298-299; “Agromyza trifolii Burgess,” p. 129.

Lintner, J. A. The insects of the clover plant [read Jan. 19, 1881]. Trans. N. Y.
State Agr. Soc., v. 33, 1877-1882, p. 1S7-207, 6 fig., 1884.

“ Agromyza ( Oscinis ) trifolii Burgess,” p. 205-206.

Riley, C. V. The cabbage Oscinis. ( Oscinis brassicce , n. sp.). U. S. Comr. Agr.
Rpt., 1884, p. 322, pi. 8, fig. $.

“Agromyza brassica Riley.”

SciiinER, J. R. Fauna Austriaca. Die Fliegen (Diptera). T. 2, Wien, 1864.

“Agromyza pusilla Meig.,” p. 301.

DESCRIPTION OF PLATE

Plats V. Leaves of different species, showing the work of the serpentine leaf-miner *
(Agromyza pusilla ). Fig. i. — Mines in a leaf of rape. Fig. 2. — Mines in leaves
of white clover. Fig. 3. — Mines in leaves of alfalfa. (All nearly natural size.
Original.)

(88)

ADDITIONAL COPIES of this publication
iL may be procured from the Superintend¬
ent of Documents, Government Printing
Office, Washington, D. C., at 15 cents per copy

Plate V

JOURNAL OF AGRICULTURAL RESEARCH

DEPARTMENT OF AGRICULTURE

Voe. I Washington, D. C., November io, 1913 No. 2

THE OCCURRENCE OF A COTTON BOLT WEEVIL IN

ARIZONA

By W. Dwight Pierce,

Agent and Expert , Investigations of Insects Affecting Southern Field Crops ,
Bureau of Entomology

The preliminary announcement by Mr. O. F. Cook, of the Bureau of
Plant Industry, in February, 1913, of the occurrence in Arizona of a
weevil resembling the Mexican cotton boll weevil, appears at this time
to have been an announcement of considerable importance. In com¬
pany with Mr. Harold Bell Wright, Mr. Cook found this weevil breeding
in the bolls of a wild shrub known as Thurberia thespesioides in Ventana
Canyon, Santa Catalina Mountains, Arizona.

In May the writer obtained a large quantity of bolls of Thurberia from
Mr. W. B. McCleary, of the Bureau of Plant Industry, who collected
them in the lower part of Stone Cabin Canyon, Santa Rita Mountains,
Arizona. This material was very heavily infested by the weevil.

During August Dr. A. W. Morrill, State Entomologist of Arizona,
together with the writer, located this weevil in Ventana Canyon, Santa
Catalina Mountains, and in Sawmill Canyon, Santa Rita Mountains,
breeding commonly upon the same plant.

A close examination of the material received early in the year disclosed
many minor points of difference from the usual form of the cotton boll
weevil, Anthonomus grandis Boheman. The Arizona form averages
slightly larger and is a little more robust. The punctation of the male
beak is a little more pronounced, and the sculpturing throughout is
slightly stronger than in the Texas form. The scaly vestiture approaches
a golden color, while in the Texas form it is usually grayish. The sides
of the prothorax in front are rarely emarginate, while the emargination
is usually very noticeable in the Texas form. Minor differences also
appear in the shape of the teeth on the legs. All in all, the adults of the
Arizona weevil present an assemblage of characters differing from the
eastern form sufficient to suggest a new species.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Vol. I, No. 2
Nov. 10, 1913

Journal of Agricultural Research

Vol. I, No. a

In addition to these differences in characters, specimens of the Arizona
form were found in hibernation in their cells until September i, while
the eastern form is never found in its cells in cotton bolls after March 15.
The Arizona insect seems to be confined to one, or not more than two,
annual generations, while the cotton boll weevil has many generations.
The former lives on Thurberia, the latter on Gossypium. The Arizona
weevil was found at 4,000 feet altitude, while the Texas weevil has never
been found above 2,000 feet altitude. The two forms are geographi¬
cally isolated by mountain divides. When the Arizona weevil was seen
in the field, it displayed a tendency to oviposit at a different place and
to seal its egg puncture differently; the egg itself was of a slightly
different shape.

The Mexican cotton boll weevil has never been known before this year
to feed readily or breed in any other plant, although suspected of being
capable of adapting itself to other foods if forced to it. When oppor¬
tunity was given the Texas boll weevil to attack Thurberia squares and
bolls, it fed readily and eagerly, sometimes displaying a preference for
Thurberia over cotton when both were available. The Thurberia-feeding
weevil, on the other hand, was able to feed upon and breed in cotton
squares.

Mr. B. R. Coad, of the Bureau of Entomology, has succeeded in rear¬
ing undoubted crosses between the two varieties from females of each
form, although these hybrid offspring were somewhat undersized.

It will be seen from further evidence in this paper that the two forms
must represent merely two subspecies, or varieties, or geographic races of
a single species. The Arizona form is therefore to be known as Anthono-
mus grandis thurberiae, new variety. Its technical description is as fol¬
lows :

Anthonomus grandis thurberiae, n. var. — Stout, subovate, rufo-piceous, and
clothed with coarse, pale-yellowish pubescence. Beak long, slender, shining, and
sparsely pubescent at the base; striate from base to the middle, striae rather coarsely
punctured; apical half finely and remotely punctured. Antennae slender, second
joint of funicle longer than the third; joints 3 to 7 equal in length but becoming gradu¬
ally wider. Head conical, pubescent, coarsely but remotely punctured, front foveate.
Eyes moderately convex, posterior margin not free. Prothorax one-half wider than
long; base feebly bisinuate, posterior angles rectangular; sides almost straight from
base to middle, strongly rounded in front; apex slightly constricted and transversely
impressed behind the anterior margin; surface moderately convex, densely and sub-
confluently punctured; punctures irregular in size, coarser about the sides; pubescence
more dense along the median line and on the sides. Elytra oblong, scarcely wider
at the base than the prothorax; sides robust to subparallel for two-thirds of their length,
thence gradually narrowed to and separately rounded at the apex, leaving the
pygidium moderately exposed; striae deep, punctures large and approximate; inter¬
stices convex, rugulose, pubescence somewhat condensed in spots. Legs rather stout,
femora clavate, anterior strongly bidentate, inner tooth long and strong, outer one
acutely triangular and connected with the former at the base; middle femora with
small second tooth and posterior femora unidentate. Tibiae moderately stout, anterior
bisinuate internally, posterior straight; tarsi moderate, claws broad, blackish, and

Nov. io, 1913

A Cotton Boll Weevil in Arizona

rather widely separate; tooth almost as long as claw. Length, 5 to 5.5 mm. (0.20 to
0.22 inch).

This variety differs from Anthonomus grandis on cotton by its greater robustness
(PI. VI); the more golden appearance of the scales; the slighter constriction of the
prothorax (figs. 1 and 2); its stouter and more coarsely sculptured beak (figs. 3 and 4);
its slightly more compact antennae (figs. 5 and 6), with funicle of a lighter color than
the club; its stouter legs, with a distinct second tooth on the middle femora (figs.
7 and 8) ; the wing (fig. 9) , which shows a slightly more distinct spot. It also differs in its
food plant ( Thurberia thespesioides), its altitude (4,000 feet upward), its breeding
season (August 15 to November), and in certain physiological and biological char¬
acters. The most obvious diagnostic characters are as follows:

Anthonomus grandis thurberiae

Antennal funicle of a distinctly lighter color than
the club; punctation of elytral striae strongly and
clearly defined; prothorax usually very feebly con¬
stricted and not emarginate or but very slightly so;
elytra often robust; vestiture of ochreous scales
intermixed with black hairs; breeds in Thurberia
thespesioides; range, above altitude of 4,000 feet.

Fig. r. — Anthonomus grandis , var. thur¬
beriae: Prothorax. Much enlarged.
(Original.)

Anthonomus grandis

Antennal funicle and club concolorous; puncta¬
tion of elytral striae not clearly defined from the
striae; prothorax strongly constricted at apex and
usually emarginate in front; sides of elytra usually
parallel; vestiture of grayish to ochreous scales in¬
termixed with very inconspicuous grayish to very
dark-brown hairs; breeds in Gossypium spp.; range,
below altitude of 2,000 feet.

Fig. 2. — Anthonomus grandis Boh.: Pro¬
thorax. Much enlarged. (Original.)

Hibernation. — It is not known whether Anthonomus grandis thurberiae hibernates
as an adult outside of its cell, but it is known positively that many individuals pass
the winter and even the summer in the cells formed during the preceding fall. In
May, 1913, from the material sent by Mr. McCleary, the 'writer found 18 live adults in
their cells in an examination of 743 bolls, 220 of which were infested. On August 27
Dr. Morrill found six live boll weevils still in their last year’s cells at about 4,500 feet
altitude in Sawmill Canyon, Santa Rita Mountains, and on August 30 the writer
found another live weevil in its cell in Ventana Canyon, Santa Catalina Mountains.

As further evidence of the prolonged rest of this variety, no immature stages were
found, beyond a one-fifth grown larva in squares. The extreme lateness of the plants
in the canyons where the boll weevil was found indicated that the weevils could not
have had buds on which to feed for much more than two weeks in August. Plants
grown from seed at Victoria, Tex., and Tallulah, La., did not begin to pioduce buds
until well along in August. The natural dormant period of the Arizona boll weevil
therefore lasts about nine months.

It is interesting to note that the Thurberia weevils extracted from their cells in May
and sent to Victoria, Tex. , immediately began to feed and breed upon cotton and pro¬
duced several generations.

Journal of Agricultural Research

Vol. I, No. »

The Arizona form has either acquired by long years of adversity an ability to sur¬
vive for a longer period without food, assuming Anthonomus grandis Boh. to be the
original species; or if the Thurberia weevil is the true original form, then the ability
to obtain a plentiful supply of early food has caused the species to lose some of its
resistance to adversity.

Feeding. — The adults feed upon the squares and bolls in much the same manner as

the typical Anthonomus grandis.

Feigning death. — The adults
are not quite so easily disturbed as
those of the cotton-feeding form,
but when disturbed they feign
death and drop to the ground or fly
away.

Ovtposition. — On the first day
that any adults were seen, August
25, in the Santa Rita Mountains,
the males were the most abundant
and usually were not feeding, but
were perched on the tips of squares
or on the foliage in an attentive
attitude, evidently waiting for
females.

The egg puncture is almost always made at the base of the square, and the hole is
sealed by a gelatinous scale exuded by the plant, over which there is often a small
mass of excrement. On removal of this scale the egg can often be seen. A majority
of the eggs seen were twice as long as broad, and only one was of the same proportions
as usually found in Anthonomus grandis . In the bolls the position of the egg punc¬
ture is more general.

Development. — The developmental period of the Arizona weevil on its native host
has not been studied, but it has
been watched by Mr. Coad at Vic¬
toria, Tex. ,on cotton. Theperiod
is practically the same as for
Texas weevils, beginning on the
same day: In June, 16 days; in
July, 12.5 days; in September,

17.2 days. The period in bolls
in September is naturally longer,
and no specimens had been car¬
ried completely through at the
time of writing this article.

The most interesting point in
the Victoria work lies in the fact
that in June, when this boll
weevil was removed from hiber¬
nation and transplanted on cot¬
ton, it was able to begin its generations immediately and to continue reproduction
throughout the season.

The food plant of this new variety is known botanically as Thurberia
thespesioides , although it has also been called Gossypium thurberi and
Ingenhouzia triloba. It occurs in southwestern Chihuahua and Guada¬
lajara, Mexico; in the Santa Catalina, Santa Rita, Tanque Verde, Rincon,

Fig. 4. — Anthonomus grandis Boh.: Head and beak: A, Fe¬
male; male. Much enlarged. (Original.)

Fig. 3. — Anthonomus grandis, var. thurberiae: Head and beak:
A, Female; B, male. Much enlarged. (Original.)

Nov. io, 1913

A Cotton Boll Weevil in Arizona

Mule Pass, Huachuca, and Chiricahua Mountains, and also in Fish Creek
Canyon of the upper Salt River valley, and at Dragoon, Fort Bowie,
and Davidson Springs, all in Arizona.

Thurberia grows at altitudes from 2,250 feet to 7,000 feet, and is found
in the bottom of the canyons, on the canyon walls, and on top of the
ridges, growing usually where protected more or less from the greatest
heat of the sun.

The plant begins flowering in some localities in July, but in others it is
just beginning to bud in the latter part of August. Flowering continues
into October.

In appearance Thurberia is so nearly like cotton that the Mexicans and
natives call it “wild cotton.” The leaves are simple, or 3 or 5 lobed,
and in the two latter forms resemble the okra-like form of Upland cotton
{Gossypium hirsutum) or the normal leaves of the Mexican species Gossy-
pium palmeri Watt, and G. schottii Watt. The leaf has a nectary on the

Fig. 6. — Anthonomus grandis Boh.;
Antenna of female. Much en¬
larged. (Original.)

Fig. 5. — Anthonomus grandis , var.
thurberiae : Antenna of female.
Much enlarged. (Original.)

midrib, like cotton, and this nectary is as attractive to insect life as the
leaf nectaries of Egyptian or Upland cotton. The buds differ from cotton
buds by the truncate calyx cup and the linear involucral bracts, but the
three nectaries, which also prove a great attraction to insects, are present
as on cotton squares. The flowers resemble cotton flowers very closely.
The bolls are small, not over three-fourths of an inch in length, and are
3 to 5 celled, with two rows of seed in each. There is a very tiny fiber
on the cell walls.

The plants are perennial, growing to be over 10 feet high, with a spread
of about 10 feet, and having a large, strong, woody trunk. They are
very prolific fruiters. The species is often killed back by frosts, as is
evidenced by the dead terminals with the old bolls of previous seasons.
The heavy wash in the mountain canyons is one of the principal means of
dispersion of the plant.

Thurberia is exceedingly like cotton in most essentials, the relationship
being most clearly demonstrated by the many insects which attack both.

Journal of Agricultural Research

Vol. I, No. a

At least two species of parasites attack the Arizona “ wild-cotton' ’ boll
weevil in the Santa Rita Mountains. One of these is a species of Ceram-
bycobius and the other is a braconid. There are also some predators
which attack it.

Without further information it is idle to speculate as to the direction
of the adaptation which has evidently taken place in Anthonomus grandis .
If further research should locate this boll weevil breeding upon another
genus of plants closely related to cotton, such as Eremoxylum, a genus
of western Mexico, or upon one of the small wild species of Gossypium in
Mexico, the direction of adaptation might be traced. Some of the

Fig, 7. — Anthonomus grandis , var.
thurberiae: A, Front leg; B, middle
leg; C, hind leg. Much enlarged.
(Original.)

Fig. 8. — Anthonomus grandis Boh.:

A , Front leg ; B , middle leg ; C, hind
leg. Much enlarged. (Original.)

differences in the condition of the two varieties which show the range of
adaptivity of the insect are as follows :

The rainfall in the vicinity of Tucson, Ariz., for 40 years has averaged
only 11.66 inches per annum, not reaching 3 inches in any month. July
and August are the months of greatest precipitation.

The rainfall at Victoria, Tex., for 20 years has averaged 36.63 inches
per annum, with over 3 inches in seven months of the year. May is the
month of greatest precipitation.

The rainfall at Opelousas, La., for 17 years has averaged 57.12 inches
per annum, with over 5 inches in six months of the year. July is the
month of greatest precipitation.

Nov. io, 1913

A Cotton Boll Weevil in Arizona

The altitude of Opelousas is 83 feet, of Victoria 145 feet, and of Tucson
2,390 feet. The Arizona boll weevil is found at 4,000 feet altitude and
higher. The highest altitude at which the Texas form has been found
on cotton is under 2,000 feet.

The maximum temperature at Opelousas and Victoria is 104° V., and at
Tucson ii2°. The minimum temperature at Opelousas is 20, at Victoria
6°, at Tucson io°. The mean temperature at Opelousas is 67.3°, at Vic¬
toria 70°, at Tucson 68°. The average date of first killing frost in the
fall for Opelousas is November 17; Victoria, December 10; and for
Tucson, November 22. The average date of last killing frost in spring
for Opelousas is March 5; for Victoria, February 20; and for Tucson,
March 26. At Tucson, August is the only month in which the minimum
temperature does not run below 56° F., which is the zero of effective
temperature for Anthonomus grandis in Texas. At Victoria and Opelou¬
sas the minimum never goes below 56° in July or August.

Of course, in the mountains where Anthonomus grandis thurberiae
occurs the temperature does not reach quite as high a point as at Tucson,
and the minimum temperature is lower. The chilly nights and warm
days probably would
retard the develop¬
ment and hibernation
of the cotton boll
weevil in the same
manner if transplanted
to Arizona mountain
conditions.

The points of great¬
est adaptation are evidently atmospheric pressure and humidity, and
possibly high temperature, although typical individuals of Anthonomus
grandis have been known to survive 1140 F. at Dallas, Tex., while the
excessive drought experienced for several years in northern Texas prac¬
tically exterminated the species.

Cotton is cultivated in the Imperial Valley and the Colorado River
valley in California, in the Salt River valley, the Gila River valley in
eastern and central Arizona, and also in the Santa Cruz River valley of
Arizona.

The varieties grown are mainly long staple — Egyptian and Durango,
with some Triumph. The crops, which are irrigated, are very promising
and can be made with very little water if it is properly applied.

The Arizona “wild cotton,” Thurberia, occurs in nearly every moun¬
tain range in southwestern Arizona where there is any moisture. In the
vicinity of the Santa Cruz Valley cotton is grown within 5 miles of
Thurberia plants growing in the mountains. The boll weevil was not
found on the nearest Thurberia plants, nor were many of the nearest
canyons investigated, but it was found to be abundant not more than 10
miles distant. This is the first year of cotton in the Santa Cruz Valley,
and it is expected that a large acreage will be planted in 1914.

Fig. 9. — Anthonomus grandis , var. thurberiae; Wing.

Journal of Agricultural Research

Vol. I, No. 2

Thurberia is known to occur in Fish Creek Canyon, one of the sources
of the Salt River. This valley has the most extensive cotton plantings
in Arizona. However, the boll weevil has not been observed there.

No observations have been made in the vicinity of the Gila River
valley, but as Thurberia occur§ in the mountains both north and south
of this valley, it undoubtedly also occurs in some of the ranges bordering
the valley.

The Arizona weevil may be able to cover considerable distances by
flight, especially if compelled to seek sustenance elsewhere. However,
it will probably cleave to its native food plant as long as this gives
sufficiently abundant food, though a great increase of weevils or a
decrease of food might drive them to seek other food. They would
take more readily to cotton than anything else, and once they find the
rich, succulent cotton, with its plentiful food and moistened soil, they
will probably do serious damage. It is to be feared that a wholesale
destruction of the native food plant might invite a quicker than natural
adaptation to cotton on the part of this western weevil. This matter
is now under investigation, but at the present time it is the writer’s
personal opinion that the safest plan is to preserve the status quo of
the weevil in the mountains. An introduction of parasites from the
cotton boll weevil would be of considerable assistance in reducing the
Arizona weevil and would not cause its dispersal.

There is danger of a distribution of weevil-infested buds through the
drainage system by summer freshets. After such occurrences the cotton
should be watched very closely for several weeks for the appearance of
weevils.

The cotton boll weevil has never been able to successfully invade the
drier cotton sections of western and northwestern Texas, although it has
been expected that it will gradually adapt itself to the more rigid con¬
ditions of these sections. It is of extreme importance that the Arizona
weevil be kept out of western Texas and any part of the southeast, except
when under very careful isolated observation of specialists. If acci¬
dentally introduced into other sections, the Thurberia weevil might be
able to stand much greater variations of climate than Anthonomus
grandis Boh. and become a much more powerful pest. Furthermore,
there is every reason to believe that Anthonomus grandis thurberiae could
withstand the rigors of the climate of western Texas.

It is therefore important that restriction by quarantine be considered,
and this matter will be taken up at an early date by the Federal Horti¬
cultural Board.

DESCRIPTION OF PLATE

Plate VI. Figs, i, 2, 5, and 6. — Anthonomus grandis thurberiae : Type specimens;

actual length ,5.5 to 6 mm . Figs. 1 and 5 . — Side and dorsal views of male .
Figs. 2 and 6. — Side and dorsal views of female . Enlarged . ( Original . )

Figs. 3, 4, 7, and 8. — Anthonomus grandis : Typical specimens; actual
length, 5.5 to 6 mm. Figs. 3 and 7. — Side and dorsal views of female.
Enlarged . ( Original . )

Fig. 9. — Thurberia thespesioides: Section of boll, showing cell of Anthono¬
mus grandis thurberiae. Enlarged. (Original.)

Fig. 10. — Thurberia thespesioides: Seed, showing cell of A nthonomus grandis
thurberiae. Enlarged. (Original.)

Fig. 11 .j — Thurberia thespesioides: Boll, showing egg puncture of Anthono¬
mus grandis thurberiae. Enlarged. (Original.)

(98)

A Cotton Boll Weevil in Arh

Plate VI

W 1

¥ K 1

m 1 1

5 A

THE DIAGNOSIS OF DOURINE BY COMPLEMENT

FIXATION

By John R. Moheer, Adoeph Eichhorn, and John M. Buck,
Pathological Division , Bureau of Animal Industry

INTRODUCTION

Dourine is a specific infectious disease affecting under natural condi¬
tions only the horse and the ass, transmitted from animal to animal
by the act of copulation, and due to a single-celled animal parasite or
protozoan, the Trypanosoma equiperdum . It is characterized by an
irregular incubation period, the confinement of the first symptoms to
the genital tract, the chronic course which it runs, and by finally pro¬
ducing complete paralysis of the posterior extremities, with a fatal
termination, as a rule, in from six months to two years.

HISTORY OF DOURINE IN THE UNITED STATES

In the United States the disease was first suspected in 1885 and
recognized in 1886 by Dr. W. L. Williams, who was then a veterinary
practitioner at Bloomington, Ill. Officials of the State of Illinois took
hold of the outbreak, and as a result of rigid prophylactic measures the
disease was eradicated from the State in 1888, but not before an affected
stallion had been shipped to Gordon, Nebr., thereby starting up a new
center of infection in that locality.

In 1892 dourine was again brought into public notice by an outbreak
among the breeding horses of northwestern Nebraska, the history of
which suggested that it originated with this Gordon stallion. After an
expenditure of about $5,500 by the Bureau of Animal Industry the
disease was considered to have been eradicated from that section of
the country. Five years later the infection again made its appearance
in the same part of Nebraska, and early in 1899 the Bureau again began
the work of eradication. Many inspections were made, and those
animals which were found diseased were purchased and killed. Many
obstacles were encountered, and the disease evidently kept smoldering
during 1900.

In 1901 the infection reappeared with increased vigor, this time in
the Pine Ridge and Rosebud Indian Reservations in South Dakota, in
addition to northern Nebraska, and more stringent measures were
immediately inaugurated to control the spread of the disease. However,
eradication in this region was extremely difficult, owing to the wildness

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(99)

Vol. I, No. 2
Nov. 10, 1913
A— a

IOO

Journal of Agricultural Research

Vol. I, No. 2

of the country as well as of the horses and the fact that many horse
owners would try to conceal from the inspectors animals which they
knew to be affected with the disease. In 1906 the last suspicious cases
of dourine were destroyed in South Dakota.

In the meantime, during the year 1903, dourine was reported in Van
Buren County, Iowa, and successful steps were immediately taken to
stamp it out. No connection could be established between this outbreak
and that in Nebraska, but it was quite definitely determined that an
imported Percheron stallion purchased by a company of farmers was
responsible for its appearance.

Another outbreak of dourine was discovered in Taylor County, Iowa,
in 1911. The diseased animals, together with all exposed stallions and
mares, were immediately quarantined by the State. Those showing
lesions of the disease and those exposed horses that reacted to the
complement-fixation test were purchased by the Government and
destroyed. It is now believed that the infection is entirely eradicated
from Iowa. The source from which this center of infection was derived
is only a matter of conjecture, but there is apparently no connection
between this and any of the previous outbreaks. No authentic informa¬
tion as to the origin of the outbreak was discovered, but all cases lead
back to a Percheron stallion which was imported in 1909 and brought
direct to Tenox, Iowa.

Early in July, 1912, the State Veterinarian of Montana reported several
suspicious cases of dourine in eastern Montana and forwarded blood
sera from the suspected animals for the complement-fixation test.
All but one sample gave positive results, thus establishing a new center
of infection of dourine. From present indications this outbreak appears
to be more extensive than any of the previous outbreaks, involving also
two Indian reservations in North Dakota and South Dakota; but a force
of 12 Federal veterinarians assisted by State representatives is at work
on the disease, and the infection is well under control.

SEARCH FOR A METHOD OF DIAGNOSIS

The difficulty of diagnosing chronic and latent forms of dourine is
generally recognized, and owing to this fact the control and eradication
of this disease in horses has been of slow progress and sometimes in¬
effective. In such outbreaks it has been the custom to trace the disease
as far as possible to its origin and then to keep under observation all
mares and stallions which directly or indirectly have been exposed to
the disease. At the same time animals which show clinical evidences of
the affection are destroyed without delay. By this means several of
the outbreaks which have occurred in the United States have been
checked and eradicated.

The attempt to make a microscopical demonstration of the Trypano¬
soma equiperdum in affected horses is very frequently unsuccessful,

Nov. io, 1913

Diagnosis of Dourine

101

although our more recent experience proves that the organism may occa¬
sionally be found in the serous exudate of the plaques and also in the
fluid of the edematous swellings of the genital organs in the stallions
as well as in the mares.

Of course, this procedure of diagnosis can be attempted only when
the disease occurs in farming localities where the animals can be readily
observed and examined as desired. On the other hand, in the present
outbreak in Montana and adjoining States the conditions make the
diagnosis by the demonstration of trypanosomes impossible, and, like¬
wise, animal inoculations can not be satisfactorily utilized for this purpose.
Horses in that locality are bred under range conditions; they run wild
and a round-up takes place only once a year. The difficulty of an exami¬
nation, even clinically, of such animals is obvious, since they have
not been broken to the halter and are troublesome to handle.

Our experience with the disease in Montana showed that only a limited
number of animals were clinically affected. Nevertheless, the associa¬
tion of all the animals without any restriction in the breeding periods
indicated that a larger number of animals would be found infected,
which, as a matter of fact, has been proved by subsequent tests, as
hereinafter shown.

Owing to the fact that until the last few years the eradication of
dourine in this country was supposed to have been complete, the disease
has received only slight attention as compared with other menacing
diseases of our domesticated animals. It was not until the outbreak
in the State of Iowa in 1911 that the necessity for devising a method of
diagnosing this infection began to be fully realized. The value of being
able to detect the latent and to verify the clinical cases became apparent.
Otherwise, the necessity existed of maintaining a long-continued quar¬
antine in those sections of the country where cases have been discovered.
While little difficulty has been experienced in recognizing the advanced
cases, a clinical examination alone naturally permitted many infected
animals to escape detection, only to facilitate the further spread of the
disease until the appearance of symptoms made the diagnosis unques¬
tionable.

Inasmuch as the complement-fixation method of diagnosis has been
employed with gratifying results in connection with numerous other
diseases, the possibility of applying this method to dourine naturally
suggested itself, and steps were therefore taken to determine the feasi¬
bility of its application to this disease.

It was very early discovered that the problem of preparing a satis¬
factory antigen would offer considerable difficulty. Efforts were pri¬
marily directed toward utilizing for this purpose the different organs of
those horses that had succumbed to the disease. Several of the clinical
cases were shipped from Iowa to the Bethesda Experiment Station
during the outbreak referred to, in order that a more complete observa-

102

Journal of Agricultural Research

Vol. I, No. a

tion might be made of the development of the disease and that material
might at the same time be available for the preparation of an antigen.
From time to time, as these animals died, certain tissues were obtained
which it was suspected might furnish the desired results, but although
shake extracts of the spleens, livers, kidneys, and bone marrow, as well
as alcoholic and acetone preparations, were employed under various
conditions, the results were rather discouraging.

Subsequent to this time there came under our observation publica¬
tions by numerous investigators who had given this subject considera¬
tion. It will suffice to mention the publications of Landsteiner, Muller
and Potzl, Levaditi and Yamanouchi, Hartoch and Yakimoff, Citron,
Weber, Manteufel, Manteufel and Woithe, Zwick and Fischer, and
Schilling, Claus, and Hosslin. The results in these instances appeared
to have been unsatisfactory, which was also the case in the extensive
work on the diagnosis of dourine by the Wassermann method by Trajan
Pavlosevici, as he concluded that while antibodies can be demonstrated
by this method in laboratory animals infected with trypanosomes, the
method can not be utilized in stallions affected with dourine.

Later, Winkler and Wyschelessky, Mohler, and also Watson in their
work on complement fixation as an aid in the recognition of trypanoso¬
miasis indicated the good results obtained in the diagnosis of dourine.
Likewise, Mattes in his work on the agglutination of trypanosomes ob¬
tained gratifying results, while Braun also concludes that complement
fixation can be utilized for the diagnosis of trypanosome affections.

In the recorded publications it was observed that the more promising
results were obtained by those who employed suspensions of pure try¬
panosomes. The organ extracts and other preparations of antigens
generally used for this purpose proved unreliable. The procedure as
recommended by various workers in obtaining an antigen from pure
trypanosomes and using such a suspension as the antigen has also been
tried by the writers with uniformly good results. The practical applica¬
tion of this procedure, however, would be very laborious and require a
great deal of time, especially in cases where a large number of horses
have to be tested by this method. Accordingly it was deemed advisable
to devise a means by which an antigen could be prepared which would give
similarly good results but would not require such delicate and laborious
technique. In place of the specific trypanosome of dourine being util¬
ized, the writers selected the surra organism, as it had been previously
ascertained by several investigators that the reaction obtained was not
absolutely specific for any one trpyanosome infection but was rather of a
group nature. As dourine is the only known trypanosome affection of
horses existing in this country, the value of even a group reaction was
immediately appreciated, and attention was directed to the carrying
out of this idea in our diagnostic work.

Nov. 10, 1913

Diagnosis of Dourine

103

In place of preparing suspensions of the trypanosomes, however, an
antigen was made of the blood and macerated spleens of rats killed at the
height of surra infection. This material was placed in a bottle containing
glass beads and shaken for six hours, filtered through gauze, and car-
bolized. The results from this antigen proved satisfactory, and it was
used repeatedly on the blood of the horses affected with dourine that
were left of the Iowa shipment.

The smallest quantity of the serum which gave a positive reaction with
the antigen was 0.05 c. c. ; however, the various comparative tests indi¬
cated that fixation in tubes containing 0.2 c. c. of serum is sufficient for
diagnostic purposes. Sera from normal animals, also those affected with
various other diseases, failed to give a reaction. This antigen proved
active on 10 consecutive days, but failed to produce fixation of complement
on subsequent tests. Later attempts by the same procedure also resulted
less satisfactorily, and it was therefore deemed advisable to try other
methods in order to procure an antigen of more uniform action.

The following procedure was next employed :

After successive examinations of the blood of a dog infected with
surra, about 200 c. c. of blood were drawn from the jugular vein when
the microscopic examination revealed an extreme infestation with the
parasite. The blood was drawn into a 1 per cent potassium-citrate
solution in large centrifuge tubes of 100 c. c. capacity. A quantity of
potassium-citrate solution was used equal to the amount of blood drawn
into each tube, and 0.5 gram of saponin was added to each tube in order
to dissolve the red blood corpuscles. After a thorough shaking and after
complete hemolysis had taken place, the tube was centrifuged for 30
minutes at 2,500 revolutions, and the supernatant fluid was siphoned off.
The residue, which was of an opaque color and consisted principally of
trypanosomes, was then thoroughly mixed and shaken up with salt solu¬
tion, when it was again placed in the centrifuge; this washing was
repeated three times. After the last washing the thrown-down opaque
mass was emulsified with 50 c. c. of salt solution and titered as to its
merits as an antigen for dourine tests. The results were highly satisfac¬
tory, and the titer was established at 0.5 c. c. of this emulsion per tube.
However, the disadvantages of this method — namely, the difficulty in the
preparation of this antigen and also the small quantity which was obtain¬
able from a single bleeding of a dog — were soon apparent.

In July, 1912, the outbreak of dourine in Montana was discovered, as
already mentioned. Several samples of blood sera from clinical cases
were forwarded by the State authorities to the Bureau of Animal Indus¬
try for verification. Positive reactions were obtained in numerous
instances with antigens thus prepared, establishing conclusively the
presence of the disease in that State, as well as suggesting the possibilities
of the test as a means of its eradication. It was not long before dis-

104

Journal of Agricultural Research

Vol. I, No. a

covery was made that the disease was quite widely spread in Montana
owing to the previous failure to recognize it. In an endeavor to comply
with the request of the State authorities to make diagnoses in a large
number of animals, it was soon apparent that a different method would
necessarily have to be devised in order to make the desired progress.

PREPARATION OF ANTIGEN

Various organs from rats just dead from surra were tried out in both
fresh and preserved states, and the results which were obtained from the
fresh suspension of the macerated spleen of a rat just dead from surra
were the most promising. In order to establish whether such an antigen
would constantly, or at least in most instances, give the results desired,
it was repeatedly tested on positive sera of horses affected with dourine,
as well as on horse serum known to be free from immune bodies of
dourine. After repeated tests on horses clinically affected with dourine
had shown the antigen to be uniformly constant in its action, the pro¬
cedure of diagnosing dourine by this method was definitely adopted.
It was at this time that our present method of preparing antigen was
first employed, which is as follows:

Gray or white rats are infected with surra by the injection of 0.2 c. c.
of blood from a rabbit infected with that disease. Since tests have to be
made every day to keep up with the large number of cases submitted
and as the antigen proves effective only when prepared fresh, it was
arranged that at least two rats should die daily with the disease. When
the rats appeared to be at the point of death late in the afternoon it was
found that placing such rats in the ice chest until they died furnished a
better antigen than when they have died in the cage during the night
and have to be used the following morning.

The spleens of the rats are removed, placed in a mortar, and ground up
with a small amount of salt solution to a pulpy mass. From time to
time more of the salt solution is added, and the suspension thus obtained
is filtered twice through a double layer of gauze into a test tube. The
quantity of the suspension from each spleen is made up to 40 c. c. by
dilution with salt solution.

This suspension constitutes the antigen for the tests of the suspected
dourine sera. Dr. Jacob Traum, who was temporarily assigned to this
work, found that when the suspension was titered against sera in gradu¬
ated quantities from a known positive and a known negative case the
best results were obtained, and this method has since been adopted. The
quantity of antigen employed is double the amount necessary to pro¬
duce complete fixation with positive serum. The following table gives
the method practiced in titrating the antigen :

Nov. 10, 1913

Diagnosis of Dourine

105

Table showing method of titration of antigen for the complement-fixation test in dourine .

1 "

Positive serum.

Tube No.

NaCl so¬
lution.1 *

Serum.

Antigen.3

Comple¬

ment.3

Hemo¬

lytic

serum.4

Blood cor¬
puscles.6

C. c.

C. e.

O. 15

0. 05

• TS

. I

• 15

. 2

• *5

■25

•15

■3

Negative serum.

0. 15

0. I

• 15

• i5

•

•3

•4

w O

0 £

33 3

H O

• 15

• 5

u.B

• J5

. 6

1 0.85 per cent NaCl solution.

3 Suspension of macerated spleen from rat.

3 The determined smallest quantity established by titration,

4 Sensitized rabbit serum.

3 S per cent suspension of red blood corpuscles of sheep.

Half the quantity of antigen which in the negative serum does not
inhibit hemolysis, provided this quantity is at least double the amount
necessary to produce complete fixation with the positive serum, indi¬
cates the titer of the antigen. For instance, if tubes Nos. 1, 2, 3, and 4
of negative serum show complete hemolysis and Nos. 5 and 6 slight
inhibition, and at the same time tubes Nos. 6, 5, 4, 3, and 2 of positive
serum show complete fixation and No. 1 partial fixation, the quantity
of antigen for the test proper would be 0.2 c. c. of the antigen.

Occasionally the antigen does not prove satisfactory for the test and
has to be discarded. In these cases the fixation in all tubes is apparently
due to the excessive amount of proteids from the spleen. Experience
has shown that the excessively large spleens contribute such an antigen.
This, of course, is indicated by the titration undertaken prior to the regu¬
lar test. At other times it was found that the antigen proved satisfactory
the following day, after it was allowed to stand in the test tube over¬
night and the supernatant fluid drawn off for the antigen. This is then
retitered and the titer established in accordance with the results of the
test.

THE COMPLEMENT-FIXATION TEST

The test proper for the diagnosis of dourine is carried out in a manner
similar to that practiced for the diagnosis of glanders.1

1 A more detailed description of the technique of this method as applied to glanders is given by Mohler

and Bichhom in Bulletin 136, Bureau of Animal Industry, entitled “The diagnosis of glanders by com¬

plement fixation/’

io6

Journal of Agricultural Research

Vol. I, No. a

The hemolytic system consists of sensitized rabbit serum, serum from
a guinea pig, and a 5 per cent suspension of washed sheep corpuscles.

The serum to be tested is, of course, inactivated for one-half hour at
56° C. and is used in the tests in quantities of 0.15 c. c., since it has been
found that fixation in this quantity is obtained only with sera of horses
affected with dourine. Tests to determine the smallest quantity of serum
of horses having dourine which will give a fixation showed that in
several instances even 0.02 c. c. of serum was sufficient to give a com¬
plete fixation.

The complement from the guinea pig is always titered previous to the
test, as it is absolutely necessary to use the exact amount of the comple¬
ment to obtain the best results, since a deficiency or an excess of the com¬
plement would interfere greatly with the reaction.

In the numerous cases which have been tested the results were almost
invariably definite, and only on a very few occasions was it found neces¬
sary to make retests on cases which appeared atypical. The reaction is
always very marked, and in our work only a complement fixation with
the quantity of serum mentioned is recognized as a positive reaction.
It is only proper that in the tests the usual number of checks should be
employed in order to insure reliable results.

Since the testing has been undertaken by the method described,
8,657 samples have been examined from Montana and the Cheyenne and
Standing Rock Indian Reservations in North Dakota and South Dakota.
Of these, 1,076 gave positive reactions, which appears to be a very large
proportion, but when it is remembered that these animals were kept
under range conditions without sanitary or veterinary control and also
that before the disease was recognized as dourine it had been diagnosed
for a long period as some other affection, it will be apparent that the
opportunity for the spread of the disease was ideal.

With the present system of diagnosis, by which even the latent cases
can be determined, it is hoped to eradicate the disease quickly. All the
horses in the infected localities will be submitted to the complement-
fixation test, and by cooperation with the State authorities means will
be devised to dispose of the affected animals in such a way as to make
the further spread of the disease impossible. The animals which were
destroyed as a result of the disease in the above-named localities and
which were diagnosed by the complement-fixation test showed in most
instances some lesions indicative of the disease. In some of the cases
there were no indications of a progressive paralysis, but the lesions
existing in the genital organs of either the male or female were sufficient
for confirmation of the diagnosis by the complement-fixation test.

It is therefore evident that the diagnosis of trypanosome infections
of both man and animal by the complement-fixation test is of very great
importance, especially in countries where only one of these protozoan

Nov. io, 1913

Diagnosis of Dourine

107

diseases exists. By this means it is possible to determine all infected
persons or animals within a short time and adopt such hygienic measures
as would be best suited for the control of the infection. Furthermore,
the introduction of a disease like dourine into any country could also
be guarded against by a compulsory requirement of this test on all
horses imported from countries in which dourine is present.

BIBLIOGRAPHY

Braun, H. Uber das Verhalten der Trypanosomen Antikorpem gegeniiber. Centlbl.

Bakt. [etc.], Abt. 1, Ref., Bd. 54, Beil., p. 11-16, 1912.

Citron, Julius. Die Komplementbindungsversuche bei Erkrankungen mit bekann-
ten, aber nicht zuchtbaren Erregem. Kraus, Rudolf, and Levaditi, C.: Hand-
buch der Technik und Methodik der Immunitatsforschung, Bd. 2, Jena, 1909,
p. 1112.

Hartoch, O., and Yakimorr, W. Zur Frage der Komplementbindung bei experi-
mentellen Trypanosomen. Wiener Klin. Wchnschr., Jabrg. 21, No. 21, p. 753—
755, Mai 21, 1908.

Landsteiner, K., MtjllEr, R., and P6tzl, O. Uber Komplementbindungsreak-
tionen mit dem Serum von Dourinetieren. Wiener Klin. Wchnschr., Jahrg. 20,
No. 46, p. 1421-1422, Nov. 14, 1907.

- Zur Frage der Komplementbindungsreaktionen bei Syphilis. Wiener Klin.

Wchnschr., Jahrg. 20, No. 50, p. 1565-1567, Dez. 12, 1907.

Levaditi, C. , and Y amanouchi, T. La reaction des lipoldes dans les Trypanosomiases
et les spirilloses experimentales. Bui. Soc. Path. Exot. [Paris], t. 1, No. 3, p.
140-144, 1908.

ManteuREL. Untersuchungen uber spezifische Agglomeration und Komplement¬
bindung bei Trypanosomen und Spirochaeten. Arb. K. Gsndhtsamt. [Germany],
Bd. 28, Heft 1, p. 172-197, Marz, 1908.

- and Woithe. Uber die diagnostische Bedeutung der Komplement-

bindungsreaktion bei Trypanosomeninfektionen. Arb. K. Gsndhtsamt. [Ger¬
many], Bd. 29, Heft 2, p. 452-477, 1908.

Mattes, Wilhelm. Agglutinationserscheinungen bei den Trypanosomen der Schlaf-
krankheit, Nagana, Dourine, Beschalseuche, und des Kongokustenfiebers.
Centlbl. Bakt. [etc.], Abt. 1, Orig. Bd. 65, Heft 6/7, p. 53^-573, Aug. 10, 1912.
MohlEr, John R. Dourine. Report of committee on diseases. Proc. Amer. Vet.
Med. Assoc., 1912, p. 99-115, 1913.

PavlosEvtci. Recherches sur Tapplication de la m£thode Wassermann, dans le diag¬
nostic de la dourine. Arch. Vet. [Bucharest], v. 7, No. 2, p. 69-82. Mar.-Apr.
1910.

Schilling, Claus, and Hosslin, V. Trypanosomen- Inf ektion und Komplement¬
bindung. Deut. Med. Wchnschr., Jahrg. 34, No. 33, p. 1422-1425, Aug. 13, 1908.
Watson, E. A. The serum reactions and serum diagnosis of dourine. Proc. Amer.
Vet. Med. Assoc., 1912, p. 411-420, 1913.

Weber, Hans. Uber Immunisirungs- und Behandlungsversuche bei Trypanosomen-
krankheiten. Ztschr, fur Expt. Path. u. Ther., Bd, 4, Heft 2, p. 576-626, 1907.
Winkler and WyschelESSKy, S. Die Agglutination, Prazipitation, und Komple¬
mentbindung als Hilfsmittel zum Nachweis der Trypanosomenkrankheiten im
besonderen der Beschalseuche. Berlin. Tierarztl. Wchnschr., Jahrg. 27, No. 51,
P- 933“936> T>ez. 21, 1911.

Zwick and Fischer. Untersuchungen uber die Beschalseuche. Arb. K. Gsndhtsamt.
[Germany], Bd. 36, Heft 1, p. 1-103, 1910.

THREE UNDESCRIBED HEART-ROTS OF HARDWOOD
TREES, ESPECIALLY OF OAK

By W. H. Long,

Forest Pathologist , Investigations in Forest Pathology , Bureau of Plant Industry

INTRODUCTION

During an investigation made in 1912 of the pathological condition of
the oaks in the Ozark National Forest, of Arkansas, and in other sections
of the United States the writer found a large percentage of the trees,
especially in some regions of Arkansas, attacked by various fungi which
rot the heartwood. Twenty different kinds of heart-rots were found.
Of this number eight have been previously described and assigned to
their causative fungi; two were caused by well-known fungi, but no
detailed specific descriptions of the rots have yet been published; one
proved to be a true root-rot caused by Polyporus dryadeus ; three
have not yet been connected with their causative organisms; while six
have been for the first time definitely associated by the writer with the
fungi which produce them. Only three of these last six rots will be
discussed in this paper.

INVESTIGATIONS OF HEART-ROTTING FUNGI

The writer found in the Ozark National Forest ideal conditions for the
study of heart-rotting fungi, as thousands of white-oak trees ( Quercus
alba L.) were being worked into 36-inch staves for whisky barrels. Trees
over 16 inches in diameter were felled and sawed into about 3-foot
lengths; these were immediately split into what are known as bolts.
As only perfectly sound timber can be used for whisky staves, all rotten,
wormy, water-soaked, and stained pieces were rejected and left on the
ground where the tree was cut. It was therefore very easy to determine
the character and extent of the rot in each tree. As the areas being
cut were in a virgin forest, all ages of trees down to about 160 years old
(16 inches in diameter) were included. The majority of the trees were
cut very close to the ground; the stumps averaged 12 inches in height,
but in many cases were much lower. This aided in the investigation,
since the nearer the ground the trees were cut the more complete was
the record as to the rot in the trunks.

Of the twenty rots found in oak, only the following eight were pres¬
ent to any extent in the trunks and tops of the trees:

(1) A rot which produces hollows caused by Hydnum erinaceus;
(2) a brown, checked rot caused by Polyporus sulphur eus; (3) a

(109)

Journal of Agricultural Research.

Dept, of Agriculture, Washington, D. C.

Vol. I, No. a
Nov. 10, 1913
G-a

IIO

Journal of Agricultural Research

Vol. I, No. 2

whitish heart-rot, piped in its earliest stages and common in the upper
half of the trees, due to P . dryophilus; (4) a string and ray rot in
the butts of the trees, due to P. berkeleyi; (5) a straw-colored rot
caused by P. frondosus; (6) a white piped or pocketed rot caused by
P. pilotae; (7) a brown, brittle rot, cause unknown; and (8) a tough,
spongy, whitish rot caused by Fomes lobatus .

Of these eight rots the bulk of the damage to the timber in the butts of
the trees is caused by the following fungi, named in the order of their im¬
portance: Hydnum erinaceus, Polyporus pilotae , P. sulphur eus , P. berke¬
leyi , and P. frondosus . However, P. dryophilus causes a most common
and very injurious heart-rot of the upper trunk and limbs of oaks in the
Ozarks.

Although 64.8 per cent of the felled oak trees studied in the Ozarks
were affected with butt-rots, the amount of merchantable timber actually
destroyed by these fungi was comparatively small, owing to the fact
that these rots do not ascend very high in the trees. More than 2,100
felled oak trees were carefully studied by the writer, and extensive data
concerning each tree were recorded. Of the entire number 1,938 were
white oaks.

Table I shows the various heights of each rot in the trees down to a cer¬
tain limit, together with the corresponding stump diameter, the diameter
of the rot for each tree, and the number of trees for each recorded rot
height. For example, the first line, reading across the page, shows
the name of the rot — “ hollow-producing rot”; cause — “Hydnum erina -
ceus”; diameter of the stump — “26 inches;” diameter of the rot in the
stump — “17 inches” ; height of rot in the bole of the tree — “28 feet” ; and
the number of trees with this height of rot — “1.” Where more than one
tree has a particular rot of a given height the diameters of the stumps
and the diameters of the rot in the stumps are averaged, and the resulting
numbers are shown in the proper columns.

Nov. 10, 1913

Heart-Rots of Hardwood Trees

Table I. — Data on jive types of butt-rots found in white oak (Quercus alba L.).

Name of rot.

Cause.

Diameter
of stump.

Diameter
of rot in
stump.

Maxi¬
mum
height
of rot in
butt.

Number
of trees
having rot
of the
given
height.

Inches.

Feet.

' 26

3<5

Hollow-producing rot . . .

Hydnum erinaceus ....

2 6

Pocketed or piped rot . .

Polyporus pilotae .

Brown, checked rot. . . .

Polyporus sulphureus . .

String and ray rot .

Polyporus berkeleyi . . .

■ 28

Straw-colored rot .

Polyporus frondosus. . .

1 12

Journal of Agricultural Research

Vol. I, No. 2

Table I. — Data on jive types of butt-rots found in white oak (Quercus alba L.)— Contd.

SUMMARY.

Average —

Total

Name of rot.

Cause.

Diameter
of stump.

Diameter
of rot in
stump.

Height
of rot in
butt.

Age oi rot.

number
of trees
infected.

Hollow-producing

Hydnum erinaceus

Inches .
26. O

Inches.
12. 6

Feet.

3-9

Years.

648

rot.

Pocketed or piped

Polyporus pilotae . .

25. 6

I3- 7

3*9

156

408

rot.

Brown, checked

Polyporus sulphu-

25-8

13. 6

3-o

270

rot.

String and ray rot . .

reus.

Polyporus berke-

28. O

19. 0

3- 5

190

Straw-colored rot. .

leyi.

Polyporus frondo-

27. O

14. O

2- 3

sus.

In the above summary are given certain data for each of the most
important butt-rotting fungi in white oaks, and from them some
idea can be obtained as to the amount of damage done by these heart-
rotting fungi in the virgin timber of the Ozark National Forest. All of
the rots listed in the table are also found in black oak (< Quercus velutina
Lam.), as well as in white oak, but on account of the limited number of
trees of this species examined no data are now given for it. All height
and diameter measurements given in this article, unless otherwise stated,
were taken from the tops of stumps 12 inches high.

In determining the age of the rot only trees were used in which the
fungus had undoubtedly entered at an old fire scar long since healed over.
The annual rings of wood were counted from the point where the callus had
completely closed the wound, so that the heart-rotting fungus must have
entered before the wound was covered. Therefore, the figures given here
represent the minimum age for each infection. The rot might have
entered sooner and therefore be older, but it could not have entered
later and therefore be younger, as the callus had closed the wound. No
stumps with open wounds of any kind were used in estimating the age
of the rot.

The writer realizes that this method of determining the length of time
the fungus has been in a tree is open to the following criticism :

(1) The fungus might have entered underground through injuries
which reached to the heartwood of the root and thence moved upward
into the bole of the tree; (2) the wound made by the fire may have healed
above ground, but not below on the stool and roots of the tree, thus

Nov. io, 1913

Heart-Rots of Hardwood Trees

113

leaving a permanent opening into the heartwood of the trunk just below
the surface of the ground. Through such hidden openings the mycelium
of any heart- rotting fungus capable of growing in the forest debris could
enter the tree, and thus the resultant rot would be directly associated with
the old fire scar; (3) some of these heart-rotting fungi may be able to
enter through sound, unbroken living roots and then move upward as a
heart-rot into the bole of the tree. In this case they would also be true
root parasites and not simply mere heart-rotting fungi.

None of the three rots discussed in this article are known to be true
root parasites. As to the first objection mentioned, the writer has inves¬
tigated several hundred uprooted oak stumps, many of which had heart-
rot, and in no instance was any evidence found indicating that the heart-
rotting fungus entered through the roots and thence worked upward in
the tree. On the contrary, repeated instances were found where the rot
began at the surface of the ground in an old fire scar or other wound and
moved downward in the heartwood of the root and upward in the bole of
the tree. In every case where the rot had entered the roots it had
evidently come from above and not from below, as the rot was limited to
the heartwood of the root, while the sapwood was alive and sound.
However, there is a large wood borer which lives in the roots of oaks,
and when its burrows reach the surface of the roots an opening would be
made for any fungi to enter from the soil. It is well known that in the
roots of oaks the amount of heartwood compared to that of sapwood is
very small. This in itself makes improbable the entrance of heart-rotting
fungi through the roots, especially sound ones.

In regard to the second objection mentioned, the writer has recently
examined more than 200 oak trees with fire-scarred bases, and not a single
one was found in which the wounds having healed above ground had
not also completely healed over below the ground. As a rule, forest fires
injure the tree but a short distance, 2 to 3 inches, below the collar of the
tree, owing to the protection of the soil. Therefore, it is not impossible
for these three heart-rotting fungi to enter through the root system; but
taking the above facts into consideration it is improbable that they did
enter by this route, even granting that they are capable of leading a
purely saprophytic existence in the soil and forest debris — a condition
yet to be proved.

The very close association of the heart-rots with the old fire scars in
the trees studied is so evident that undoubtedly the causal fungi entered
the tree by this route. So marked is this association of fire scars with
heart-rots in the Ozarks that one could tell the areas in the forest which
had been most frequently burned over from the percentage of trees
affected with heart-rots.

The writer has found three types of heart-rotting fungi in living trees :

(1) Those limited to the base and lower portion or butt of the tree, for
example, Polyporus berkeleyi and P. frondosus; (2) those which are able

Journal of Agricultural Research

Vol. I, No. 2

114

to enter either at the butt or in the top of the tree, such as Hydnum erina-
ceust Polyporus sulphur eus, and P. pilotae ; (3) those which enter the
upper portion of the tree and work in both directions from the point of
entrance, but rarely, if at all, enter through fire scars at the butt, such
as P. dryophilus and Fomes everhartii.

THREE UNDESCRIBED TYPES OF HEART-ROTS

In a later article the writer expects to discuss a large number of heart-
rots of the oak, limiting this paper to a detailed description of the following
rots : A pocketed or piped rot of the oak, chestnut, and chinquapin caused
by Polyporus pilotae; a string and ray rot of the oak caused by P.
berkeleyi; and a straw-colored rot of oak caused by P. frondosus .

A POCKETED OR PIPED ROT CAUSED BY POLYPORUS PILOTAE

The rot produced by P. pilotae has been found by the writer directly
associated with the sporophores of this fungus in the following species of
trees: Quercus alba L., Q . velutina Lam., Q. texana Buck!., Q . coccinea
Muenchh., Castanea pumila (L.) Mill., and C. dentata (Marsh) Borkh.

A Pocketed or Piped Rot in White Oak

The description of the pocketed or piped rot which follows was made
from the diseased wood of a white-oak tree ( Quercus alba) which was cut
on July 23, and on August 27 the sporophores of Polyporus pilotae shown
in Plate VII, figure 1 , were found fully developed on the end of the log.
There could be no question as to the identity of the fungus producing
the rot in this case, as less than 30 days had intervened between the
felling of the living tree and the formation of the sporophore of P. pilotae .

The first indication of this rot in white oak is a slight browning of the
heartwood. Later white, oval, or circular cellulose patches from deligni-
fication appear in this discolored wood. These white areas by dissolution
of the fibers often become holes, which show in both radial and cross
section (PI. VII, fig. 2). The delignification seems to originate in the last
layers of the summer-wood fibers and spreads in a very irregular manner.
In later stages long strings of white cellulose fibers are found. This is
especially true where an abundance of air and rain water can reach
the rotting area, especially in old dead logs or in trees with cracks or in
hollow, open butts. The delignification and absorption of the fibers
do not follow the spring wood as closely as they do in the scarlet oak
(1 Quercus coccinea ).

Another type of cavity may be formed which seen in radial view is
0.5 to 1 mm. by 1 to 2 mm. in size. These cavities are lined with the
ends of the white cellulose fibers and usually occur in and at right angles
to the large spring vessels, but they may also extend radially from one
annual ring to the next in a more or less winding or interrupted course.

Nov. io, 1913

Heart-Rots of Hardwood Trees

Under the microscope the large, thick-walled, colorless hyphae are plainly
seen in these holes, and to them the holes undoubtedly owe their origin.
The edges of the perforated vessels as well as the adjacent cells have been
delignified. This type of cavity was especially abundant in the wood
immediately adjacent to the sporophore.

The final stage of this rot in white oak seems to present one of two
conditions: If an abundance of air and water is present, all the wood
fibers will be changed to cellulose, then dissolved, leaving a very light,
brittle, rotted wood of a dark-brown color, which later gradually crumbles
into a dirtlike mass. This is the type of rot usually found in dead trees
or living trees with hollow, open butts. If, on the other hand, only a
limited amount of air and no rain water is present, as is the case in living
trees with no open wounds reaching to the diseased heartwood, the rotting
wood may become honeycombed with empty, cellulose-lined, elliptical
cavities (PI. VII, fig. 3) or it may decompose into a fibrous mass consist¬
ing of long, white cellulose strands and partially decomposed vessels and
medullary rays. Large quantities of these white cellulose strands are
often found in the butts of freshly cut trees which externally appear
perfectly sound but have this rot in the heartwood.

A Pocketed or Piped Rot in Scarlet Oak

The following description of the pocketed or piped rot was made from a
wind- thrown scarlet oak ( Quercus coccinea) , which on falling split on the
upper side for 7 or 8 feet. From this fissure a sporophore of Polyporus
pilotae protruded. The rot began in the top of the tree and had reached
the ground. The tree was sawed into 6-foot lengths and split open on
March 5, and on May 30 fresh sporophores were beginning to form on the
ends of the split pieces.

In this host the fungus first attacks the spring wood immediately
around the larger vessels, turning it to a light-tan color. This change in
color is accompanied by the absorption, more or less irregularly, of the
cells of the spring wood, while the wood fibers intermixed with these
cells are delignified from within outward. The tan color of the affected
areas is due to the walls of the wood fiber and other cells adjacent to the
vessels turning a golden yellow. At this stage of the rot the spring
wood is badly decomposed and consists of cells and vessels much eroded,
leaving fragments of both intermixed with apparently unchanged cells
and vessels. This partial destruction of the spring wood causes it to
separate readily into circular sheets along these lines of weakness.

The next stage of the rot going inward toward the center of the tree
is the almost complete change of the summer-wood fibers and tracheids
into a yellowish white cellulose. Under the microscope the rotten wood
is seen to consist of delignified wood fibers intermixed with the remnants
of the spring wood and of nearly unchanged medullary rays, while the
entire mass of rotted wood is ramified by large, colorless, thick-walled,

ii6

Journal of Agricultural Research

. Vol. I, No. a

much-branched fungus hyphae 5 to 10/z in diameter. These hyphae
are especially abundant in the spring wood. In this stage the rotten
wood easily pulls loose in thin flakes, the line of cleavage being between
the medullary rays. Many white and yellowish white unabsorbed
cellulose wood fibers are found in the rot at this stage.

The third and final stage of the rot is found in the center of the tree
and is of a reddish brown color, there being a rather sharp line of demarca¬
tion between this and the light-tan color of the second stage. In this last
stage there are found remnants of the vessels, a few unabsorbed fiber
tracheids, wood fibers, and partially decomposed medullary rays inter¬
mixed with the colorless hyphae of the fungus. Not enough hyphae are
present, however, to bind the rotted wood into a tough mass. The
wood at this stage at first is rather brittle when dry and can be partially
crushed into fragments between the fingers, but finally it crumbles into a
brownish dirtlike mass, which remains in a cavity thus formed inside the
tree, unless removed by squirrels, etc. On the split surface of the rotting
wood which was exposed directly to the air and rain water a dark, reddish
brown mycelial layer of a gelatinous nature was found. This gelatinous
mass might, of course, be a foreign growth and not a part of the myce¬
lium of the fungus Polyporus piloiae. The reddish cast is due to the
formation of reddish brown bodies on or among the hyphae ; sometimes
several of them form a conidialike chain.

In general, the delignification seems to begin in the layer of wood fibers
forming the boundary line between the summer growth and the spring
layer of wood formed the following year and spreads most rapidly in the
spring wood, leaving more or less intact the largest vessels and the cells
immediately adjacent. At this stage many of the medullary rays contain
a chestnut-brown, humuslike substance.

A Pocketed or Piped Rot in the Texan Oak

The rotted wood from which the following description was made was
obtained from an old log of Texan oak (Quercus texana), just beneath a
very large sporophore of Polyporus piloiae.

The rot in this host is much like that described for the scarlet oak,
consisting of long strands of white to creamy white, cellulose fibers inter¬
spersed with the partially changed spring wood and medullary rays.
There is a zone of one-fourth to one-half inch of discolored wood be¬
tween the sound wood and the zone where the delignification is evident.

A Pocketed or Piped Rot in Chinquapin

This description of the pocketed or piped rot was made from material
obtained from a fallen log of chinquapin (1 Castanea pumila) on which a
sporophore of Polyporus piloiae was found. The rot was seen a number
of times in fallen chinquapin trees in the Ozark National Forest. In

Nov. xo, 1913

Heart-Rots of Hardwood Trees

117

living trees of this species, as in the white oak, the rot may vary
somewhat.

In the chinquapin the fungus first delignifies the latest formed summer-
wood fibers, those immediately adjacent to the large vessels, and spreads
finally to all the wood fibers lying between the spring wood of any two
successive . years. As the summer wood is composed largely of wood
fibers, the ultimate result is an almost complete separation of the layers
of spring wood. The concentric layers of the spring wood are separated
at first by the white to yellowish white, cellulose fibers. Tater this cellu¬
lose is entirely absorbed, leaving only the concentric layers of the spring
wood loosely held together by the remnants of the wood fibers and the
few small vessels found in the summer wood (PI. VII, figs. 4, a, and 4, b).
The vessels and other cells of the spring wood have in the meantime
become more or less corroded and have assumed a reddish brown color.
In the final stage of the rot the wood when dry is brittle and can be
easily broken between the fingers. In old, weathered chinquapin logs
attacked by this fungus the rot is very characteristic, consisting of con¬
centric layers of rotten wood which are so loosely held together that
one can easily pull off layer after layer.

A Pocketed or Piped Rot in Chestnut

The material examined for the following description of the pocketed
rot was obtained from the diseased wood of living chestnut trees ( Castanea
dentata) located near New Berlin, N. Y. In the hollow butts of these
trees the resupinate form of Polyporus pilotae was found. Some trees
were examined which had recently been made into railroad ties. Ample
opportunity was thus given for a thorough study of the various stages
of the rot in different regions of the tree trunks.

The first indication of the rot is a watery brownish discoloration of the
heartwood. In cross section this discolored area or “ soak” often appears
as a central circular patch (PI. VII, fig. 5), often flanked by one or more
very narrow crescent-shaped discolored areas, lying between the diseased
portion and the sap wood, or sometimes the “soak” maybe eccentrically
placed in the heartwood of the tree. These rings of diseased wood follow
very closely certain annual rings and usually appear first in the immediate
vicinity of the large spring vessels. Sometimes only one annual ring
will show the disease, and this may extend for several feet longitudinally
in the tree beyond that portion of the rot where delignification is evident.

The mycelium of the fungus travels much more rapidly longitudinally
in the tree than radially. It is first seen in the large spring vessels. The
adjacent wood fibers soon show signs of delignification, which usually
occurs most abundantly in the latest formed summer wood, where small,
irregular, oval patches of cellulose are produced. These patches usually
lie opposite the largest vessels and immediately adjacent to them. This
association of cellulose and large vessels is especially noticeable in cross

Journal of Agricultural Research

Vol. I, No. 2

118

section, where the delignified areas may usually be seen in the summer
wood. The delignification may continue without much absorption of
the cellulose till long white bands of cellulose are found lying alongside of
the vessels. This formation of bands of cellulose is especially marked
when an abundance of air and rain water can penetrate the rotting wood.
Such a condition obtains in fallen logs with large hollows or cracks in them.

If, on the other hand, the rot is in the center of the heartwood of a
living tree, the small, oval-shaped cellulose patches increase in size,
hyphae from the adjacent vessels gradually absorb the cellulose until
lens-shaped cavities are formed which at first are filled by a dense growth
of rather coarse hyaline hyphae. The sides of these cavities ar£ lined
with the projecting ends of the delignified wood fibers much like the rot
produced by Trametes pini. Later both the hyphae and the cellulose
lining may disappear and leave an empty cavity, thus producing a
pocketed or honeycomb type of rot.

In the earlier stages of the rot the diseased heartwood surrounding the
white cellulose patches is of a cinnamon color. The wood at this stage
of the rot is rather firm, contains small cellulose patches (PI. VII, fig. 6),
and has vessels filled with colorless hyphae from 6 to io//, or even less,
in diameter. The white, cellulose, oval areas gradually encroach upon
the summer wood till they extend from one annual layer of vessels to the
next. By this time much of the cellulose has been absorbed, and small,
distinct cavities are formed. At this stage of the rot the diseased wood is
much lighter in weight and can easily be broken into pieces between the
fingers. Finally, a condition is reached in which the reddish brown
rotten wood is very loosely held together and tends to split up into con¬
centric sheets corresponding to the annual rings. Short oval holes run¬
ning radially through two or three annual layers of wood are also common
at this stage. In rare cases the cells surrounding the vessels are com¬
pletely absorbed, while the summer- wood fibers are delignified without
the formation of cavities. Many of the trees attacked by this fungus had
hollows in them, but whether the hollow was caused by this fungus or by
a subsequent attack of another fungus, as Hydnum erinaceus , could not
be determined. While this rot is a butt rot of the chestnut, it is also able
to enter through dead limbs and thus produce a top rot. The rot when
it enters by means of a dead branch follows the heartwood of the branch
down to its juncture with the heartwood of the tree. The fungus then
travels both upward and downward in the bole of the tree (PI. VIII,

i).

Of the chestnut trees in the region examined around New Berlin,
N. Y., fully 75 per cent had tops attacked by this fungus. This large
percentage was probably due to numerous dead limbs on each tree, thus
affording the fungus ample opportunity to enter the tops. Of 302 felled
chestnut trees which were studied by the writer in this region 119, or
39.4 per cent, had this rot in the butts. This large percentage of infection

Nov. 10, 1913

Heart-Rots of Hardwood Trees

119

was mainly due to the fact that practically all of the trees came from a
coppice growth, and if the original stump was diseased, the later generation
of trees springing from its base were also infected through their union
with the old diseased stump. Officials of the Unadilla Railroad claim
that chestnut ties having only a small amount of this rot in their centers
last only three to five years when placed in their roadbed.

This rot in the chestnut is apparently identical with the piped rot of
chestnut described by Von Schrenk and Spaulding.1 Their description
of the piped rot of the oak in the same publication apparently includes
two distinct rots; viz, this rot caused by Polyporus pilotae and the
common heart-rot of the oak caused by Polyporus dryophilus2 which is
also a piped rot in one of its stages and will be described in a later
publication.

Resui/ts op Investigations op the Pocketed or Piped Rot

The most common and constant characters of this rot, taking all the
hosts into consideration, are the presence of long, continuous strands of
cellulose, the delignified wood fibers and fiber tracheids, and the white-
lined pockets so common in the living oak and chestnut in the early
stages of the rot. In the white oak the changing of the wood fibers into
cellulose is not so complete as in the other hosts, so that the wood is not
broken down as much. In both white oak and chestnut there are holes
which run tangential to the tree through the spring wood or radially from
one annual ring to another. This condition is especially noticeable in
the older stages of the rot in the butt of the trees and in the vicinity of
freshly formed sporophores of the fungus.

Sporophores of Polyporus pilotae were formed on living white oaks, on
the ends of white-oak logs cut only one month, on old logs which evidently
had been cut for several years, on a standing fire-killed yellow oak
(Quercus velutina ), on a fallen and very rotten log of Texan oak (Q.
texana)f on the trunk of a wind-thrown scarlet oak (Q. coccinea ), on old
dead logs of chinquapin ( Castanea pumila) , on the inside of a hollow in a
living chinquapin tree, and on chestnut trees (C. dentata). In the last
instance the sporophores were resupinate and growing in the hollow
butts of the living trees. Of the 302 chestnut trees studied in New York
1 19 had this rot. The average diameter of the rot per tree was 6.5 inches,
the average diameter of the stump 16.6 inches, and the average height of
the rot per tree was 5.4 feet. The maximum diameter and height of the
rot in any one tree was found in a tree 27 inches in diameter. The diam¬
eter of the rot in this tree was 20 inches and the height of the rot was 20
feet.

1 Schrenk, Hermann von, and Spaulding, Perley. Diseases of Deciduous Forest Trees. Bur. Plant In¬
dus., U. S. Dept. Agr., Bui. 149, p. 39, 1909.

2 Hedgcock, George G. Notes on some diseases of trees in our national forests. Phytopathology, v. a,
no. 2, p. 73, 74, Apr., 1912.

120

Journal of Agricultural Research

Vol. I, No. a

A comparison of the average height of this rot in the chestnut (5.4 feet)
with its average height in the white oak (3.9 feet) shows that it extends
higher up the bole in chestnut than it does in white oak. This difference
is still further accentuated by the difference between the average diameter
of the diseased chestnut trees (16.6 inches) and that of the diseased
white oak (25.6 inches). The average age of the chestnut was probably
not over 100 years, while that of the white oak was about 250 years. The
very large and numerous vessels in the chestnut made it possible for the
fungus to travel to greater heights in this wood in a given time than it
could in the white oak, which is a much denser, slower growing wood. Of
course, the amount of rainfall and other environmental factors would
have to be taken into consideration when comparing the relative heights
of this rot in the chestnut and oak.

On the same area in New York where the chestnut mentioned above
was studied, a record was made of 477 felled white oaks. Of this number
only 4, or less than 1 per cent, had the piped rot so common in the
chestnut. Its average height in these 4 trees was 3 feet, its average
diameter was 8 inches, and the average diameter of the affected trees
was 1 5 inches. This small percentage of infection was probably due to
the fact that no fires had been allowed in these woods and therefore
practically no opening into the heartwood of the trees was offered and
to the further fact that the oaks did not originate from a coppice growth.

On an area in Virginia which had been in timber for about 60 years the
writer checked the stumps of 565 chestnut trees which had been recently
cut. The majority of these trees originated from sprouts and had made
a vigorous growth, the average age of the trees being about 50 years. Of
the 565 chestnut trees only 18, or 3 per cent, had piped rot in the butts.
Of this same area 201 white-oak stumps were also checked, of which
number 13, or 6 per cent, had piped rot in the butts. This area was an
old abandoned field which had been used as a pasture for many years
and, so far as the writer could ascertain, had not been burned over in
50 years.

The rate of growth of the various rots in individual trees, as shown by
the records made in the Ozarks, varies greatly. For instance, Polyporus
sulphur eus had been in one white oak 200 years and had made a growth
in height of only 6 inches during that time, while the same fungus had
been in another white oak for 50 years and had made a growth in height
during that time of 3 feet. A similar wide range in growth is found for
the rot produced by P. pilotae in white oak, where it was in one tree for
280 years and had made a growth in height of only 6 inches, while in
another white oak the same fungus had made a growth of 4 feet in only
60 years. However, taking into consideration the average and maximum
height of each of these rots and their average rate of growth in a tree, it
is evident that they do not grow with any thing like the rapidity — at
least in white oak— that might be expected.

Nov- io, 1913

Heart-Rots of Hardwood Trees

121

Of the 1,938 white oaks studied in the Ozarks 408 trees had this rot.
The average diameter of the rot per tree was 13.7 inches, the average
diameter of the stump was 25.6 inches, and the average height was 3.86
feet. The maximum diameter and height of the rot in these trees was
found in a tree 400 years old. The diameter of the tree was 40 inches,
the diameter of the rot was 36 inches, and the height of the rot was 24
feet. The oldest rot was 280 years and was found in a tree 310 years old.
The average age of the rot in 92 trees was 156 years. The average rate
of growth of the rot was 1 foot in height and 3.5 inches in diameter for
every 40 years of time. The youngest white oak found with this rot
was 180 and the oldest 400 years old.

The exact range of this fungus is not known. It is very common in
oak and chinquapin in the Ozark National Forest and has been found in
Virginia on scarlet oak.

The writer has also examined authentic sporophores of this fungus on
the following hosts and from the following localities: ‘

“On underside of log” (resupinate sporophore), Pennsylvania; “on
log,” North Carolina; “in hollow oak log,” Ohio; “on rotten oak
log,” Indiana; “on underside of old log” (resupinate sporophore),
West Virginia; “on dead oak logs,” New York; “on oak,” North Caro¬
lina; from Iowa, no host given; “on punky chestnut log,” no locality
given; from Florida, no host given; from South Carolina, no host given;
from Tennessee, no host given; “on end of log,” Canada; “on oak,”
Canada; “on old logs,” Canada. Three specimens were also seen from
Europe, where it is known as Polyporus croceus (Pers.) Fries: “On
living oak,” Sweden; “on old oak and chestnut,” apparently from
France, no locality given; and “on old oak,” locality not given. It
probably occurs east of the Rocky Mountains in the United States on
oak, chinquapin, and chestnut wherever the hosts grow and also in
Europe on oak and chestnut. It is by far the worst heart-rot found in
chestnut timber, occurring in this host as both a butt and top rot. It
stands second in destructiveness to white-oak timber in the Ozark
National Forest, both as to number of trees infected and height attained
in the tree. Hydnum erinaceus is the most destructive heart-rotting
fungus of the oak found in the Ozark forests (see Table I, p. 111). The
rot caused by P. pilotae was found associated with the rot produced by
Hydnum erinaceus in 105 trees, with string and ray rot in 3 trees, with
Polyporus sulphurous rot in 8 trees, and with both Hydnum erinaceus and
Polyporus sulphur eus rots in 5 trees.

The sporophores of P. pilotae were in an actively growing stage during
the month of September in the Ozark National Forest. This fungus
usually enters the oak at the base of the tree, probably through fire
scars in most instances. The rot was also found occasionally in the
upper part of the tree, while the base was not infected. The fungus,

13000 0 — 13 - 3

122

Journal of Agricultural Research

Vol. I, No. 2

therefore, can enter the tree through fire scars in the butt and also
through broken branches or other wounds on the bole and in the top of
the tree. There is also a honeycomb rot in oak and in chestnut caused
by a species of Stereum. This honeycomb rot in its earlier stages resem¬
bles so closely certain stages of the rot caused by P. pilotae that it is
very difficult to determine which fungus produced the rot, unless the
sporophores are present.

A STRING AND RAY ROT OF OAKS CAUSED BY POEYPORUS BERKELEYI

The inittel stage of the string and ray rot in the white oak when seen
in a radial longitudinal section is characterized by the presence of large
amounts of cellulose tissue, causing the rotted wood to have a yellowish
white appearance. This stage of the rot may extend for 4 to 8 inches
longitudinally, when it terminates rather abruptly in apparently sound
wood. The cellulose tissue is composed exclusively of delignified wood
fibers, which constitute the bulk of the summer wood. The middle
lamellae have entirely disappeared, so that each delignified wood fiber is
separate from its neighbor.

The next stage of the rot is the rather rapid and complete absorption of
these delignified fibers, leaving both the spring and summer vessels, the
cells immediately adjacent, and the medullary rays intact. The rot at this
stage is most characteristic, consisting of a rather dry mass of medullary
rays interwoven with long, flat strings of wood (PI. VIII, fig. 2). These
strings are sometimes 8 to 10 inches long by one-sixteenth of an inch
wide and consist of the vessels held together by the unabsorbed adjacent
cells. The rot in this stage is reddish brown and on account of its peculiar
and characteristic structure has been named the “string and ray rot”
by the writer. This second stage of the rot may extend from a few
inches to several feet up the tree. At first the flattened strings of wood
are rather tough, but this gives place to a condition in which the strings
get brittle and can be crumbled between the fingers into a brownish,
coarse powder. Finally the entire mass of rotting wood becomes over¬
run with a colorless mycelium. In this condition the rot is very moist,
almost wet, and consists of fragments of vessels and of the medullary
rays, interwoven with the colorless hyphae of the fungus. It can now
be compressed with the hands into rather firm balls which may be thrown
with force and yet will not break into pieces.

Finally the entire mass of rotted wood and mycelium gradually disap¬
pears till a hollow is left in the base of the tree. Over the surface of this
vanishing mass brittle white or creamy white layers of mycelium are
formed, on the undersides of which are cottony masses. Shakes, checks,
or worm holes in the wood may have a slight mycelial felt in them.

The string and ray rot seems to be one of the very few heart-rots of the
white oak capable of the complete absorption of the heartwood of the
tree, thereby producing hollows. The slow rate of travel upward in the

Nov. io, 1913

Heart-Rots of Hardwood Trees

123

tree compared to its radial rate of growth and the subsequent rather
complete absorption of the entire heartwood in the stool of the tree
produce a peculiar condition when the tree is cut. A tree in which this
rot has reached its last stages in the stool will be rotted to or nearly to
the sapwood for 1 to 3 feet from the ground, and such a tree will fall as
soon as the thin shell of sound wood is severed, carrying with it the par¬
tially rotted heartwood, which easily pulls loose from the badly rotted
mass in the stool. The butt end of the felled tree will then have attached
to it a cylinder of rotted wood some 1 to 2 feet long in the string and
ray stage, thereby leaving a hollow stump in the bottom of which there
will be the wet, very rotten mass of wood held together by the threads
of mycelium.

This rot has a very strong but pleasant odor, somewhat like that of
anise oil. This odor disappears after the exposure of the rot to the air
for several weeks, but is so marked when the tree is first cut that it can
be detected at a distance of from 20 to 30 feet.

Studies were made of 1,938 white-oak trees which were cut for staves.
Of these, 57 had this rot. The average diameter of the rot in these 57
trees was 19 inches; the average height per tree was 3.5 feet; and the
average age per tree was 280 years. The maximum diameter and height
for this rot in any one tree were found in a tree 380 years old. The
diameter of the rot was 32 inches and the height was 13 feet. As a rule,
this rot does not extend very high in a tree, as compared to its extent in
diameter, and ends very abruptly in perfectly sound wood. It was
also found in the butts of two black oaks (Quercus velutina) ; the sporo-
phores of the fungus were seen several times on the roots of both white
and black oaks which had not been felled. The writer repeatedly found
from one to three sporophores of Polyporus berkeleyi (PI. VIII, fig. 3)
attached to the roots of the trees in which this characteristic heart-rot
was present. The direct connection of the rot in the stump with the
sporophore could easily be traced by following the rot down into the
stool and thence through the rotted heartwood of the root to the
sporophore. This was done in the case of at least a dozen trees.

The youngest tree found with this rot was 170, the oldest 500 years
of age. The rot was usually found in mature and overmature trees
from 25 to 32 inches in diameter which grew in rich soil on north slopes.
In 6 of the stumps of the 57 white oaks found affected with this rot some
evidence as to the age of the rot was obtained. The oldest rot was 380
years and was found in a tree 420 years of age. The average age of the
rot in these six trees was 190 years. The average rate of growth of the
rot was 1 foot in height and 5.4 inches in diameter for every 60 years
of age. The fungus producing this rot usually enters the tree through
some wound at the butt, such as fire scars. No evidence was found
that it could enter through broken branches. In no instance was the

I24

Journal of Agricultural Research

Vol. I, No. 2

rot found in the top of a tree. It originates at the butt and travels
upward in the heartwood of the tree.

Of the sporophores of Polyporus berkeleyi found by the writer all
occurred at the base of oak trees, either plainly growing from the exposed
root or on the ground near the base of the tree. In the latter case a
careful examination of the basal portion of the sporophore showed that
it was attached to the roots of the tree. The writer has never found
it growing on the bole of the tree above the surface of the ground, though
it is not impossible that it could grow as brackets on the trunk, but it
is doubtful if it does. P. sulphur eus Fr. and P. schweinitzii Fr., two
closely related polypores which produce heart-rots in living trees, are
often found growing on the roots at the base of the diseased trees as well
as on the boles proper.

There was no evidence to indicate that the fungus could fruit on the
trunk after the trees were felled, even if the rot should continue to
grow in the felled tree. A small sporophore was found at the base of
a 20-foot white-oak snag, while a large sporophore was found at the
base of a dead standing white oak, indicating that the fungus could
continue to grow and fruit after the trees were dead. The only external
evidence that trees are attacked by this heart-rot is the presence of the
sporophores of the fungus on the roots. Sometimes the base of the
diseased tree is slightly “ swell butted.” This last character, however,
is common to trees attacked by other butt-rots.

This rot was found associated with the rot produced by Hydnum
erinaceus in 7 trees, with the pocket rot caused by Polyporus pilotae
in 3 trees, and with the rot produced by P. sulphureus in 1 tree.
Hydnum erinaceus was repeatedly found attacking and completely
destroying wood previously rotted by the following fungi: Polyporus
berkeleyi , P. pilotae , Fomes everhartiij Polyporus hispidus , P. jrondosus ,
and P. dryophilus , but no evidence was found of its attacking the rot
produced by P. sulphureus , although it was found associated with
this rot in the same tree. Fresh sporophores of P. berkeleyi were com¬
mon during the latter part of August and probably could be found
during September. No fresh sporophores were seen in December.

The writer has also examined authentic material of Polyporus berkeleyi
on the following hosts and from the following localities: “At base of
white oak,” Canada; “on roots of living white oak,” Missouri; from New
York, West Virginia, and Missouri no host was given; “from dead place
near ground in living oak,” Pennsylvania; “on base of stump,” North
Carolina; “on oak,” New York; “on chestnut,” New York; “at base
of tree,” Ohio; “at base of ash stump,” Ohio; “at base of oak stump,”
Pennsylvania; from West Virginia, Pennsylvania, Ohio, North Caro¬
lina, and Canada no host was given; “near roots of large oak,” Canada;
and “under oak,” Massachusetts. Apparently this fungus is found only
in America. The writer has never seen it growing on anything but oak,

Nov. 10, 1913

Heart-Rots of Hardwood Trees

125

but from the above record it also occurs on chestnut and on ash, while
Dr. Weir, of the Office of Investigations in Forest Pathology, reports
it on larch in 1913.

From the studies made in the field the writer finds no proof of the
ability of this fungus to grow permanently as a saprophyte in humus
and decayed forest litter. All sporophores seen certainly grew from
mycelium inside the living, diseased trees at whose base they were
found and not from mycelium ramifying in and drawing nourishment
from the soil or leaf litter.

Weir reports 1 the finding of sporophores of Polyporus berkeleyi
attached to the roots of the larch in Montana, but from observations
made in that region reached the conclusion that the mycelium ramified
in the deep forest litter and drew its food from that source as well as
from the rotten roots to which the sporophores were attached. It will
prove very interesting if this rot in the larch should prove to be similar
to that produced by this fungus in the oak, especially since the anatom¬
ical character of the wood of these trees is so different.

A STRAW-COEORED ROT OF OAKS CAUSED BY POEYPORUS FRONDOSUS

The initial stage of the straw-colored rot of the white oak (Quercus
alba) is characterized by the dissolution of the middle lamellae and the
delignification of some of the wood fibers, leaving the fibers now con¬
sisting of cellulose free from each other. (PI. VIII, fig. 4.) The advanc¬
ing line of the rot upward in the tree consists of irregular, rather indefinite
white patches, conforming more or less in size and shape to the largest
medullary rays, or of narrow white bands projecting into the sound
wood. Five or six inches below the boundary line between the advanc¬
ing rot and sound wood the color in radial sections is more evenly white,
as the patches have coalesced more or less at this stage. The unpolished
split surface feels velvety, owing to numerous projecting free ends of
the cellulose fibers. A tangential view of the advancing line of rot
shows a whitish surface consisting of white delignified fibers interspersed
with unchanged medullary rays and unchanged or only partially deligni^
fied vessels and their immediate adjacent tissue. In cross section the
rot has a whitish cast surrounded by the natural color of sound heart-
wood.

The amount of delignified tissues in the earlier stages of this rot is
much less than that found in the earlier stages of the string and ray
rot. Eight to twelve inches behind the advancing point of the rot numer¬
ous colorless hyphae are found in the larger vessels. At this stage
in the rot some of the delignified tissue has been entirely absorbed.
The delignification and absorption begin with the inner layer of the
wood fibers and proceed centrifugally, so that the lumen of the cell

1 Weir, J. R. Some observations on Polyporus berkeleyi. Phytopathology, v. 3, no. 2, p. 101-103, pi. 9,
1913.

126

Journal of Agricultural Research

Vol. I, No. 2

gradually increases in size as the rot progresses. Marked delignifica-
tion occurs in the tracheids and cells immediately adjacent to the larger
vessels in which the fungous hyphae are found, but the medullary rays
and walls of the large vessels are still strongly lignified, as are also the
numerous tyloses seen in these vessels. The walls of the tyloses were
punctured in many places by the fungous hyphae. Six to eight inches
farther down, or 18 to 24 inches behind the advancing line of the rot,
the rotted wood is soft and spongy to the touch and is of a straw color.
In this stage the rotted wood consists of partially changed medullary
rays, some unchanged wood fibers, and vessels with fragments of these
in various stages of absorption, all strongly permeated with fungous
hyphae. Some medullary rays are still intact, while others have their
outer radial cells either partially or entirely delignified and absorbed,
so that in pulling apart the rotted wood tangentially, the medullary
rays often pull out, leaving holes in one piece similar in size and shape
to the rays, while the rays themselves remain attached to the other
piece of the rotted wood.

The final stage of the rot differs but little from this condition, since there
are still portions of all the elements present either unchanged or only
partially changed. The rotted wood is rather tough and can be bent and
twisted without breaking if taken in pieces 12 to 18 inches long and 4 or 6
inches thick. It is rather soft and spongy, but the fungus apparently
never completely disorganizes the wood, thereby producing hollows.
On weathering for two or three months the rot in the tops of the stumps
and in the ends of the rejected butt cuts turns reddish brown and becomes
firmly agglutinated, a condition so characteristic of this rot that one could
identify the rot by this feature alone, without the presence of sporophores.

The rot has no odor. A section through the reddish discolored wood
shows an abundance of light-brown hyphae. The remnants of the remain¬
ing lignified tissues are also colored light brown. In a freshly cut stump
which had this rot it would be hard to identify the rot in a cross section.
Even when the wood is split open, there are no very pronounced
macroscopic characters to distinguish it, like the string and ray stage
of the rot caused by Polyporus berkeleyi.

The following is a brief description of the gross appearance of this rot
caused by Polyporus frondosus , made as soon as the tree was cut.

The rot seen in a radial longitudinal view consisted of long white lines
advancing 6 to 10 inches beyond the more completely rotted wood below.
These lines apparently were caused by the fungous hyphae following the
vessels in certain annual rings. There was a watery reddish discoloration
or “soak” about 2 inches in advance of the white lines. The older rot
was of a light-tan or straw color and with a slight mycelial weft in checks.
Some 2 to 6 inches below the upper end of the white lines, white downy
masses of mycelium could be seen by the aid of a hand lens in the large
spring vessels situated in the white lines. In cross section the rot had a

Nov. iot 1913

Heart-Rots of Hardwood Trees

127

coarse, fibrous surface, due to the stiff unabsorbed ends of the vessels,
partially isolated by the absorption of the wood fibers and the sub¬
sequent tearing apart by the saw when the tree was felled. This fibrous
character was not evident except where the tree was sawed.

This rot was identified in only 12 trees out of the 1,968 white oaks
examined. No idea was obtained as to its age in a tree, as all of the trees
found affected by it had open scars at their bases. It was apparently
through such scars that the fungus entered the tree. Sporophores of
Polyporus frondosus were found attached to the roots of 6 of the trees
in much the same manner as those of P. berkeleyi , usually on that side of
the tree which had the fire scar. The average height in the 12 trees
attacked was 2.3 feet, the average diameter of the rot 12 inches from the
ground was 14 inches, and the average age of the trees attacked was 270
years. The minimum age of the trees attacked was 220 years and the
maximum age was 340 years. The maximum diameter of the rot in a
tree was 24 inches and the maximum height in the tree was 4 feet.

The only external evidence of this rot in a tree was the presence of the
sporophores attached to the roots of the diseased tree. The connection
between the attached sporophores and the heart-rot in the tree was
easily established in every case. This fungus may not continue to grow
in the diseased trees after they are cut, for no sporophores have been
found on felled trees nor have any been reported as occurring on logs. It
seems to be strictly a butt-rot, as no evidence is known to the writer
of its occurrence in the tops or on the branches of trees. One tree was
found in which this rot was associated with the rot produced by Hydnum
erinaceus. The writer has also found sporophores of P. frondosus on the
roots of Quercus digitata at Arlington, Va., and has examined authentic
herbarium material of the plant on the following hosts and from the
following localities: “In evergreen woods,” Canada; “under oak,”
Massachusettes ; “at base of oak,” Massachusetts; “at base of red oak,”
New York; from Ohio, no host given; “on old stump,” Ohio; “at roots of
fallen oak,” Ohio; “at roots of oak,” Maryland; “on dead trunks,
lAceris negri,’ ” Missouri; “on roots of chestnut,” Germany; “on roots of
chestnut,” Italy; “on Castanea vesca” France; “at base of large oaks,”
Saxony; “at base of trunk,” Italy; and “on roots of chestnut,”
Bohemia (?).

This fungus, which has been known to mycologists for many years,
is represented in nearly all the more complete lists of European fungi.
It is evidently very widely distributed, inhabiting frondose woods in
North America and Europe, in direct association with oak and chestnut
trees.

The writer is under many obligations to the officers in charge of the
New York Botanical Garden for the many courtesies extended to him
while there, and to Dr. W. G. Farlow for free access to the Cryptogamic
Herbarium of Harvard University.

DESCRIPTION OF PLATES

Plate VII. Fig. I. — Polyporus pilotae ; A sporophore on the end of a white-oak log
from Arkansas. Photograph made 43 days after tree was felled.

Fig. 2. — Polyporus pilotae: Rot appearing in the butt of a white-oak
log from Arkansas, showing the holes and white cellulose areas char¬
acteristic of this rot in a cross section of a living oak.

Fig. 3. — Polyporus pilotae: Radial-longitudinal view of a white-oak
log from Arkansas, showing the honeycomb type of the rot with the
white cellulose lines and elliptical hollows.

Fig. 4. — Polyporus pilotae: Rot occurring in a log of Castanea pumila
from Arkansas; A , concentric layers of the rotted wood; B, white
cellulose fibers.

Fig. 5.— Polyporus pilotae: Cross section of a chestnut log from New
York, showing the central circular rotted zone.

Fig. 6. — Polyporus pilotae: Radial-longitudinal view of the rot in a
chestnut log from New York, showing the white pocketed stage.

VIII. Fig. 1. — Polyporus pilotae: Radial-longitudinal view of the rot in a
chestnut log from New York. This rot enters at a dead branch and
then moves down the heartwood of the branch into the trunk.

Fig. 2. — Polyporus berkeleyi: Radial-longitudinal view of the rot in
white-oak timber from Arkansas, showing the string and ray form char¬
acteristic of its second stage.

Fig. 3. — Polyporus berkeleyi: A sporophore on a white-oak root from
Arkansas.

Fig. 4. — Polyporus frondosus: A sporophore on roots of white oak from
Arkansas.

(128)

Plate

•Rots of Hardwood Tr

INDIVIDUAL VARIATION IN THE ALKALOIDAL CON¬
TENT OF BELLADONNA PLANTS

By Arthur F. Silvers,

Chemical Biologist , Office of Drug-Plant and Poisonous-Plant Investigations , Bureau

of Plant Industry

INTRODUCTION

It has long been recognized that a necessity exists for the improvement
of the important medicinal plants. Within recent years agricultural
science has been largely concerned with the improvement of crops by
the application of the methods of plant breeding, but thus far practically
no attempts have been made to extend these methods to drug plants
with a view to improving their medicinal qualities. The chief aim in
applying such methods should be to increase the active medicinal con¬
stituents rather than to improve the appearance of the plants. That
the amount of a chemical constituent in a plant can be favorably modified
by selection has been amply proved by work which has been done on the
sugar beet, and there is reason to believe, therefore, that similar efforts
with the economically important medicinal plants will be attended with
success.

One of the first steps necessary to inaugurate such a plan is to deter¬
mine the variation of the active constituents in individual plants and the
extent to which such variation is influenced, if at all, by the various
factors affecting the growth and cultivation of the plants. This article
deals entirely with such a study. The results herein set forth furnish a
basis for the application of the principle of selection as the next step in
the solution of the problem.

Atropa belladonna was selected as a suitable plant with which to work,
since it is probably the most important of the group of solanaceous
plants which depend for their therapeutic action on mydriatic alkaloids.
The supply of this plant in the wild state is largely exhausted and future
supplies must necessarily depend on cultivation. The alkaloids which it
contains can be definitely determined by chemical assay, which is a dis¬
tinct advantage in a problem of this kind.

The writer wishes to emphasize the fact that the work thus far done
constitutes but a preliminary step toward the application of the methods
of selective breeding, which has already been begun. Considerable
interest attaches to the results presented in this article because they
represent the first extensive study of the variation of the quantity of
alkaloids in this important economic plant.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(129)

Vol. I, No. 2
Nov. 10, 1913
G — 3

130

Journal of Agricultural Research

Vol. I, No. 2

VARIATIONS IN THE ALKALOIDAL CONTENT OF EEAVES OF DIFFERENT

BELLADONNA PLANTS

METHOD OF INVESTIGATION

The object of this investigation was to study the variation in the
alkaloidal content of the leaves of individual belladonna plants and to
establish, if possible, some correlation between the appearance of the
plant and the variation of active constituents, should any variation exist.
It was decided that the plants to be used for this purpose should be
selected entirely at random, and they were therefore taken from all sec¬
tions of the plat without reference to size or thriftiness. This afforded
an opportunity to study the relationship of growth to alkaloidal content
of the leaves. The field work was carried on at Arlington, Va., Bell,
Md., and Madison, Wis. The Arlington plat was the largest, and the
large number of plants at that place furnished the most complete data.

The general plan followed was to pick the leaves from each plant at
different times during the growing season so as to be able to determine
the proper time of the year in which the leaves should be picked in order
to insure the greatest percentage of alkaloids. This should have the
further advantage of showing whether individual plants which contain
an abnormally high or low percentage of alkaloids in the leaves at one
time of the year possess the same feature at other times.

Unfortunately this program could not be followed the first year,
owing to pressure of other work. In some cases, especially early in the
season, the plants were too small to furnish sufficient leaves for an assay
without being entirely denuded. The smallest amount of dry material
that could be used in the assay was 2 grams, and in order to insure a
duplicate assay it would be necessary to have at least 25 grams of green
leaves. Immediately after picking, the leaves were spread out evenly
on a table in a dry, well-ventilated room until air-dry. They were then
placed in small cloth bags until assayed.

The development of this investigation has been somewhat retarded
through the loss of a number of the plants under observation. The loss
was especially severe in the lower section of the plat, where the drainage
is poor. The plants wilted suddenly and rapidly and the roots became
entirely decayed. The loss was greatest after a prolonged wet spell,
and after the trouble had once manifested itself the plants only occa¬
sionally recovered. Holmes 1 says that the cultivation of belladonna
can rarely be continued beyond the third year, as the increased weight
of the plants has a tendency to split the roots, thus permitting the water
to enter and rot them. This may possibly be the trouble encountered
here, but there is little evidence to show that the weight of the plant or
mechanical injury is responsible, as both young and old plants suffered
from the trouble.

1 Holmes, E. M. The cultivation of medicinal plants in Lincolnshire. Phann. Jour., s. 3, v. 12, Sept.
17, p. 237-239, 1881.

Nov. io, 1913

Alkaloidal Content of Belladonna Plants

131

RESULTS OE THREE YEARS' OBSERVATION

During the summer of 1909 three rows of belladonna plants were
started at the Arlington Experimental Farm directly from field sowing.
The plants made a fair growth in the late summer and fall, but did not
bear seed. The following spring they made a good growth and 24 plants
were carefully staked out for this investigation. Since the plants made
only a partial growth in the preceding year, they were considered as first-
year plants and are so referred to throughout this article. The only
picking from these plants in the first year was made in June, when most
of the plants were in full bloom, although some were bearing berries of
considerable size. Table I shows the general physical condition of the
plants and gives the percentage of alkaloids in the leaves of each indi¬
vidual plant.

132

Journal of Agricultural Research

Vol. I, No. 2

Table I. — Description of individual first-year belladonna plants and percentage of
alkaloids in the leaves of each at Arlington Experimental Farm in June , igio.

Plant

No.

Stage of growth.

Des

Height.

cription of ph

Spread.

mt.

Num¬
ber of
stems.

Remarks.

Alka¬

loids.

Inches .

Feet.

Per cent.

Not yet flowering. . . .

i by ij4

o. 645

Few flowers .

2 by 2%

. 618

. do .

3^ by 2%

. 528

Slightly past flower-

iKby 3

•495

ing.

Flowers and some

i by 3

•334

berries.

. do .

3 by 3X

Squatty .

•459

. do .

®Kby 3X

Tall .

.667

Many flowers .

iXby3

One branch tall and

• 657

erect.

Flowers and a few

i/^by 4

• 563

berries.

Many flowers .

3Xby4K

• 53b

Flowers and berries . .

3 by 3^

• 587

. do .

2 by 3K

• 603

. do .

3Xby4X

. 700

. do .

3 by 3

. 656

. do .

3 by 4

■ 555

. do .

iXby3X

• 544

. do .

2^ by 4

.485

Flowers and a few

2 by 2%

Very backward in

. 462

berries.

development.

Flowers .

i XA by i H

■2

. 440

Not yet flowering ....

i>^by 2

Poorly developed. . .

•473

. do .

i by \x/2

. do .

. ^87

Flowers and berries. .

2 bV4

. 622

Not yet flowering ....

i/<by i X

Poorly developed. . .

.412

Flowers .

i^by iX

• 503

Average .

• 547

Nov. io, 1913

Alkaloidal Content of Belladonna Plants

133

Since one picking showed such a wide range of variation among the
24 plants, 35 additional plants were staked out the following spring.
Twenty-six of these were in the same plat as the first plants, while the
remaining 9 were in a neighboring plat on practically the same kind of
soil and were separated from the others by a space of only about 100 feet.
These 9 plants are distinguished in Table II by the letter “ w.”

Table II represents the results of the second and third years. In 1911 ,
five pickings were made, extending from May 12 to October 17. At each
picking the height of each plant was measured, until the full stage of
development had been reached. At the first picking, on May 12, nine of
the plants were not sufficiently advanced to furnish samples of leaves.
Some of the more advanced plants were beginning to have flowers. On
May 22, when the second picking was made, the plants were all in the
full flowering stage. The third picking was made on June 17, when the
flowering was mostly over and the berries generally were well developed.
The plants had made considerable growth since the previous picking,
but by the latter part of June they had reached their maximum growth.
At the time of the fourth picking, September 6, they had assumed their
characteristic late-summer and fall appearance. The berries were ripe
and the leaves were small and sparse. At this stage the picking of the
leaves is a very tedious process. Later in the fall, after the berries are ripe,
new leaves begin to appear on the plant. Many of them develop on the
new sprouts which mature during the summer, and not a few appear
as the result of suckers which sprout directly from the roots. It has
frequently been observed that some plants develop so many of these
suckers that they have the appearance of plants just before flowering.
At this stage, October 17, the fifth and last picking was made.

Table; II. — Description of belladonna plants and percentage of alkaloids in the leaves of each at different stages of growth in ign and 1912.

134

Journal of Agricultural Research

Vol. I, No. 2

Nov. io, 1913

Alkaloidal Content of Belladonna Plants

135

136

Journal of Agricultural Research

Vol. I, No. 2

In 1912 the same line of observation was followed in connection with
the same plants and the results are also included in Table II. Unfor¬
tunately, the disease described elsewhere killed fully one-half of the plants
by the end of the season. Therefore, the results given in the table are
not as complete as those of the previous year, especially with regard to
the fourth and fifth pickings. The stages of growth at which the pickings
were made correspond closely to those of the previous year, as the dates
indicate.

At the drug-testing garden at Bell, Md., where the soil is quite different
from that at Arlington Experimental Farm, 19 individual plants were
under observation and three pickings of leaves were made. Owing to a
delay, no picking was made at the time of the first picking at Arlington,
although the plants at both places were at the same stage of development.
Consequently, the picking on May 27, which is designated as the first at
Bell, corresponds to the second picking of the Arlington plants. Table
III shows the results.

Table III. — Description of individual belladonna plants and percentage of alkaloids in
the leaves of each at different stages of growth , at Bell , Md., in IQII.

Plant No.

Description of plant.

Alkaloids (per cent).

Number
of stems.

Height (inches).

May 27.

June 22.

First
picking
May 27.

Second
picking
June 22.

Third
picking
Oct. 17.

Average

for

season.

0. 329

O. 288

0. 422

0. 346

*474

. 502

•395

*457

■ 485

. 408

. 64I

• 5ii

• 639

.686

• 57°

* 632

• 659

■ 637

.415

• 57°

•577

■ 637

• 559

. 624

• 654

. 722

. 482

. 619

.467

.464

. 466

. 526

• 59S

• 350

•477

• 752

. 600

. 418

• 590

* 571

•485

• 579

• 545

■ 548

.424

.486

•695

•587

• 75°

.677

.407

• 605

. 506

• 43*>

■ 448

.442

. 466

•390

• 511

* 456

.823

.665

• 527

• 675

• 754

. 689

* S°2

.648

. 782

•783

• 556

• 707

Nov. io, 1913

Alkaloidal Content of Belladonna Plants

*37

At the drug- testing garden at Madison, Wis., observations similar to
those at Arlington have been made for two years, and the results are
given in Table IV. The first nine plants were under observation in 1911
and 1912, while the last eight were sent to Madison from Arlington as
young seedlings in the spring of 1912. No notes were taken of the indi¬
vidual plants with regard to height, spread, and number of stems, since
they were all very much alike. Each plant acquired a height of about
2 feet and had an average of three or four stems each.

The stages of growth at which these pickings were made correspond
closely to the first, second, and third pickings at Arlington, irrespective
of the dates. Since Madison is farther north than Washington, the plants
came up later in spring than in the vicinity of Washington and did not
reach the full flowering stage until in July or early in August.

13000"— 13 - 4

Table IV. — Description of individual belladonna plants and percentage of alkaloids in the leaves of each at different stages of growth at Madison t

Wis., in ign and igi2 .

Journal of Agricultural Research

Vol. I, No. 2

138

Aver¬

age al¬
kaloids
for two

seasons.

Per ct .

00 On O t Nt « On

O to N H 00

^ 'J t 4 t Nf 't

d .

*454

Season of 1912. j

■ *■< j

ijil

VO to to « « W N

fO 00 O t- !-t N VO

V tO V K) to »o ^

d .

OCOwNTfl^WO
NONOOt-iOtO'tto
toVrOTCTj-N to V

i-c.

t" © tt 00 t'- tv CO CO

On On n O On to t' O

ro V V V V N to 10

d *

»n

13
>— »

too O 'O w H *OOn

tv © vo*-* wvO O 0

to (O to v to N to o-

o’*

On m N if) nO 't'tj-

w 0 H Tt m ro O

NO to Tf rj- to vo O'

d .

w 00 « Ol N N O

VO NO ON ON O O O

tO tv (ON (O VO Vo

d .

00 v 0 f-0 H to *>*

H M

■s

no ^ 0 oo 00 if 00

TO N to V V « M

O O O N On f' bo

(O »o to V to « fi

Good .

. do .

Excellent .

- do .

Fair .

. do .

s»

•s

> <u 2

<* SP 8
tfl «

t- 0 (O'© vo O t" vo

«0»0'VHM<>0«'0
© VOVO^-IOIOTJ-IO'O

d .

-0

>“>

00 0 0 00 t a 10 ifl

©CVtVtvO©MCC.H
t- to to v »o © <r xt \o

d .

°2.00 H f. vo © « On O

ft *"■ O tv V © 'Trow

J'tV0TtV0Tr»3-V0C0

0^0 N to OO ; 00 w

X 0N 00 On ON H .VO'O
^ u- <(• t , 10 vo

d . : * *

General appearance and size of plant.1

CVTftVTOtOtONMCV

£ 00 Cl 00 N 00 00 00 00 'O

•3

J 00 Cl >C GO CO 00 « N 00

> H H H HMHH

Excellent .

Good .

Poor .

Excellent .

. do .

Pair .

Poor .

• * * * * * * *••«*««•
OMCfr 0 ^ \0 N

hhhhhhmih

1 These data were taken at the time the last picking was made. By the end of the growing season, October 15, the plants had greatly increased in spread and also somewhat in
height. This was due largely to the new fall growth.

2 In 1912 this plant did not appear above ground until August 1.

* This plant was dead in 1912.

Nov. 10, 1913

Alkaloidal Content of Belladonna Plants

139

RELATION OR THE ALKALOIDAL CONTENT OF THE LEAVES TO THE STAGE
OF GROWTH OF THE PLANT

Opinions have been expressed from time to time as to the proper
stage in the growth of the belladonna plant at which the leaves should
be picked in order to insure the greatest percentage of alkaloids. Owing
to the standard required by the Pharmacopoeia, this is a question of no
small economic importance. Gerrard 1 has found that the plant is
not rich in alkaloids before flowering, but that the full development
is reached at the period of flowering and is maintained in both the roots
and leaves into the fruiting season.

The large number of assays of the leaves of individual plants here
involved presents exceptional opportunity for the study of the above
question. The proper season for the picking of belladonna leaves does
not, however, depend entirely on the percentage of active constituents
present. This will become very evident when the data at hand are
thoroughly interpreted. Table V shows in condensed form the number
of plants in which there was an increase or decrease in the percentage
of alkaloids in the leaves at the various pickings.

Table V .—Number of belladonna plants which showed an increase or decrease in per¬
centage of alkaloids in the leaves at the second , third, fourth , and fifth pickings as
compared with the preceding picking at Arlington Experimental Farm in ign and IQI2 .

Stage of growth.

Season of 1911.

Season of 1912.

Total

number

plants.

Number of plants
which showed —

Total

number

plants.

Number of plants
which showed —

Increase.

Decrease.

Increase.

Decrease.

Second picking .

7°

Third picking .

Fourth picking .

Fifth picking .

Table V shows that in 1911 the leaves of most of the plants were
richer in alkaloids at the second picking than at the first, which is in
accord with the observations of Gerrard, already noted. In 1912,
however, the opposite is true. It will be seen further that in the fourth
picking of both years the greatest number of plants showed an increase
in the alkaloidal content of their leaves. Referring to Table II, it is
seen that in the fourth picking in 1911 the average quantity of alkaloids
for the leaves of all the Arlington plants was 0.633 per cent, or more
than one-tenth of 1 per cent than at the flowering stage. In 1912, at

1 Gerrard, A. W. On the alkaloidal value of belladonna plants at different periods of growth. Year¬
book of Pharmacy, 1881-1882, p. 400-404, 1882.

140

Journal of Agricultural Research

VoL I, No. a

this same stage, the average was 0.568 per cent of alkaloids, which is
0.065 Per cent higher than the average at the flowering stage, although
lower in this case than at the early stage. There appears to be but
a slight difference so far as the alkaloidal content is concerned between
the flowering stage and the early fruiting stage. At the last, or fifth,
picking, the plants had acquired much new growth and, judging from
the average results, the percentage of alkaloids present in the leaves at
that stage was not much different from the second and third stages.

Although the experiments show that the leaves are richest in alkaloids
at the late fruiting stage of the plant, collection at that time for commer¬
cial purposes is practically out of the question because the leaves are of
very small size. After the flowering period is over and the berries are
ripening many of the large leaves fall off and numerous small, bractlike
leaves develop. These, while apparently rich in alkaloids, could not be
picked to advantage in large quantities.

RELATION OF SIZE AND APPEARANCE OF PLANTS TO ALKALOIDAL CONTENT

OF LEAVES

When this investigation was first undertaken it was hoped that some
relationship might be found to exist between the physical appearance
of the plants and the alkaloidal content of their leaves, for should such
relationship exist the process of distinguishing between the good and the
poor plants with regard to their active constitutents would become a
much simpler matter than by use of the assay method, since the latter
is necessarily tedious.

The variations in the physical appearance of belladonna plants
depend largely on the height and the number of stalks or stems. When
height is referred to here, the actual length of the stems from the ground
to the tips is meant rather than the vertical distance of the topmost
branches from the ground. This distinction is necessary because many
of the branches droop or grow at an angle. The spread of the plant,
that is, the distance around, is largely dependent upon the angles at
which the branches are growing and on the number of stems of the
plant. The height of the plant and the number of stems, therefore, are
the two distinguishing features as regards size. These indicate also
the relative health and vigor of the plant. An attempt was made
to differentiate between various types of leaves, with reference to size
and color and between different types as regards blooming and fruit¬
ing tendencies. It was found difficult, however, to find individuals
which conformed definitely to any particular type. Where certain
characteristics existed they were not as a rule general over the entire
plant, but were usually found on only one side or on only certain stems.
Thus, in some cases, one or two stems of a plant bore what appeared to
be leaves of a larger size than usual and of a different shade of green,

Nov. io, 1913

Alkaloidal Content . of Belladonna Plants

141

while the remainder of the plant was in every respect like most of the
other plants. The same would be true of the number of flowers and ber¬
ries. In such cases it could not be assumed that the plant represented
any special type. It was also noticed that some of these distinctive
features were subject to gradual changes, so that their distinctiveness
was soon lost.

While the number of plants that have been under observation was
probably not sufficiently large to show conclusively that there is no
definite correlation between physical appearance and active medicinal
properties in the leaves, yet from the data at hand such a condition is
at least indicated. Henderson,1 in commenting on the great variation
in the alkaloidal content of different lots of belladonna roots, points out
that appearance is no criterion of the quality, the best appearing roots
being often the poorest in medicinal value.

To show by actual examples that there is apparently no relation
between the appearance of the plant and its alkaloidal content it is
necessary only to refer to the tables. For example, in Table I plant No.
10 has a height of 42 inches and a spread of 3X by 4^ feet; in fact, it is
the largest plant in the list, yet its leaves contain only 0.536 per cent of
alkaloids, which is a trifle less than the average of all the plants. On
the other hand, plant No. 8, which is only half as high and much smaller
in spread, shows- 0.657 Per cent of alkaloids in its leaves. Again, in
Table II (season of 1911) plant No. 15 is the largest in the plat in point of
height, yet its leaves assayed only 0.494 Per cent, or less than the average
quantity of alkaloids. A similar statement may be made in regard to
large plants Nos. 4, 43, 45, and 46, while, on the other hand, the leaves
of the comparatively small plants, Nos. 21, 29, and iw, contained 0.630,
0.756, and 0.682 per cent of alkaloids, respectively. The data show
that in the following year these same plants failed again to compare
favorably with others as regards size, yet the percentages of active
constituent in their leaves stand out prominently above the average.
However, plants can be pointed out in the same table which are larger
and apparently more vigorous than the average and which also contain
above the average percentage of alkaloids in their leaves. The lack of
correlation is therefore very evident.

VARIATION AMONG PLANTS

Among the facts brought out by this investigation probably the most
important is the great variation in the percentage of alkaloids found in
the leaves of individual plants at each of the three testing gardens.
That some variation should exist was to be expected, since variations
are often noted in the chemical constituents of different plants of many

1 Henderson, H. J. Percentage of alkaloid in belladonna root. Pharm. Jour., v. 75, no. 3485 (s.4, v. si*
no. 1832), p. 191, 1905. * *

142

Journal of Agricultural Research

Vol. I, No. 2

other species. The knowledge of the existence of such individual varia¬
tions should have an important bearing on the question of the improve¬
ment of drug plants by selection and cultivation.

To show the great variation found among the comparatively limited
number of plants under observation Table VI is here presented.

Table) VI. — Range of variation in percentage of alkaloids in the leaves of belladonna
plants at each stage of growth , at Arlington , Madison , and Bell stations , in different
years.

Alkaloidal content of the leaves (per cent) .

Stage of growth.

Arlington, Va.

Madison, Wis.

Bell, Md.

1910.

1911.

1912.

1911.

1912.

1911.

High.

Low.

High.

Low.

High.

Low.

High.

Low.

High.

Low.

High.

Low.

First picking .

Second picking .

Third picking .

0. 700

0.334

0.852

.879

•925

.891

• 733

0.303

. 262

.277

•311

. 200

0. 869

• 747

.88 2

.806

.678

0. 404

. 292

.328

•359

. 296

0. 580

. 820

. 767

0. 418

•427

.419

0. 500

• 5i9

0. 268

.316

0. 823 .

• 783

• 750

0.329

.28 8

•395

Fourth picking .

Fifth picking .

Season average .

Average .

.766

.306

.768

•353

.665

•430

•452

.312

• 707

•34<S

. 841

.277

.792

•339

. 708

•423

.490

. 298

. 766

■ 339

From this tabulation it appears that the active principle is more than
three times as great in the leaves of some plants as in those of others
at the same period of growth, although the plants are in the same plat
and therefore grow practically in the same soil and under the same
climatic conditions. Under such circumstances the existing variation
can hardly be attributed to anything but the inherent characteristic of
the individual plant. Much has been written concerning the influence
of soil and climate on the formation of alkaloids in the plants. Gerrard* 1
has found that a chalky soil favors the formation of atropin. Cheva¬
lier 2 concludes from his experiments with fertilizers that the alkaloidal
content of certain Solanaceae can be increased by means of nitrates and
farmyard manures. Ransom and Henderson,3 however, who are working
along the line of Chevalier's experiment, have not found thus far that
artificial manures materially affect the percentage of alkaloids in the
dried leaf, but note in several cases a large increase in the yield of the

1 Gerrard, A. W. Op. cit.

! Chevalier, J. Influence de la culture sur la teneur en alcaloxdes de quelques Solan&es. Compt. Rend.
Acad. Sci. (Paris), 1. 150, p. 344-346, 1910.

1 Ransom, Francis, and Henderson, H. J. Belladonna: the effects of cultivation and fertilizers on the

growth of the plant and its alkaloidal content. Chemist and Druggist, v. 81, no. 1703, p. 53-55. 191a.

Nov. io, 1913

Alkaloidal Content of Belladonna Plants

143

green plant per acre. Carr 1 claims to have found a certain relationship
between the amount of sunshine during the growth of the plant and the
percentage of alkaloids found in the stems and leaves, claiming that
plenty of sunshine and limited rainfall have a tendency to stimulate the
production of alkaloids.

Although soil and climate may have considerable influence on the
alkaloidal content of plants, yet to establish this as a fact beyond all
doubt is a difficult matter because of the individual variation involved.
Until experiments have been conducted upon a large number of plants
which show a minimum variation in their alkaloidal content, nothing
definite can be said upon this point. In working with a limited number
of plants collectively, an abnormally low or high percentage of alkaloids
in the leaves of a few might so affect the yield as to make the average
entirely misleading. Likewise, this individual variation becomes an
important matter in the sampling of large quantities of leaves and roots.
In order to secure a reliable sample, it should be of considerable size and
selected only after the leaves or roots have been thoroughly mixed.

INDIVIDUAL VARIATION THROUGH SEVERAL SEASONS

Having definitely established the fact that great variations exist in
the alkaloidal content of the leaves of individual plants, the question
remains to be answered whether such variations exist only during one
growing season or whether they manifest themselves in the same propor¬
tion in following seasons. If plants which are rich in alkaloids one season
are correspondingly poor the following season, then it is logical to assume
that the production of alkaloids in the plant is dependent on factors
which change from year to year. If it Were definitely known what r61e
the alkaloids play in the metabolism of the plant, it might be easier to
determine what factors influence their development. As has been
shown, the physical appearance, or, in other words, the vitality and
growing power of the plant, appears to bear no definite relation to the
development of alkaloids. Furthermore, if soil and climate are the
potent factors, then their influence ought to be felt by all plants alike
when all are grown on similar soil and in the same locality. Such,
however, has been found not to be the case, and reference to the tables
shows that there were plants rich and poor in alkaloids in every year
during which the observations extended. On the other hand, if it
should be found that a plant with leaves containing an unusually high
or low percentage of alkaloids in one season shows the same characteristics
in following years, it would be safe to assume that there is a definite tend¬
ency in that plant to produce a small or a large quantity of alkaloids in the

1 Carr, F. H. The effect of cultivation upon the alkaloidal content of Atropa belladonna. Chemist
and Druggist, v. 8i, no. 1703, p. 43-44, 19x2.

144

Journal of Agricultural Research

Vol. I, No. a

course of a season's growth, just as in other plants there are well-defined
tendencies toward certain physical characteristics.

This investigation, however, has hardly progressed far enough to yield
any definite conclusions. In Table VII a comparison is made between

Fig. i. — Diagram showing the percentage of alkaloids in thel eaves of individual belladonna plants
at the Arlington Experimental Farm, Va., during the seasons of 1911 and 1912.

the years 1911 and 1912 of the 59 plants grown at Arlington, showing
the variation of alkaloidal content above and below the average for each
of the years mentioned. Figure 1 shows graphically the seasonal com¬
parison.

Nov. io, 1913

Alkaloidal Content of Belladonna Plants

145

Table VII. — Percentage of alkaloids above and below the average 1 in the leaves of indu
vidual belladonna plants at Arlington, Va.t in ign and IQI2.

[The figures given are based on the season averages of all the pickings. In each of the 40 plants designated
by a star (*) the percentage of alkaloids above or below the average of the entire lot in 1911 varies by
not more than one-tenth of 1 per cent from that in 1912.]

5-
6*
7.
8*
9-
10*
11 .
12*
13*
14*

15*

16*
17.
18*,
19*,
20*,
21*.
22 .
23*.

24.

25.

26.

27.
28*,
29*,
30*.

Plant No.

Alkaloids above
(+) or below (— )
the average (per
cent).

Plant No.

Alkaloids above
(4-) or below (— )
the average (per
cent).

19x1.

1912.

1911.

1912.

— 0. 068

4- ■ 179

4- .023

32* .

— .006

—0. 003

— .125

— . 116

33* .

— .046

4- .044

— .014

4- .014

34* .

— . 118

— • 139

-f .047

— .071

35 .

— . 102

4- .071

— .008

— .073

36* .

- .086

— . 062

4- .070

— .081

37* .

— . 017

4- .003

4- .064

+ • 144

38* .

4- .063

4- .061

. 048

— • 116

39* .

- .141

— .072

— . 029

- • 125

40* .

4- .028

— .012

-f .087

- -OSS

41* .

— . 104

— .076

4- . 109

+ *093

42 .

4- .001

4- .134

4* • 107

— .023

43* .

— -033

— . 012

— *044

- ■ 125

44* .

— .043

— .082

— .038

4- .005

45* .

— .044

— .085

4- .077

4- .086

46* .

— .142

— . 139

— .059

4~ • 066

47* .

— *079

— . 040

4- .019

— . on

48 .

4- .036

- .093

— .005

— .059

49* .

4- .024

— . 064

4- .036

+ • 133

5° .

— .005

— . 192

4- .098

4- . 199

IW* .

4- . 150

+ -174

—I

4- ■ 041

2W* .

— . 001

4- .015

— . 129

- ■ 144

5W* .

4- .052

4- .081

6w* .

4- . 234

4- • 243

— . 226

— .056

7W* . .

4* • 172

4- . 127

+ -03s

4- . x8i

8w .

4- .045

- .077

— . 090

4- .in

9W* .

— .018

— .004

— . 019

4- .027

IOW* .

4- .055

4- .065

4- * 224

4- . 159

IIW* .

- .088

— .001

— . 002

4- .027

1 Average for 1911, 0.532 per cent; for 1912, 0.545 per cent.

In the plants in Table VII there are a number which are conspicuous
because of the high or low percentage of alkaloids in their leaves. Plants
Nos. 3, 23, 34, and 46 are without doubt greatly inferior to the others
from a medicinal point of view. On the other hand, Nos. 21, 29, iw,
6w, and 7w are greatly superior to any others in the list. Furthermore,
these plants manifested the same characteristics not only on the average
but at each picking. The recapitulation given in Table VIII shows
this very clearly.

146

Journal of Agricultural Research

Vol. I, No. 2

Tabus VIII. — Alkaloidal content of the leaves of belladonna plants , rich and poor in
alkaloids , at various stages of growth , in igil and 1912.

Plants with leaves of low alkaloidal content (per cent).

Stage of growth
(picking).

No. 3.

No. 23.

No. 46.

1911.

1912.

1911.

19x2.

1911.

1912.

1911.

1912.

First .

0. 384

*375

.277

•549

•451

0. 496
.366

■341

0-335

o- 337

.285

.308

.588

*431

0.418

■334

.480

•483

*314

Second .

0*393

.448

0.348

•354

.487

*425

0. 292

• 520

Third .

• 526

*532

. 200

Fourth .

Fifth .

Average .

.407

.429

■403

.401

.414

.406

•390

.406

Plants with leaves of high alkaloidal content (per cent).

Stage of growth
(picking).

No. 21.

! No. 29.

No. iw.

No. 6w.

No. 7W.

1911.

1912.

1911.

1912.

1911.

19x2.

1911.

1912.

1911.

1912.

First .

Second .

Third .

Fourth .

0. 535

•633

. 669
.684

0. 732

.719

.781

0.655

.914

.908

• 547

0.737

.647

• 729

0. 638

•835

*587

.738

. 612

0.737

.642

• 777

0. 596
.879

•925

• 7H

. 722

a 847

* 747

.882

. 804

.558

0.558

.831

.832

. 727

•571

0. 782

.666

.646

.694

•573

Fifth .

Average .

. 630

* 744

• 756

. 704

.682

• 7i9

. 766

. 768

• 704

.672

SUMMARY

From the point of view of the percentage of alkaloids present in the
leaves and the quantity of material available, the leaves can be picked
to best advantage from the time of flowering until the early berries begin
to ripen. Although the leaves are richer in alkaloids later in the season,
they are then too small and sparse for harvesting.

Thus far nothing has been found to indicate that any correlation
exists between the physical appearance of the plant and the alkaloidal
content of its leaves. Luxuriant growth is by no means a criterion of
the medicinal value of the plant.

The variation of the percentage of alkaloids in the leaves of the dif¬
ferent plants is exceedingly large. This makes it a difficult matter to
determine to what extent soil and climate influence the development of
alkaloids. Where such wide variations exist among individual plants,
the testing of a general sample from all plants collectively is not always
a safe means of judgment.

A considerable number of plants with leaves rich in alkaloids in one
season are found to have equally rich leaves in the following season.
Furthermore, they frequently manifest the same characteristics at the
various stages of growth during the season in comparison with other
plants. The same facts are true with regard to plants which bear leaves
with a low percentage of alkaloids.

THE PUBESCENT-FRUITED SPECIES OF PRUNUS OF
THE SOUTHWESTERN STATES

By Sii/AS C. Mason,

Arboriculturist , Crop Physiology and Breeding Investigations , Bureau of Plant Industry

INTRODUCTION

The species of the genus Prunus described in this article occupy a
unique position in the flora of the western United States from the fact
that their relationship with the wild plums of the country is remote
and they are more closely allied to some of the Asiatic species of this
genus.

Their economic importance arises chiefly from their close adaptation
to the climatic and soil conditions of the Southwest, where fluctuations of
heat and cold, severe drought, and considerable alkalinity of the soil
must be endured by most tree crops.

Adaptable stocks for the cultivated forms of Prunus capable of meeting
such conditions are eagerly sought. Species with such characters which
are capable of being hybridized with the old-established cultivated forms
of the genus offer attractive possibilities to the plant breeder. This is
especially true of the one edible-fruited form, Prunus iexana , which
affords in aroma and flavor of fruit most attractive characters for combi¬
nation with other stone fruits of larger size and more staple commercial
character.

Instead of forming a homogeneous group, as has usually been be¬
lieved, these species fall into small groups of quite diverse character
and affinities. To the plant breeder and student of their economic
possibilities these relationships are of such importance that the following
detailed study of them is deemed essential to an intelligent use of them
in plant-breeding work.

In parts of the country beyond the Rocky Mountains a few ranchmen,
occasionally a solitary mining prospector, and a few local botanists
know of curious bushy plants growing in desert wastes having plumlike
bark and twigs, oddly shaped leaves, and small downy fruits with thin
dry flesh which have won for them the local names “wild almond”
in the Great Basin region, “wild peach” or “desert almond” for another
form in the Mohave Desert, and “wild apricot” or “wild almond” for a
third form in the foothills bordering the Salton Basin in southern
California.

A fourth form has been known for many years to the pioneers of
eastern Texas, who have enjoyed eating the “wild peach” of their sandy

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Vol. I, No. a
Nov. io, 1913
G-4

148

Journal of Agricultural Research

Vol. I, No. a

country, the only really edible fruit of the group. However, this fruit
is still strangely ignored by horticulturists and botanists alike.

A fifth form, growing in the limestone plains of central Texas, has a dry
and inedible fruit which has not sufficiently attracted the attention of the
cattlemen and goat herders in this sparsely settled region to earn a local
name.

A sixth form, growing in the high altitudes of both northern and
southern Mexico, though the first of all to receive botanical notice (1823)
is still very rare in herbaria and has been seen in its native habitat by but
few botanical explorers. It was first collected by Humboldt in his
famous journey through the Mexican plateau region. A seventh species,
Havard’s wild almond, still very imperfectly known, has recently been
described from the region inclosed by the Big Bend of the Rio Grande in
western Texas.

We have, then, native to the region of North America, lying west of
the Mississippi drainage area, six or seven members of the plum family
differing in a very marked way from the familiar types of American wild
plums.

They are united by the common character of a woolly or pubescent
fruit, and all are deep-rooted, with remarkable drought resistance.
This fruit character, so at variance 1 * with the true plums of America or of
the Old World, would at first seem to ally these species with the almond
or apricot sections of the genus, as their common names suggest. A close
examination of their botanical characters shows, however, that they fail
to agree with those groups and must be regarded as occupying interme¬
diate ground between the true plums on the one hand and the almonds
or apricots on the other. Aside from the common character of pubescent
fruit and their deep-rooting habit, these species differ widely from one
another, which is to be expected from the wide geographic range which
they occupy and the resulting differences in climate and soil.

HABITAT AND ENVIRONMENT

Ranging farthest north is the commonly named * 4 wild almond”
(. Prunus andersonii)t which is found around the shores of Pyramid Take,
Nev., in the Honey Take region of California, and along the basin slopes
of the Sierras, having an altitude range of from 4,000 to 8,000 feet in
the Upper Sonoran and Transition life zones. (See map, fig. 1.) This
is consequently subject to severe cold in winter, as much as 20° F. below
zero in some instances, and to extreme drought and severe heat in the
summer. It is usually found in gravelly or sandy soils.

Its near relative, the “wild apricot” {Prunus eriogyna ), found along
the desert slope of the San Bernardino and Santa Rosa Mountains and

1 Prunus oregana Greene, of Oregon and northern California, has fruit with a fine, soft pubescence, but it

is a true plum, near to P. subcordata.

Nov. io, 1913

Pubescent-Fruited Species of Prunus

149

southward into Lower California, is an inhabitant of much lower altitudes,
at least in California. There it occurs at from 500 to 3,000 feet in the
upper margin of the Lower Sonoran, but chiefly in the Upper Sonoran
zone, extending a little below the zone of light winter snow, though subject
to intense heat and prolonged drought in summer. (See map, fig. 1.)

Most similar to this species in habitat and requirements, though remote
in relationship, is the “desert almond”1 (Prunus fasciculata) . This fruit

Fig. i. — Map of the southwestern part of the United States, showing the range of Prunus andersonit ,
Prunus fasciculata, and Prunus eriogyna, n. sp.

occurs in widely scattered localities over a range which includes southern
Nevada and California, together with the adjacent portions of Utah and
Arizona. It overlaps portions of the areas of both Prunus andersonii and
Prunus eriogyna , but, like the latter, is found in the upper margin of the
Lower Sonoran and in the Upper Sonoran zones. (See map, fig. 1.) It

1 Called “desert range almond ” by Dr. C. H. Merriam in notes on the distribution of trees and shrubs in
the deserts . . . U. S. Dept. Agr., Bur. Biol. Survey, North American Fauna, no. 7, p. 301, 1893.

150

Journal of Agricultural Research

Vol. I, No. a

is nowhere subject to the severe cold endured by the Nevada “wild
almond” in its most northern habitat. It usually grows in gravelly
formations or along washes or sandy slopes where deep root penetration
is possible.

This “ desert almond” is remote geographically from the two species of
the group to which it is most nearly allied, the Texas wild almond, Prunus
minuli flora, and its Mexican cousin, Prunus microphylla, which may
consistently be called the “Mexican wild almond.”

The Texas species has a range not yet well worked out, but it is appar¬
ently confined to the Cretaceous limestone region of the southwestern
portion of the State, extending across the Rio Grande into the State of
Chihuahua, Mexico, and probably occurring in Coahuila. Its known
localities are entirely in the upper portion of the Tower Sonoran zone.
(Fig. 2.) It is found over an area ranging in altitude from 750 feet near
San Antonio to 3,000 feet near the mouth of the Pecos River, with an
average rainfall of about 20 inches, but subject to periods of prolonged
drought. There is an absolute temperature range for the years recorded
of from zero to no° F., with the liability to sudden drops from winter
northers, peculiar to this region.

Of the conditions under which the Mexican species grows we have but
indefinite knowledge, but it occurs at high altitudes — 6,000 to 8,000 feet,
the Upper Sonoran zone of this southern latitude, probably a mild tem¬
perate climate with light winter rains and heavy summer showers. In
common with the other species it grows in a region where the setting of
the fruit is frequently prevented by late spring frosts.

The little-known Havard’s wild almond, Prunus havardii , apparently
a near relative of these two species last mentioned, has been found so far
only in western Texas.

The Texas “wild peach,” Prunus texana , occurs in scattered localities
over a region of eastern Texas from near sea level to nearly 2,000 feet in
elevation, lying wholly in the Lower Sonoran or Lower Austral zones.
This includes a portion of the western extremities of the com and cotton
belts, where an apparently sufficient annual rainfall is so unevenly dis¬
tributed that long periods of drought make agriculture somewhat pre¬
carious and render irrigation a needful adjunct. It is adjacent to the
area of Prunus minuti flora, but the division with its sharp demarcation
is not one of climate, but of soils. Prunus minutiflora follows the Cre¬
taceous limestone of the plateau region, while Prunus texana occurs on
the mellow granitic sandy soil of the “Burnet Country” or the sandy
loam of the Coastal Plain and is wholly wanting on limestone soils.
(See map, fig. 2.)

Nov. 10,1913 Pubescent- Fruited Species of Prunus 151

BOTANICAL CHARACTERS OF THE GROUP

The botanical characters of the seven species under consideration, even
the obvious character of the leaves and fruit, are so distinct from those
generally recognized as belonging to wild or cultivated plums that it is
not surprising that the Mohave Desert form was first assigned by Dr.
Torrey to a new genus, Emplectocladus, from the Greek words referring
to its interlocking branches. This was later placed in the genus Prunus
by Gray, but as a separate section. Schneider, while including all these

Fig. 2. — Map of Texas, showing the known areas and probable range of Prunus minutiflora and Prunus

iexana.

species under Prunus,1 groups them in the section Emplectocladus along
with Torrey’s original species, Prunus (. Emplectocladus ) fasciculata, and
the Old World P. pedunculata. Several authors have assigned some or
all of the species to Amygdalus.

The study of the entire group from abundant material and the field
examination of all but Prunus microphylla and P. havardii convince the
writer that they are separable into three distinct sections.

1 Schneider, C. K. Illustriertes Handbuch der Taubholzkunde. Bd. i, Lfg. 4, Jena, 1905, p. 589-590.

152

Journal of Agricultural Research

Vol. I, No. a

The so-called wild almond (Prunus andersonii) , chiefly found in Nevada,
though also occurring along the eastern slope of the Sierras in California,
is upon careful comparison found to be very closely related to the wild
apricot ( Prunus eriogyna) of the Colorado Desert in southern California.
These two species are clearly separated from the peach and almond by the
characters of the leaves both in vernation and when mature, by floral
characters, and by the seeds.

The entire group (the genus Amygdalus of some authors) of the genus
Prunus which includes the almonds and peaches has leaves folded length¬
wise in the bud (conduplicate), the flowers sessile or subsessile, the stones
rugose and pitted.

The Nevada wild almond, notwithstanding the fact that it has been
described as being “ a true almond in its affinities,” 1 and the desert apricot
agree with the section Armeniaca, the apricots, in three important points:
First, the leaves in the bud are rolled from the margin toward the middle,
or convolute; second, the flowers are stalked, some on pedicels three-
fourths of an inch long; and, third, the stones are smooth or but faintly
pitted and decidedly wing-margined.

These characters are found also in some of the true plums, but a distinct
separation from the plums is met in the rose-colored flowers and in the
only slightly fleshy, pubescent fruits.

The presence of stomates in the upper surfaces of the leaves is a
character distinguishing these two species from both the Amygdalus and
Armeniaca sections.

Their characters as a whole, however, seem to unite them most closely
with the apricots, and apparently there is nothing among the European
and Asiatic forms of Prunus to which they are as closely related. Con¬
sequently the two species are here placed in a new section, Penarmeniaca
(near-apricots).

The California desert almond (Prunus fasciculata) , the Texas wild
almond (P. minutiflora) , and the Mexican wild almond (P. micro phylla) ,
agree in three important characters which separate them clearly from
the three other species of this group. All three are dioecious by the abor¬
tion of either stamens or pistils; the number of the stamens is usually
reduced to io or 15 and a portion of them inserted on the walls of the
calyx cup. They further agree in having the inner face of the cup finely
hairy instead of having a nectariferous surface as in apricots, peaches, and
almonds. Havard's wild almond probably belongs in this same group.

Prunus fasciculata has leaves with stomates in the upper surface, in
which it resembles P. andersonii and P. eriogyna , while the other three
species have no stomates in the upper surface. However, on the strength
of the characters possessed in common, especially of the remarkable one
of the dioecious character of the flowers, Prunus fasciculata is placed with

Greene, E. E. Flora Franciscana. [Pt. i], San Francisco, {1891] p. 49.

Nov. 10, 1913

Pubescent-Fruited Species of Prunus

153

the Texas and Mexican wild almonds in the subgenus Emplectocladus of
Prunus. This has been done with a full realization that most definitions
of this genus describe the flowers as perfect, though Sargent 1 and Schnei¬
der extend the definition to include polygamo-dioecious flowers. No
reference to dioecious or polygamo-dioecious characters in any Asiatic
forms of Prunus has been found.

While a more complete knowledge of the Asiatic forms 2 may disclose
closer affinities for these three species, they are retained provisionally as
the sole member of the subgenus Emplectocladus. With our present
knowledge of these forms the seven species of Prunus studied in this
paper should be grouped as follows :

SCHEME OF CLASSIFICATION

PRUNUS

Subgenus Emplectocladus

Low divaricate or erect shrubs with more or less spinescent branches. Bark on new
growth gray or brownish, glabrous or more or less pubescent. Leaves conduplicate in
vernation; borne singly on vigorous young growth or apparently fascicled on budlike
suppressed branchlet, with or without stomates in upper epidermis.

Flowers solitary or gemminate, sometimes crowded on short fruiting spurs, subsessile,
precocious or coetaneous with the leaves, dioecious by the abortion of stamens or
pistils; calyx cup obconic or campanulate, glabrous or faintly puberulous on the
outer surface, minutely hairy within; stamens usually. 10 to 15 on short filaments, in
three more or less well-defined circles, inserted on the margin of the cup and on the
walls below; ovary and base of style pubescent.

Fruit seldom more than 1 cm. long, pubescent, subglobose or irregularly ovate, with
thin, dry flesh splitting tardily, and smooth or obscurely ridged stone.

Four species: Prunus fasciculate Gray, Prunus minutiflora Engelm., Prunus micro -
phylla Hems., and Prunus havardii (Wight), n. comb.

Subgenus Euprunus

SECTION PILOPRUNUS, N. SECT.

Low, much branched, often procumbent, scarcely spinescent shrubs, with gray
or brown, pubescent young wood.

Leaves conduplicate, without stomates in upper epidermis, tomentose, glandular
serrate.

1 Sargent, C. S. Silva of North America. Boston, 1892, v. 4, p. 7.

2 Prunus pedunculata (Pall.) Maxim, and P. pilosa (Turcz.) Maxim, of Mongolia are said by Koehne (PI.
Wilsonianae, pt. 2, p. 273) to have the calyx cup dry within and minutely hairy at the insertion of the
stamens. Schneider figures (Laubhk., v. 1, p. 598, fig. 335 <*) the whole interior of the calyx cup of P. pedun¬
culata as finely hairy. Tittle is known as to the flower characters of Prunus boissierii Carr, from Asia Minor
referred to P . pedunculata by Schneider, but which differs in having sessile flowers. These plants are
referred to the section Emplectocladus by Schneider, but his figures of P. pedunculata show a perfect flower
and no hint is given in descriptions of the other forms of their flowers being dioecious. These species, as
well as the little-known P. mongolica and P. dekiscens Koch., grouped along with them by Koehne (PI.
Wilsonianae, pt. 2, p. 274), and P. petunnikowi Utw. doubtfully referred to this group by Schneider ( Uaubhk
v. 2, p. 974), all need to be studied carefully so as to permit of a careful comparison with the American forms
here referred to the section Emplectocladus.

13000°— 13 — s

154

Journal of Agricultural Research

Vol. I, No. 2

Flowers white, appearing with the leaves, fascicled on short pubescent peduncles,
perfect, highly fragrant; calyx cup campanulate, pubescent without, nectariferous
within, with glandular serrate lobes; ovary finely pubescent.

Fruit 1.5 cm. to 2.5 cm. long, pubescent, the juicy, fragrant, highly flavored
flesh clinging to the stone by a persistent velvety pile; stone rounded, smooth or
scarcely furrowed.

One species: Prunus texana Dietr.

SECTION PENARMENIACA, N. SECT.

Dense shrubs with angled and thorny branches or of more smooth and erect arbores¬
cent growth reaching 3 meters in height; young twigs glabrous, reddish or yellow
brown.

heaves convolute in vernation, glabrous, more or less glandular serrate, with
stomates in the upper epidermis.

Flowers rose colored, pale pink, or rarely white, solitary or in fascicles of two or
three, on stalks from 5 to 15 mm. in length; stamens 20 or 30, inserted near the rim
of the calyx cup; calyx cup campanulate, with nectariferous lining; pistil as long
or longer than the stamens; ovary and base of style pubescent.

Fruit oval or subglobose, 1 to 2 cm. long, pubescent, somewhat fleshy while imma¬
ture, harsh and astringent but with an acid, fruity flavor, opening along suture when
mature; stone thick walled, furrowed, with obscure reticulations or smooth or
somewhat pitted; kernel in some varieties edible, often strongly flavored with prussic
acid.

Two species: Prunus andersonii Gray and Prunus eriogyna , n. sp.

THE WILD PEACH

The earlier botanical descriptions of the important species Prunus
texana are so meager that the following description in greater detail
seems necessary :

Prunus texana Dietr.1

Amygdalus glandulosa Hooker, Icon. PI., v. 3, pi. 288, 1840.

Prunus glandulosa (Hooker) Torr. and Gray, FI. N. A., v. 1, p. 408, 1840.

Prunus texana Dietr., Syn. pi., v. 3, p. 45, 1843.

Prunus Hookeri Schneider, Daubhk., v. i, L,fg. 5, p. 597-598, fig. 335, i, k, 1, 1906.

Amygdalus texana (Dietr.) W. F. Wight, Dudley Mem. Vol., p. 131, 1913.

Illus. Hooker, loc. cit.; Schneider, loc. cit.

Low, squarrose shrubs, sometimes reaching a height of 2 meters, with a spread
of 2 to 2.5 meters; stems usually slender but occasionally erect and stout branches >
rarely spinescent; bark dark iron gray, roughly furrowed on old wood, on young growth
grayish brown or silvery gray, densely pubescent.

The leaves, conduplicate in the bud, are usually narrowly elliptical, with rounded
apex and rounded or wedge-shaped base; thick, strongly veined, serrate or crenately
doubly serrate, with glandular teeth, dull green, thickly pubescent above, canes-
cent beneath, 1.5 to 4 cm. long, 6 to 18 mm. broad; petiole short, rather thick,
stipules 3 to 4 mm. long, narrowly lanceolate, with glandular teeth.

The small flowers, which appear with the leaves in February and March, are
fragrant, perfect, 1 to 1.5 cm. broad, borne singly or in fascicles of two or three on
short, finely pubescent peduncles; the campanulate calyx tube is finely pubescent,

1 There being a Prunus glandulosa of Thunberg, 1784, Hooker’s Amygdalus glandulosa can not be trans¬
ferred to the genus Prunus and the name Prunus texana , given by David Dietrich (Synopsis plantarum, v.
3, Vimariae, 1843, p. 45), has priority and is a most appropriate one, as this interesting species has so far
been found only within the limits of Texas. This conclusion as to the priority of Dietrich’s specific name
is confirmed and published by Dr. C. S. Sargent in Trees and Shrubs, v. 2, pt. 3, Boston, June, 1911.

Nov. io, 1913

Pubes cent- Fruited Species of Prunus

155

the strongly reflexed, short rounded lobes being glandular ciliate margined, with
fine soft hairs on both surfaces. The inner face of the tube is lined with an orange-
colored, nectariferous layer. The thin white petals, 5 mm. long, are broadly ovate,
often truncate at the base, attached by short, stout claws. The ovary and two-thirds
of the length of the style are finely pubescent. The fruit is roundish oval or oblong,
usually with a ventral shallow furrow, 1 to 2.5 cm. in length, a sharp depression at
the base, pedicel 5 to 8 mm. long. The skin is rather thick, coated with fine pubes¬
cence, yellowish, greenish yellow, or rarely taking a rich reddish flush on one side;
flesh yellowish or greenish yellow, finely netted, juicy and luscious, sometimes
very richly flavored, clinging to the rather large stone by a curious tough, persistent
elastic pile, like coarse plush, which, when scraped away, leaves an ovate obtusely
pointed, thin- walled seed without pits or furrows. The
kernel is plump, roundish pointed, slightly furrowed, and
with a strong flavor of prussic acid. (PI. IX, and fig. 3.)

It is plain that with its strongly glandular
pubescent leaves and luscious, fleshy fruit with
the pilose or velvety stone it has little near rela¬
tionship with the five species of the group in
which it has been included. It has accordingly
been placed in the subgenus Euprunus and in
a new section, Piloprunus. Analogy for the
pubescent fruit is found in the Prunus oregana
of Greene and for the netted flesh clinging to the
stone in the sand plum, P. watsoni.

With a promising wild species of distinctly
limited range it is of first importance to learn
under what conditions of soil, temperature, and
rainfall it has been able to reach its present
standing in the plant world. In a State afford-

. , ,, „ Fig. 3.— Prunus texana Dietr.: Ar

mg so vast an open range,” so to speak, as the section of calyx, x 3; b, detail

State of Texas, the restriction of a species to a of calyx Iobes- showinK slandl1-

range must mean certain limitations in endur- calyx from flower of the horticui-

ance. If it stops rather sharply as soil types tural variety Ramsey, P. Uxana

1 . w . ^ X Wild Goose plum, X 4.

change, with no other apparent reason for not

extending farther in that direction, we must suspect a soil preference
amounting to limitation. A fairly well-defined northern boundary is
pretty sure to mark the limit of cold endurance, provided soil and moist¬
ure conditions would seem to invite farther advance in that direction.
Therefore, the geographic range or distribution of the wild peach should be
studied and also related conditions of soil, temperature, and moisture.

Journal of Agricultural Research

Vol. I, No. 2

156

DISTRIBUTION AND SOIU *

The range of the wild peach is wholly within the State of Texas, but
its local distribution is not yet worked out. As shown by the map
(fig. 2), there are two principal areas of its growth. The first of these is
what is called the “Burnet Country/' a region of granitic uplift occupy¬
ing the greater portion of Llano County, and small areas of Burnet, San
Saba, Mason, Gillespie, and Blanco counties. It is also found along a
narrow alluvial strip next to the Colorado River in Lampasas County.

It is upon the sedentary soils from granitic disintegration, small areas
from sandstone and schistose rocks of the earlier stratified formations
bordering and upturned by the granitic protrusions, and on narrow
strips of river alluvium that the “wild peach ” occurs. Only one instance
is known of its occurrence upon the calcareous areas which surround and
in isolated patches overlap the granitic protrusions.

The second considerable area known for this species lies in the south¬
eastern part of Bexar County and in the adjacent counties of Guadalupe,
Wilson, and Atascosa, extending eastward into Gonzales and southward
into Bee County. As this region is a part of the area of sandstone forma¬
tion known geologically as the Marine Eocene region and the plants are
found only on rather mellow sandy soils, we must conclude that the species
has so strong a preference for granitic or sandy soils as to practically
exclude it from limestone regions. It was learned in the neighborhood
of Lavemia that extensive areas of this “May plum," as it is called in
that section, had been destroyed in the clearing up of fields. Isolated
patches have been found at points as remotely separated as Van Zandt
County at the north, the coast dunes of Aransas County, and a consider¬
able area in the sandhills of Hidalgo County at the south, where the
fruit is much esteemed by the Mexicans under the name “durasnillo,”
or “little peach.”1 It seems probable that a more complete survey of
the eastern portions of the State would show that the wild peach has a
botanical range extending over a greater portion of the sandy formation
of the Marine Tertiary region, restricted probably by lack of moisture in
the southwest portion of that formation.

All the plants studied have a deep- rooting habit, enabling them to
penetrate to layers of soil where the moisture is fairly permanent as is often
the case where the soil has a sandy foundation. This aids them greatly
in surviving the long periods of drought to which the country is subject.
The thickly pubescent upper surface of the leaves and the almost felted
undersurface are features which reduce transpiration and must aid
materially in drought resistance.

1 Prof. S. W. Stanfield, of the Texas State Normal School, states that in southern Bexar County this fruit
is called “albaricoque,” which is the Spanish name for the apricot.

Nov. 10, 1913

Pubescent- Fruited Species of Prunus

157

CLIMATIC CONDITIONS

The principal areas occupied by Prunus iexana are represented by
fairly complete weather records at Menardville, Fredericksburg, and
San Antonio 1 and by volunteer records at Burnet, Llano, and Lampasas.2
These show that the mean annual rainfall ranges from 22.6 inches in
the more westerly to 28 inches in the eastern and southern portions.

The monthly means show a fairly well-distributed rainfall throughout
the year. December to March constitutes the drier period, with Febru¬
ary as the driest month. The study of the monthly records of a number
of years, however, shows that this section is subject to occasional
heavy rainfalls almost torrential in character, as well as to periods of
severe and prolonged drought. A study of the extremes of rainfall at
San Antonio, a nearly central point in the range of this species, shows
that during the driest year of the period covered by the record, 1885
to 1903, only 15,9 inches of rain fell, while the maximum record was
40.5 inches. The structural characters enabling Prunus iexana , the
wild peach, to endure these vicissitudes are important features to study.

The temperature conditions characteristic of this section are those of
comparatively mild winters, minimum temperatures of 120 to 160 F.
being matters of common record, with occasional winters showing
minimum records of as low as — 20 to —4° F.3

Minimum temperatures of 50° to 6o° and maximum temperatures of
6o° to 750 F. may be followed in a short time by a norther which will
lower the temperature to near the zero point, or even below. The extreme
maximum temperatures experienced in this section are from ioo° to
105° F.

NATURAL HYBRIDIZATION

One of the most striking characteristics of the wild peach is the readi¬
ness with which it hybridizes with the native and cultivated plums. This
is proved by the occurrence of well-marked natural hybrids with the local
wild plums in at least five widely separated localities within its range.

The occurrence of natural hybrids between species of plants is unusual
and in many families rare or unknown. The integrity of our plant forms
could not be preserved if indiscriminate natural hybridizing were a pos¬
sibility.

Probably among trees and shrubs the most numerous examples of such
hybrids are afforded by the oaks of the Mississippi Valley and the Western
States, and a number of these have from time to time received definite

1 Henry, A. J. Climatology of the United States. U. S. Dept. Agr., Weather Bur., BulletinQ., p. 431-436,

1906.

3 U. S. Dept, Agr., Weather, Bur., Climate and Crop Service, Texas Section, v. 1-5, 1897-1901.

8 There was a record of —4.1* F., at Llano, Feb. 12, 1899. U. S. Dept. Agr., Weather Bur., Climate and
Crop Service, Texas Section, Report, v. 3, no. 5, p. 5, 1899.

Journal of Agricultural Research

Vol. I, No. 2

158

botanical description. A few wild grape hybrids are recorded in the
writings of Dr. Engelmann.

Among plums a few definite natural hybrids of the wild species have
been recognized, and the later judgment of Prof. Bailey on Prunus
hortulana , described by him as a species, is that it is a group of varieties
which are hybrids between Prunus americana and the southwestern
species, Prunus angustifolia }

On the whole, surprisingly few authentic hybrids have come into
existence without the aid of man among uncultivated plants.

Examples of natural or unassisted hybridization among cultivated
plants are somewhat more common, as though as a result of cultivation
some of the safeguards which nature had established against the inter¬
breeding of species had been broken down, but here again the sum total
of such crosses is small.

These considerations make it the more interesting and significant when
we find such a divergent form of plum as this so-called wild peach hybrid¬
izing so freely with the local forms with which it comes in contact.

How numerous these hybrids may actually be is only a matter of con¬
jecture, and only a close survey of the entire region of occurrence of
Prunus texana can disclose this. The detection of these at any stage of
active growth is rendered comparatively easy by the strongly marked
characters of this species. The narrow, pubescent, strongly glandular-
serrate leaves, as well as the pubescent calyx cup with glandular-serrate
lobes, added to the notable character of the pubescent fruit with its
peculiar pile-covered stone, all help to render one of this class of hybrids
conspicuous and unmistakable. Three of the more striking forms in
three widely separated localities had been noticed and taken into culti¬
vation years ago by observant ranchmen interested in fruit growing.
Systematic search by the writer and other observers, though for only a
few days and over a very limited area, disclosed the other eight recorded.

It is significant that in five out of seven of the most important regions
where the wild peach is associated with the wild species of plums these
spontaneous crosses have been found. In these same sections hybrid
forms between the numerous species of true plums are rare or have not been
detected. More conclusive evidence can hardly be offered that Prunus
texana crosses with a number of species of the true plums with unusual
readiness, far more readily than these species cross among themselves.
It is on account of this fact and the promise which it holds out to enter¬
prising plum breeders that it has been thought worth while to describe
in rather minute detail a number of these natural hybrids, including
several which as horticultural varieties have little or no value.

The first of these varieties was learned of during an exploring trip
around Kingsland, Llano County, in March, 1910, and through the kind-

1 Willard, S. D., and Bailey, L. H. Notes upon plums for western New York. New York Cornell Agr.
Exp. Sta., Bui. 131, p. 170, 1897.

Nov. io, 1913

Pubescent-Fruited Species of Prunus

159

ness of Mr. Henry Smith the writer was shown a group of bushes located
on the Smith ranch near the foot of Pack Saddle Mountain, about 6 miles
from Kingsland. These had been known to a few settlers in the neigh¬
borhood for many years and the fruit had been carefully gathered on
account of its value in making preserves and jam. The “ Llano ” variety,
named and propagated for distribution by Mr. L. Miller, a nurseryman
at Lampasas, was secured in this neighborhood and is so nearly identical
with those seen on the Smith ranch that a separate description is scarcely
necessary.

The next group was located near the south line of Llano County not
far from the Llano and Fredericksburg Road. Mr. F. M. Ramsey had
previously discovered a bush in this region which from its appearance he
believed to be a hybrid of the wild peach and a plum. On a trip with the
writer in search of this plant two more were found in the same neighbor¬
hood. These are described in succeeding pages under the names “ ‘ Wil¬
low/' “Sumlin,” and “Holman.” They are of considerable interest as
botanical hybrids showing the potency of the species Prunus texana
rather than for the quality of the fruit produced.

Having been informed that at Valley Springs, about 12 miles northwest
of Llano, a farmer had wild peaches growing in his garden and that with
cultivation they grew as large as plum trees, another group of hybrids
was suspected. A visit was accordingly made to the farm of Mr. N. F.
Gephart, an early settler in the Valley Springs district, in whose orchard
several plum trees just ripening fruit were found to show undoubted
evidence of Prunus texana origin. Yet three clearly distinct varieties
could be detected. The two which are described as the “Gephart” and
the “Johnson” are interlocking trees which Mr. Gephart states he found
in a wild state in clearing the ground for the orchard more than 20
years ago.

The history of the Bolen variety, with two or three others in the
garden, is rather obscure. Mr. Gephart states that there used to be a
tree of this character, long since disappeared, on a near-by farm known as
the Bolen place. He is of the impression that seeds from this original
Bolen fruit were planted in his garden and produced one or more trees,
which bore well for a number of years but are now dead. There are at
present several trees very similar in character. Whether they are from
sprouts of the original seedlings of the Bolen tree or from a second
generation of seedlings Mr. Gephart is uncertain. Apparently the first
cross of Prunus .texana was the original tree on the Bolen farm, from
which the seedlings in the Gephart garden originated.

The following year, 1911, information was received of “a plum with a
skin like a peach” growing in an orchard near Lavemia, Wilson County,
20 miles southeast of San Antonio, and on a visit to that place two small
trees were found on the farm of Mr. W. J. Stuart, who reported that he
had found a little group or thicket of these trees in a draw when clearing

i6o

Journal of Agricultural Research

Vol. I, No. i

part of his farm. These two had been transplanted to his plum orchard
and the others grubbed up. Though not perfectly identical, these trees —
though small, they were perfect trees in form — are so similar as to make it
superfluous to give separate descriptions. The more perfect of these
was selected for description and named “Stuart” for the owner. (PI. X,
figs, i and 2.) Its fruit is among the best produced by hybrids. Two
points worthy of note about this variety are that the flowering followed
the opening of the leaves and that there is a tendency to suckering or
root-sprouting.

A few days later, in a trip along the Hilderbrand Road in company
with Mr. R. E. Blair, two more hybrid varieties were found in a field of
the Whittaker Ranch. Both were small trees, evidently sprouted from
an older growth broken down. The flowering season had passed and a
small setting of half -grown fruits was noted. Later in the season Mr.
Blair returned and found a few of these ripe, but a detailed description
was not secured. It is a medium-sized, dull-red fruit of only fair quality.
The variety near the fence on the pike road was designated as the Hilder¬
brand and the one in the field as the Whittaker. On a later trip Mr. Blair
located another hybrid tree in the same neighborhood, a detailed
description of which has not been secured.

It will be noticed that in the descriptions of these hybrids no attempt
has been made to name other parents than the wild peach {Prunus
texana)y which dominates them all. There are characters which indicate
that at least three other species as parents must be reckoned with. The
northern portion of the range of this species is also the home of a number
of species of typical American plums.

Prof. Sargent has but recently described two new species from this
territory, and it is probable that others are pending. No less than eight
species of Prunus of the plum class have been credited to this territory,
several of which are with difficulty distinguished from each other. The
hopelessness of determining the other parents of these hybrids is immedi¬
ately apparent. We have no basis for more than a conjecture as to
which direction the cross may have taken, whether Prunus texana
furnished the pollen or was the pistillate parent.

The only hint we can get in this direction is from the work of Mr.
Ramsey, referred to later. He made use of pollen from the Wild Goose
plum, without removing the stamens from the flowers of the wild peach
and secured four hybrid trees out of a number of fruits set on the pro¬
tected branch. All of these four show Wild Goose characteristics in
their flowers.

The grouping of a number of closely similar varieties, as in the case of
the two Gephart trees, the Stuart group, and those on the Hilderbrand
Road, suggests that a bush of wild peach may have received visits of bees
carrying wild-plum pollen and that a number of fruits of this pollination
germinated under or near the parent bush.

Nov. io, 1913

Pubescent-Fruited Species of Prunus

161

DESCRIPTIONS OF HYBRIDS

Prunus hortulana (Wild Goose) X texana,

Hort. var. Ramsey.

A rather ragged branched tree about 2 meters high, with yellowish brown pubescent
twigs of new growth. Leaves ovate lanceolate , acuminate at apex, rounded or broadly
cuneate at base, serrate or doubly crenulate serrate, with short glandular teeth; upper
surface dull with scattered short hairs; lower surface grayish green, silvery tomentu-
lose; petiole stouter than in most of the hybrids, 5 mm. to 10 mm. long, tomentose;
stipules narrow, acute, glandular toothed.

The flowers, which appear before the leaves, about the middle of March, are white,
about 8 mm. broad, borne in three or four flowered umbels on slender, pubescent
pedicels. The calyx is pubescent, lobes pubescent on both surfaces, margins
glandular.

The fruit, ripe about June 15, is globose, 2 to 2.5 cm. long, the rather thick dull-red
skin sparingly tomentulose, the thin reddish flesh clinging to the velvety coated pit,
which is turgid , oval, pointed at either end , and with a broad ventral ridge ; the pedicels
are 8 to 10 mm. long and stouter than in most of the hybrids of the species. This fruit
is acid, rather austere, but of value in making jellies, marmalades, etc. The origina¬
tor, Mr. Ramsey, states that it is a remarkably regular bearer. It seems to thrive well
on a strongly calcareous soil and has been grown to a good size worked on peach stock.

Primus texana hybrid.

Hort. var. Llano.

A low, ragged bush, 1 to 1.5 meters high, as it occurs in thickets in the stony pastures
in Llano County, where it was first observed more than 30 years ago and where it
spreads slowly by means of root sprouts. Worked on peach stock the twigs of young
growth become long, slender, and pendulous with little disposition to spiny branches.
The twigs of young growth are reddish brown, thinly pubescent.

Leaves elliptical or ovate elliptical, apex acute or narrowed and shortly acuminate,
base rounded or broadly cuneate ; margin serrate or doubly serrate ; the teeth glandu¬
lar tipped; the upper surface dull green, with scattering short silvery hairs; russety
green with thin pubescence beneath ; 3.5 to 4 cm. long, 1.5 to 2 cm. broad; the midribs
yellowish brown; slender petioles about 7 mm. long; stipules 3 to 5 mm. long, narrow,
acute, coarsely glandular toothed. The flowers appearing with the leaves are white,
5 to 8 mm, broad; calyx tube campanulate, pubescent; lobes short, broadly ovate, with
glandular teeth and hairy inner surface ; petals obovate with short claw.

The fruit, ripening in June, is globose, a little compressed, 2 cm. in diameter; color
dull red; skin rather thick, coated with a thin, fine pubescence; flesh netted, clinging
to the pit, which is turgid; oval, obtuse at base and apex, coated with velvet pile;
pedicel short. This fruit, produced in great abundance, is of a sharply acid flavor,
but is highly esteemed for domestic use.

Prunus texana hybrid.

Hort. var. Willow.

A willowy shrub, 1 meter high, profusely branched, the branches angled at nodes,
long, slender, tapering; young growth greenish brown, pubescent, but becoming
smooth iron gray with age.

Leaves ovate lanceolate; apex acute; base rounded; margin finely and evenly
glandular serrate; upper surface dull green with scattered hairs; under surface grayish
green with a thin silvery pubescence; 3 to 4 cm, long, 1.5 to 1.7 cm. wide; venation
prominent; petioles 4 to 5 mm. long; pubescent; fruit solitary as far as seen, a small
roundish plum with the surface covered with scattering hairs; stalk 3 to 4 mm. long;
not seen in mature condition. While an evident hybrid with distinct plumlike

Journal of Agricultural Research

Vol. I, No. 2

162

characters, this variety retains more of the Prunus iexana characters than any other
hybrid noted. But one plant discovered, south of Big Sandy Creek, Llano County,
Tex.

Prunus texana hybrid.

Hort. var. Sumlin.

An erect, slender-branched shrub, with grayish brown bark on old wood and slender,
yellowish brown pubescent twigs of young growth.

Leaves ovate elliptic, acute at apex, rounded or broadly wedge-shaped at base,
serrate with glandular teeth; the upper surface dull green, with short scattered hairs;
lower surface grayish green; hairy tomentose; 4 to 5 cm. long; midrib rather conspicu¬
ous; petiole short; stipules 3 to 4 mm. long, narrow, acute, glandular toothed.

Fruit a small, roundish, pubescent-coated plum, upon a stalk 4 to 10 mm. long.
Not seen mature, but described as red in color and a desirable fruit, ripening somewhat
later than the Prunus texana parent. Some of the characters in this variety suggest
that the cross may have been derived from a local wild plum usually classed as P.
americana var. lanata Sud worth, though perhaps an undescribed species. Trees of
this form occur in the same field and, while flowering a little later, overlap P, texana
in blooming period.

Prunus texana hybrid.

Hort. var. Holmann.

An erect-growing shrub 1 to 2 meters high, of irregular branching habit, inclined to
be spiny. Young growth slender, yellowish brown, with thin pubescence; older wood
iron gray.

Leaves 3 to 5 cm. long, 1.5 to 2 cm. broad, ovate lanceolate, with rounded base and
acute apex; margin finely glandular serrate; upper surface with scattered short hairs;
lower thinly pubescent; petiole 4 to 6 mm. long.

Fruit a small oval plum with a thinly pubescent surface, borne singly or in pairs;
stalk 6 to 10 mm. long; calyx sometimes persistent. Described as being of poor
quality. Found in a scattering group of small thickets, indicating that it has ability
to spread by root sprouts.

Prunus texana hybrid.

Hort. var. Gephart.

A tree 2.5 meters high, with numerous slender semipendulous branches; young
growth reddish brown, finely pubescent; older wood silvery gray or iron gray.

Leaves narrowly elliptical, approaching oblong; apex rounded or acute, finely
doubly serrate with minute glandular teeth; base rounded or broadly wedge shaped;
upper surface dull green, covered with scattering short hairs; lower surface ashy gray
green, finely reticulated, silvery pubescent; 3 to 4 cm. long; stipules 2 to 3 mm. long,
slender, acute, glandular toothed.

Fruit borne in great profusion, smooth, plumlike in appearance, oval, 2.5 cm. long,
dull yellow, with slight pubescence; stalk 3 to 5 mm. long; a juicy fruit, the rather soft
flesh clinging to the pilose pit much as in the original species, somewhat subacid and
lacking in quality. The earliest ripening of any of the Prunus texana hybrids so far
noted (May 13 to 18).

Prunus texana hybrid.

Hort. var. Johnson.

This variety was found growing interlocked with the Gephart, but is more upright
and stiff branched in habit and quite distinct. Young twigs reddish brown, slightly
angled at the nodes, sparingly pubescent; older growth grayish brown or iron gray.
(PI. X, fig. 3.)

Nov. io, 1913

Pubescent-Fruited Species of Prunus

163

Leaves narrowly elliptical or obovate elliptical, rounded at the apex, rounded or
tapering at the base; margin finely doubly glandular serrate, dull pale green set with
scattered hairs above, ashy green, thinly pubescent beneath, 3 to 4.5 cm. long, 1 to 1.5
cm. or, rarely, 2 cm. broad; the midribs and slender petioles, which are 1 to 2 cm.
long, are dull purplish; stipules 2 mm. long, slender, acute, glandular toothed.

Flowers not seen.

Fruit in close bunches, single or paired, 2 to 2.3 cm. long, oval, slightly compressed,
covered with a fine, soft pubescence; stalk slender, 1 cm. long, pubescent, inserted in
a very slight depression. Skin dull greenish yellow, tough; flesh greenish yellow,
acid, flavor better than that of the Gephart, but not a fruit of high quality; stone
oval, flattened, acute at apex, having a soft, short, velvety pile of the Prunus
texana type.

Prunus texana hybrid.

Hort. var. Bolen.

A compact, pendulous-branched tree, about 2.5 meters high, with finely pubescent,
brown twigs of young growth.

Leaves broadly elliptical, narrowing abruptly to an acute apex; base rounded or
broadly wedge shaped; margins finely glandular serrate, upper and lower surfaces
with scattered silvery *hairs, scarcely amounting to a pubescence, 4 to 4.5 cm. long,

2 to 2.5 cm. broad, the yellowish brown midrib passing into a slender hairy petiole,
5 to 7 mm. long.

Fruit 2,5 cm. long, 1.5 to 2 cm. broad, oval, slightly oblique, and tapering to an obtuse
apex; stalk about 5 mm. long, a little stouter than in the Gephart variety. Skin dull
yellow, rather tough; flesh yellow, rather thin because of the large seed; flavor very
similar to that of the pure Prunus texana species.

Prunus texana hybrid.

Hort. var. Stuart.

A small tree with trunk 1 cm. in diameter and spreading top 2.5 meters high and

3 meters broad; branches angular but smoother and more open than in Prunus texana;
bark grayish brown. The trees show some tendency to spread from root sprouts.

Leaves ovate elliptical, rounded or broadly pointed at apex; cuneate at the base;
serrate or doubly serrate, with fine glandular teeth; dull green with fine scattered
hairs on the upper surface; grayish green, finely pubescent beneath; 3 to 3.5 cm. long;
a conspicuous midrib passing into a short, dull red petiole; stipules minute coarsely
glandular toothed.

The flowers, which appear later than the leaves, borne singly or two or three in a
fascicle, are about 6 mm. in diameter, on slender hairy pedicels from 4 to 8 mm. long;
calyx tube narrowly campanulate, surface sparsely covered with short hairs; lobes
elliptical, with scattered glandular teeth and fringed with fine hairs; inner surface
with scattered hairs; petals thin, white, broadly obovate, with a short claw; ovary
velvety pubescent, but style smooth. Mature fruit oval, about 2.5 cm. long, with an
acute cavity around the short stalk; dull yellow, with velvety surface and mellow,
luscious, highly flavored flesh. Seed oval, turgid, with heavy velvety pile.

Prunus texana hybrid.

Hort. var. Hilderbrand.

A small tree with slender, erect, rather angular branches; bark smooth, grayish.
Leaves obtuse or rounded at the ends; finely, sometimes doubly, serrate, with minute
glandular teeth; dull green, with scattered fine hairs on the upper surface; grayish
green with hairs more numerous below; 3 to 3.5 cm. long, 0.7 to 1 cm. wide; midrib
narrow, tinged with dull purple at the base ; petiole short, slender, pubescent; stipules
2 mm. long, narrow acute, glandular serrate.

164

Journal of Agricultural Research

Vol. I, No. a

Flowers not seen, apparently opening with the leaves.

Fruit oval, velvety, stalk 1 cm. long, slender, nearly glabrous. (Mature fruit not
seen; described as being red.)

Primus texana hybrid.

Hort. var. Whittaker.

A shrub of treelike form, 2 meters high; branches regular or somewhat angled at the
nodes, long, slender, with few spines; bark smooth, iron gray or brown.

Leaves thin, narrowly ellipticar, acute at both ends, doubly serrate with minute
glandular teeth; dull green with minute scattered hairs above, grayish green, more
abundantly hairy below; 4 to 5 cm. long, 1 cm. to 1.3 cm. broad; petiole slender,
pubescent, dull purple, 0.5 to 1 cm. long; stipules lancelinear, acute, glandular serrate,
about 2 mm. long.

Flowers, appearing with the leaves, small, on slender hairy peduncles about 6 mm.
long (petals not seen); calyx tube narrowly elliptical, fringed with fine silvery hairs
and sparsely coated with hairs on the inner surface.

Fruit borne singly or in pairs, oval, finely pubescent (not seen mature; color said
to be red), the pubescent stalk 6 to 8 mm. long. ♦

THE NEVADA WILD ALMOND #

The wild almond (Pis. XI and XII, figs. 1 and 2), the most striking
of all the dry-fruited members of the plum family occurring in the
United States, was first described by Asa Gray from specimens sent him
by Dr. C. E. Anderson, collected near Carson, Nev., 1863-1866, and was
named in honor of Dr. Anderson.

From field notes and abundant herbarium material collected by the
writer in person or supplied by Mr. E. W. Hudson, important characters
heretofore unnoted are brought out and this species is redescribed as
follows :

Primus andersonii Gray.

Primus andersonii Gray, Proc. of Amer. Acad., v. 7, p. 337-338, 1868.

Amygdalus andersonii (Gray) Greene, PI. Franc., pt. 1, p. 49, 1891.

Emplectodadus andersonii (Gray) Nelson and Kennedy, Muhlenbergia, v. 3, p. 139, 1908.

Illus., Schneider, C. K., I^aubhk., p. 598, fig. 335, d, e.

A spiny, much-branched, interlocking shrub 1 or 2 meters high, or, rarely, more
smooth, erect, and treelike, reaching 3 meters or over; bark of young branches grayish
green to reddish or yellowish brown, glabrous, on older wood breaking into coarse,
dark-gray scales. The leaves are convolute in the bud, broadly or narrowly spatulate,
with rounded or acute apex and short petiole, finely serrulate or entire, often with a
pair of small glands near the base, 1 to 4 cm. long; yellowish or grayish green, leathery,
glabrous, or faintly pilose at the base; stomates present in the upper epidermis.

The flowers, appearing with the leaves, are perfect, 1.5 to 2 cm. in diameter, on
slender glabrous pedicels, 1.5 cm. or less in length, solitary or fascicled; calyx tube
short, campanulate, leathery, glabrous, or rarely with pedicel and calyx cup puberu-
lous; lining nectariferous; the lobes triangular with ciliate margins, often persistent
on mature fruit; petals from pale to deep-rose color, or rarely white, oval, 6 to 10 mm.
long, narrowing abruptly to a short claw; stamens 20 to 30; style equal to or longer than
the stamens; glabrous or only the lower one-fourth hairy; ovary pubescent.

Fruit roundish or obliquely unsymmetrical, compressed, often with a marked
winglike ventral expansion, abruptly rounded to an apiculate apex; base distinctly

Nov. io, 1913

Pubescent-Fruited Species of Prunus

165

necked, 1 to 1.8 cm. long, dull grayish or greenish yellow with thickly pubescent
surface, usually with prominent, coarse, reticulate venation as it dries; the thin flesh
dry, leathery, and astringent, or, rarely, more succulent and with edible qualities,
usually splitting along the ventral suture at maturity after the fashion of an almond.

Stone roundish, unsymmetrical, turgid or compressed, the narrow dorsal wing
having a shallow groove; the ventral wing often much expanded; has an acute central
ridge usually flanked by parallel ridges and obscure reticulate veins; surface smooth

Fig. 4. — Prunus andersonii Gray: A, Petal, X 3; B, section of a flower, X 3;
C, calyx showing ciliate margins, X 3; dried fruit times natural

size; F , G, stone, i1/* times natural size.

or obscurely or decidedly pitted ; apex rounded to an acute point, base with a more
or less thin, attenuated neck; kernel small, pointed, grooved in some varieties, edible,
often strongly flavored with prussic acid. (Fig. 4.)

This species is one of the most distinctive of those commonly included
in* the Emplectocladus group.

On mountain sides and dry foothills of eastern California and Nevada
it is a squarrose, much-branched and spiny shrub, 1 to 2 meters in height
and diminishing to 0.5 or 0.7 meter at its upper limit of growth. In
more favorable situations, along the shore of Pyramid Lake and other
localities where better soil and a more constant supply of water occur, it
becomes a large shrub or even a small tree. Forms appear reaching
over 3 meters in height, nearly free from spines, with clean, free growing
branches and have the appearance of young peach or almond trees.
(PI. XII, figs. 1 and 2.) Well-marked varietal forms are found, not only

Journal of Agricultural Research

Vol. I, No. 2

1 66

in habit of growth and branching, and color and texture of bark, but
in size and color of flowers and character of fruit. 1 2

One variety was noted with fruits of unusual size and having a fleshy
development of the pericarp instead of the characteristic dry, leathery
coating.

A very strongly developed taproot is a characteristic of this species as
well as of several others of the group. This has been very noticeable in
growing seedlings. Seeds stratified in sand and exposed to open condi¬
tions of a severe winter of Washington, D. C., made a vigorous germination
early in March, sending down strong taproots, while the tops were but
two or three leaves above the ground. This must be recognized as an
adaptation which has enabled them to survive under peculiar local con¬
ditions. Where wild-almond thickets occur there can usually be traced
at a depth of i or 2 meters a layer of soil or sand where more permanent
moisture is afforded than prevails near the surface. After the taproot
has penetrated this layer small branches spread out into it and the
moisture made available enables the plant to survive drought and heat
which would have caused it to perish if supported by superficial roots.

The range of occurrence of this species is shown on the map (fig. i)
and is a region of such scant rainfall that little agriculture is possible
without irrigation. Taking Carson City, Nev., as a typical station,3 the
mean annual precipitation is slightly above io inches, falling as low as
5 inches in years of extreme drought. The 2 feet or more of snow forms
a considerable portion of the annual moisture, the mean precipitation
from April to September, inclusive, being but 2.4 inches. With summer
heat occasionally reaching ioo° F., and the average winter temperatures
of —20° F., some idea of the hardiness and drought resistance of this
species can be formed.

THE DESERT APRICOT

This striking apricotlike species occurs only in certain out-of-the-way
places in southern California. (Pis. XII, fig. 3, XIII, and XIV, fig. 1.)

Confined chiefly to a narrow zone on the desert side of the San Ber¬
nardino and San Jacinto Mountains, the only frequented spots of its
habitat are the village of Palm Springs at the foot of San Jacinto Peak
on the south and the almost deserted hamlet of Banner at the foot of
the mountains and just above the border of the desert below Julian in
San Diego County.

, In the Gray Herbarium the type specimen sheet has mounted upon it
a specimen bearing the label i( Prunus subcordata , Bth.,” and in print,

1 Mr. E. W. Hudson, of the Office of Crop Physiology and Breeding Investigations, while doing cooperative
work at Wadsworth Agency, made numerous collections of this species in 1910 and noted that the flowers
ranged in color from pale pink to a deep-rose color, and also varied greatly in size.

2 Henry, A. J. Climatology of the United States. U. S. Dept. Agr. Weather Bureau, Bulletin Q,

p. 920, 1906.

Nov. io, 1913

Pubes cent- Fruited Species of Prunus

167

“ Flora of Southern California, &c. Coll, by C. C. Parry and J. G.
Lemmon, 1876.” In the upper right-hand comer of the same sheet
is a specimen of very different appearance bearing the label, “Fre¬
mont’s Expedition to California, 1845-7. 3 70 — 1846,” and in pencil

“New” (in the hand of Dr. Asa Gray). At the bottom of this sheet
is the penciled label “P. Fremonti Watson, n. sp.”

The specimen first cited and referred to as collected by Cleveland in
Oriflamme Canyon bears the label “P. subcordata , var. eriogyna ,” but it
has also beneath a subsequent label the penciled inscription, “P. Fre -
monti Watson, n. sp.”

It has been noted by Mr. W. F. Wight, of the Bureau of Plant Industry,
in a memorandum placed upon the specimen in 1910, that the Fremont
specimen is Prunus subcordata , a determination supported by the glabrous
pistils and the leaf characters.

Dr. Watson clearly had before him three specimens upon which he
based his description of the new species and to which he attached the
name. While he cites the Fr6mont specimen last we may readily pre¬
sume that it was because of its lacking a definite locality label, which
the first and second citations possessed. Having incorrectly included
it in the type material, however, and having given the name “ Fremonti”
to the species, this specimen, according to the American Code of Botan¬
ical Nomenclature (section 4, canon 14, a), becomes the type specimen.
Prunus fremonti Watson, then becomes a synonym of P. subcordata
Benth., leaving the species bordering the Colorado Desert unnamed.
The name Prunus eriogyna is accordingly proposed for this species.

These two species seem to have been subject to much confusion by
the earlier collectors.

Dr. Torrey in the Botany of the Mexican Boundary Survey 1 refers
specimens collected by the expedition ac San Felipe to 11 Prunus sub¬
cordata, Benth., PI. Hartw.,” yet his description tallies well with P.
eriogyna, and the San Felipe locality renders it probable that he had
this species before him.

T|ie specimen collected by Fremont is undoubtedly Prunus subcor -
data[ Benth., the type of which was collected by Hartweg somewhere
about the upper waters of the American River in the latter part of
April, 1846.2

By an interesting coincidence Col. Fr6mont in his Memoirs, p. 476,
mentions camping March 26, 1846, at the ranch of the same Mr. Cordua
where Hartweg made his headquarters in the Sacramento country.
The month of April Fremont spent in the region tributary to the Sacra¬
mento River, now included in Butte, Tehama, and Shasta Counties,

1 Torrey, John. Botany of the boundary. Emory, W. H. Report of the United States and Mexican
Boundary Survey ... v. 2, Washington, 1859, p. 63.

2 Hartweg, T. Journal of a mission to California in search of plants. Jour. Roy, hort. soc. [Eondon],
v. 3, p. 221, 1848.

Journal of Agricultural Research

Vol. I, No. 2

1 68

and the date of Hartweg’s collection, made at a considerable altitude in
the foothills, suggests the probability that the Fremont specimen was
secured in the upper waters of one of the many mountain tributaries
which he visited.

From abundant material collected near Palm Springs and in the
Banner Canyon of San Diego County and from field notes covering
several seasons’ observations the following detailed description of this
new species has been drawn.

Prunus eriogyna, n. sp. (Fig. 5.)

Prunus fremonti1 S. Watson, in California, Geological Survey, Botany, v. 2, Cambridge (Mass.), p.
442-443, 18S0,

Amygdalus fremonti (S. Watson) Abrams, in Bull. N. V. Bot. Gard., v. 6, no. 21, p. 385, Sept., 1910.

Illus., Schneider, C. K., Laubhk., Lfg. 5, P- 598, fig. 33s, u, v.

A spiny, intricately branched and angled shrub reaching 4 meters in height. Twigs
of young growth glabrous, bright reddish brown, becoming silvery gray or brown
with age. Bark on old stems black, breaking into thin plates or scales.

Leaves variable, lanceolate, ovate or orbicular, or sometimes broader than long,
rounded or cordate at the base, narrowing abruptly to a short acute apex or often
rounded or obtuse; glandular denticulate, usually with one or more larger glands
near the base or rarely on the petiole; both surfaces pale grayish green, shining above,
firm, sometimes leathery; midrib and veins prominent on under surface; stomatesin
both upper and lower surfaces; 1.5 to 3 cm. long, 1.5 to 2.5 cm. or more broad; petiole
6 to 8 mm. long; stipules minute, narrowly acuminate, glandular denticulate.

The perfect flowers, borne in small umbels and having a faint, agreeable odor are pro¬
duced in great profusion, appearing from January to March, according to rainfall, when
the leaves are partially developed. Imbud they are white, salmon, or rose pink. Ex¬
panded they are usually 6 to 8 mm. in diameter, reaching 18 mm. in some forms, on
slender pedicels 6 to 12 mm. long; calyx tube short, campanulate; outer surface
glabrous or thinly pubescent; inner covered with a salmon or rose colored pigment;
lobes oval, half as long as the petals, finely pubescent on inner surface, glandular
ciliate, often hanging loosely in a dried condition around the pedicels of the mature
fruits; petals white, pink, or rose, 3 to 6 mm. long, oval, incurved at apex, base round¬
ing to a stout claw; stamens about 24 to 30, many imperfect; ovary and lower portion
of the style finely pubescent; stigma but little expanded.

The fruit, which ripens in May, is in appearance a small apricot, 1 to 2 cm. long, sub-
globose, ovoid or oblong ovoid, sometimes oblique, slightly or decidedly compressed;
apex mucronate; skin puberulent, dull yellow or greenish yellow, often with a dull-
rose flush, with a well-marked ventral suture along which the thin astringent flesh
opens in ripening, sometimes allowing the stone to drop, while the desiccated flesh
remains attached to the peduncle; stone smooth or slightly roughened, usually flat¬
tened or somewhat turgid, obtuse at both ends with a well-marked dorsal furrow and a
thick ventral expansion along the middle of which is a low, acute ridge separated by
smooth, narrow furrows from two obtuse parallel ridges; often one or more pairs of
obscure veins extend from the base and branch along either side; stony walls thick,

1 “ P. Fremonti. A spiny glabrous densely branched shrub or small scraggy tree (15 feet high) with short
branchlets: leaves small (4 to 8 lines long), thin, ovate or roundish, on short slender petioles, denticulate:
flowers appearing with the leaves, solitary or somewhat fascicled, 5 or 6 lines broad, on pedicels 2 or 3 lines
long: calyx lobes ciliate: ovary densely pubescent; style elongated: stone oblong, turgid, rounded on one
side and with a broad ridge upon the other, 5 lines long.

“Coast Ranges of Southern California, Oriflamme Caflon, San Diego County (D. Cleveland)', San Ber¬
nardino Mountains, Parry & Lemmon, n. 108, 1876. Also collected by Fremont in 1846, locality uncertain.
Flowering in March; fruit probably with little pulp."

Nov. io, 1913

Pubescent-Fruited Species of Prunus

169

kernel small, strongly flavored with prussic acid. Type specimen in United States
National Herbarium, C. P. B. No. 1155. Merotypes cut from the tree that yielded the
type specimen have been sent to a number of other herbaria.

The type locality of Prunus eriogyna is along the watercourse in the
boulder talus at the mouth of Tahquitz Canyon at the southern base of
the San Jacinto Mountain, near Palm Springs, Riverside County, Cal.
It is also found on dry
talus slopes in Andraeas,

Murray, and Palm Can¬
yons, along the trail to
Van Deventer Flats be¬
low Santa Rosa Peak,
and up the rocky slopes
of the San Jacinto
Mountains to an al¬
titude of over 2,000
feet, growing in barren
soil and crevices of
rocks, being apparently
extremely xerophytic.

Its range is from the
southern slopes of the
San Bernardino Moun¬
tains southward along
the desert slopes of the
San Jacinto Mountains
to San Diego County,
and into Lower Cali¬
fornia.

The plumlike appear¬
ance of the wood, es¬
pecially of the younger
growth, and perhaps a sprinkling of roundish or oblong green pubescent-
coated fruits, would excite an inquiry that would bring out the names
“desert almond” or “wild apricot.”

Fig. 5. — Prunus eriogyna , n. sp.: A , Section of calyx, X 3; B, detail
of portion of calyx with petals, from outside, showing glandular
dilation of lobes, X 3 ; C, twig showing angular habit of branching,
leaves and fruit attached, H natural size.

ADAPTATION TO DESERT CONDITIONS

The adaptations of Prunus eriogyna to the peculiar conditions which
prevail on the desert slopes of mountains are worth noting. The rain¬
fall, slight as it is, really governs plant activities, and vegetation becomes
most nearly dormant during the summer months. The rains consist of
rare torrential downpours in August and light rains from October to May,
but are nearly confined to the period from December to March, the entire
volume ranging from less than an inch to 9 inches and a fraction annually.
With warm winter days and the temperature at night falling but little

170

Journal of Agricultural Research

Vol. I, No. 2

below freezing the vegetative activity in many species of plants that
have become dormant during summer drought may be resumed at any
time when a sufficient supply of water is afforded.

In the case of Prunus eriogyna a copious November rain may start the
favorably located bushes into activity, so that a small percentage of the
many flower buds will open in January. Cool nights and light frosts may
destroy a portion of these buds, but some will set fruit, indicating a fair
degree of hardiness for this species. At the time of the main flowering
in March there may be a few scattering, nearly mature fruits, which
escape the numerous plum curculios and furnish a small supply of seeds
for germination should the rainfall be inadequate to mature the main
crop of fruit.

In seed germination this species differs strikingly from ordinary
apricots or plums. Germination is rapid, the plants appearing above the
ground in from 8 to io days. As an example, in a pot of seeds sown in
sandy soil in a greenhouse on July 31a number of plants were above the
soil on August 6. One with the plumule 1 cm. long had already sent
down a taproot of 9 cm. In desert conditions with fruit ripened in May
germination is necessarily deferred till the autumn or winter rains set in,
when the quick germination habit is essential to its success. Getting its
roots down to a zone of permanent moisture, however slight, is the
necessary thing if the seedling is to survive the dry, hot summer that
follows. A sufficient leaf expansion to afford the needed root growth is
all that is necessary and more would only hasten transpiration and waste
the limited supply of moisture.

That even the best forms of Prunus eriogyna are far from having the
quality of cultivated apricots is evident from the appearance of the
plants, but that 'this desert species of the Pacific slope has very close
affinities with the true apricot of the Orient can not be doubted. The
apricot relationship of P. andersonii , with which is placed P. eriogyna , is
not so evident, yet its convolute leaves, fascicled flowers, and slender-
stalked fruit with a slight tendency to be fleshy will ally it to the Prunus
dasycarpa type of the apricot more nearly than to the almond.

THE CALIFORNIA DESERT ALMOND

The desert almond, also called the “wild peach” and “wild almond,”
occupies a range much farther south and east than that of the Nevada
wild almond, Prunus andersonii. It overlaps the southern range of
P. andersonii in Nevada and eastern California and that of P. eriogyna
in southern California. It has been collected near the coast in San Luis
Obispo and Santa Barbara Counties and as far east as southwestern Utah
and northwestern Arizona. Its greatest abundance as far as studied is
along the foothills bordering the Mohave Desert in the neighborhoods
of Hesperia and Neenach at altitudes of 3,000 to 3,500 feet. The soils

Nov. io, 1913 Pubescent- Fruited Species of Primus

171

it favors seem to be from decomposed granite or mica schist. In washes
where the sands and silts from these rocks are deep an enormous root
development is made, the plants forming dense thickets of many sprouts,
reaching 7 or 8 feet in height. On granitic slopes above the washes the
plants occasionally grow with a single stem and a miniature tree-like
form (PI. XIV, fig. 2). The following description of this species is the
result of examination of many plants in the field and the study of abundant
herbarium material.

Prunus fasciculata Gray. (Fig. 6.)

Emplectocladus fasciculaius Forr., PI, Frdmont, p. io-ii, pj. 5, 1853. 1

Prunus fasciculata Gray, Proc. of Amer. Acad,, v. 10, p. 70, 187s.

Amygdalus fasciculata Greene, FI. Franc., pt. 1, p. 49, 1891.

Ulus., Schneider, C. K., Laubhk., Lfg. 5, p, 598, fig. 335, ff gf h; Forr., loc. cit.

A much-branched, scarcely thorny shrub, with many small branched stems from
a common crown or rarely with
a single stem and short stiff
branches, usually 1 or 2, rarely
3 meters high, with stems 6 to
10 cm. in diameter at the base.

The bark on young twigs is
usually puberulous or pubes¬
cent, at first pale green, dark¬
ening to reddish or silvery
brown, with conspicuous lenti-
cels; dark gray brown or nearly
black on older wood.

The leaves, conduplicate in
vernation,1 2 are borne singly on
young wood of free growth, but
are fascicled on short budlike
suppressed branchlets on older
growth. They are narrowly
linear spatulate with a mucro-
nate apex and cuneate base;
margin entire or with a few
fine serrations ; blade thin , pale
green, puberulous above and

below; 1 to 4 cm. long, 3 to 7 mm. broad; petiole short or wanting; stipules caducous,
slender, attenuate, minutely glandular.

The flowers, dioecious by abortion of stamens or pistils, are minute, solitary or
paired, sessile or very short stalked. In the staminate form the calyx tube, about 3
mm. long, is obconic campanulate, with blunt triangular teeth; glabrous or faintly

1 From incomplete material collected by Gen. Fremont this species was made the basis of a new genus
ky Forrey in 1853, the Latin description of which is rendered in English as follows:

Emplectocladus n. gen.- Calyx obconical campanulate; tube not at all contracted at the naked throat*
limb divided into five equal parts, persistent. Petals 5, erect-spreading. Stamens 10 to 13, biserial, pistils
1 to 2 (generally solitary), unilocular; ovules two, collateral, pendulous. Style very short, thick, slightly
oblique, stigma capitate. Fruit - .

California shrub, very much branched; branches rigid, spreading, subspinescent; leaves minute, spatu¬
late, Lorn subglobular buds, almost fascicular; stipules minute, deciduous; flowers subsolitary, sessile
terminal, small.

2 Only the most careful inspection of very young leaves as they emerge from the bud will discover that
they are conduplicate. Fhe adhering margins of the linear-spatulate leaves hold them in a tubular form
as they expand, giving them a rolled appearance which is accented by a slight twist.

Fig. 6 —Prunus fasciculata Gray: A , Section of staminate flower,
showing abortive ovary and minute hairs on interior of calyx,
X 3; -S, calyx cup, pistillate form, showing abortive stamens,
X 3; C, detail of calyx lobe, X 5; A fecundated ovary, X 3;
E, F,G, fruits, three forms, natural size; H, /, J, seed, dorsal,
ventral and side views, natural size.

172

Journal of Agricultural Research

Vol. i, No. 2

puberulous without and minutely hairy on the inner surface. Ten or twelve stamens
on short filaments are arranged in two series. The petals, 2 mm. long, are white,
broadly obovate cuneate, with erose margins and without claw.

In the pistillate form the calyx tube is rather more campanulate. There are minute
abortive stamens and the pubescent ovary is surmounted by a smooth style, 2 to 3
mm. long. The mature fruit, borne on a very short peduncle, is coarsely pubescent,
irregularly globose, 1 to 1.3 cm. long, having a distinct ventral ridge with a shallow
furrow through the center and two or three pairs of small concentric ridges arising
from the base and disappearing toward the rounded apiculate apex.

The thin, dry pericarp does not split as in Prunus andersonii and P, eriogyna.
The thin-walled stone is smooth surfaced excepting minute sharp ridges correspond¬
ing to those of the outer surfaces. Kernel scarcely edible because of the strong prussic-
acid flavor.

Mr. F. V. Coville seems to have been the first to notice that the
flowers of this species were otherwise than perfect. His description
contains the following paragraph :

The flowers are polygamo-dioecious, a fact which explains Dr. Gray's diffi¬
culty 1 in identifying Torrey 's plants with others subsequently collected. In the pre¬
vailingly male flowers the petals in our specimens are elliptical lanceolate, appressed
strigose on the back, 3 to 3.5 mm. long; the filaments 2 mm. and the anthers 1 to 1.2
mm. in length, while the style is 1 to 2 mm. long, and the pistil sterile. In the fertile
flowers the petals are ovate, glabrous on the back, 2 to 3 mm. long, the filaments
0.6 to 0.8 mm., the anthers 0.4 mm., and devoid of pollen, and the style about 2 mm.
long. The sterile flower is the one figured by Torrey (loc. cit., pi. v). The form
and length of the petals probably vary considerably.2

Schneider3 recognizes this and the two following species as “subdi-
dcisch * ’ (subdioecious) .

THE TEXAS ALMOND

The Texas almond, first collected by Lindheimer south of New Brauns-
fels, Comal County, Tex., “not far from Cebolo Cr./’ occurs in the north¬
west suburbs of San Antonio and occupies an imperfectly known region
southwestward to the Rio Grande and beyond 4 apparently restricted to
the limestone soil of the Cretaceous formation. (Fig. 2.)

The region of the lower Pecos near the Rio Grande is one of deep
deposits of soft cretaceous limestone rock, deeply eroded and very broken.
The soil over the hills is often very thin or the bare rock is wholly exposed.
In the broader washes some soil is beginning to collect in the form of
miniature bottom lands, occasionally overflowed by the run-off from
heavy rains. Along these washes there is sometimes a fringe of scrubby
growth of hackberry, oak, the western black walnut (Juglans rupesiris ),
the “chapote,” or Mexican persimmon (Diospyros texana ), and similar
arid land forms. It is in these situations that the Texas almond is found

1 Proc. Amer. Acad, x., p. 70 (1874).

2 Coville, F. V. Botany of the Death Valley expedition. Contrib. Nat. Herbarium, v. 4, p. 91, 1893.

3 Schneider, C. K. Illustriertes Handbuch der Laubholzkunde, Bd. 1, Lfg. 5, Jena, 1906, p. 598.

4 “ . . . Gravelly places and ravines between Devil's River and the Rio Grande; also in Chihuahua;
Parry. Bigelow." Torrey, John. Botany of the boundary. Emory, W. H. Report of the United
States and Mexican Boundary Survey ... v. 2, Washington, 1859, p. 63.

Nov. 10, 1913

Pubescent-Fruited Species of Prunus

173

rather than in strictly upland conditions, though in a few instances it
was found on high ground, where it benefited by no addition to the rain¬
fall by means of run-offs.

The Texas almond is a shrub scarcely 6 meters high in its northern
range. Where it was studied by the writer in Valverde Co., Tex., along
the limestone washes, it frequently forms thickets from 1 to 1 .6 meters
in height, with stems 2 to 3 cm. in diameter.

The dioecious habit of this plant is one of its most marked character¬
istics, when one has the opportunity of
examining the plants in large numbers
in its most favorable conditions.

The bushes bearing the staminate flow¬
ers are much more numerous than the
fruiting ones and the flowers more numer¬
ous and crowded, so that in the field it is
generally possible to distinguish the types
from a distance. The examination of a
large number of plants in flower in Val¬
verde County failed to show a single case
in which the flowers could be called
polygamo-dioecious. In no case were
hermaphrodite and unisexual flowers
found on the same plant. Not a pistil¬
late flower was found with fertile stamens
nor a staminate flower that did not have
the pistil abortive and much reduced in
size.

During seasons of drought and scarcity
of forage these bushes are browsed by
stock on these ranges. In the suburbs of
San Antonio, where the grazing of cows
has been heavy on vacant lots, these
bushes were fround cropped back to a
very small size and nearly all affected
with crown-gall.

Field study of two seasons of this spe¬
cies in flower and fruit has furnished the material for the following
revised description (PI. XV.) :

Fig. 7. — Prunus minutiflora Engelm.: A,
Section of flower of pistillate form, show¬
ing well-developed pistil and abortive sta¬
mens, X 4; B, section of flower, staminate
form, showing well-developed stamens
and abortive pistil, X 4; C, detail of calyx
lobes and petals. X 4.

Prunus minutiflora Engelm. (Fig. 7.)

Prunus minutiflora Engelm., in Gray, A., PI. Eindheim., pt. 2, Boston Soc. Nat. Hist., v. 6, p. 185, 1850^
Cerasus minutiflora (Engelm.) Gray, in PI. Wright, pt. 1, p. 68, 1852.

Amygdalus minutiflora (Engelm.) W. F. Wight, in Dudley Mem. Vol., p. 130, 1913.

Illus., Schneider, C. K., Eaubhk., Efg. 5, p. 598, fig. 335, m, n, o, p.

An erect but much-branched, angled and spiny shrub, from 0.5 to 1.6 meters high,
stems 1 cm. to 3 cm. in diameter, forming considerable thickets along limestone
slopes and washes in the cretaceous section of central and western Texas.

174

Journal of Agricultural Research

Vol. I, No. 2

Twigs of young growth often puberulous, reddish brown or silvery gray; older wood
with silvery gray or iron-gray bark.

The leaves, which are conduplicate in the bud, borne singly on young growth but
fascicled on short spurs on older wood, are spatulate or narrowly elliptical; apex
rounded, refuse or mucronate; base cuneate, entire or with one to several minute
teeth on either margin, and rarely one or two near the base, glandular tipped, firm
and leathery, pale bluish green, glabrous or faintly puberulous at the base, i to 3 cm.
long, 0.5 to 1 cm. wide; petiole short, slender; stipules 2 mm. long, acuminate, ciliate
margined.

The minute flowers, borne singly or paired, on short peduncles, are usually crowded
on short, budlike fruiting spurs. They appear with the leaves in February or March
and are minute and dioecious by the abortion of the stamens in the fruiting form and
of the pistils in the opposite form. In both types the inner surface of the calyx is finely
hairy. In the pistillate type the calyx tube is obconic, glabrous; lobes triangular,
acute; peduncle 3 mm. long, puberulous; ovary and lower portion of the style finely
pubescent. There are usually 15 or more abortive stamens. Petals white, about
2 mm. long, obovate cuneate, with sinuous or erose margins and short, stout claws.

In the staminate flowers the tube is slightly broader, the stamens 10 to 15 or rarely
16 to 20 on short filaments, usually with a stamen opposite each petal, one or two
against each calyx tooth, and an irregular number disposed on the upper surface of the
tube. The pistil is abortive and much reduced.

Fruit globose, apiculate and with shallow ventral furrow, pubescent, 1 to 1.5 cm.
long, the thin, dry sarcocarp scarcely dehiscent; the stone smooth with but a slight
furrow on ventral surface.

THE MEXICAN ALMOND

The Mexican almond was the first of this group to be described,
but to-day is the least known of all of them. Found in the high moun¬
tain regions of Mexico, it has been little collected and it is not known that
it has as yet been brought into cultivation.

Judged by the pubescent thin-fleshed fruit with its smooth, oval stone
its relationship would be considered near to the Texas almond ( Prunus
minutifiora) which crosses the border into Chihuahua, but its more
slender and less spinose twigs and especially the serrate, finely pubescent
leaves indicate that it is a quite distinct species. In 1823 Humboldt and
Bonpland found it growing in arid hills between Pachuca and Moran
(Estado de Hidalgo) at an altitude of 7,800 feet and describe it as a
shrub 3 feet high with sparse, reflexed, divergent, glabrous branches and
subangular pubescent twigs.

Parry and Palmer collected this shrub in the region of San Luis Potosi
‘ at an altitude of 6,000 to 8,000 feet, which would agree well with the
altitude at which the original specimens were collected by Humboldt and
Bonpland.

The majority of specimens in American herbariums have been col¬
lected by Mr. C. A. Purpus, of the University of California, to whom
the writer is indebted for the most recent information on the occurrence
and habits of this species.

The following description of this species is made from material in the
United States National Herbarium, specimens contained in the herbarium

Nov. io, 1913

Pubescent-Fruited Species of Primus

175

of the University of California, and material collected by Mr. Purpus for
the writer :

Prunus microphylla Hemsley. (Fig. 8.)

Amygdalus microphylla , H., B., and EL, Nov. Gen. et Sp. PI., v. 6, p. 243, pi. 564, 1823. 1

Prunus microphylla (H., B., and K.) Hemsley, Biol. Centr. Amer. Bot. v. i, p. 118, 1879.

Ulus., Schneider, C. K., Laubhk., Lfg. 5. P* 598, fig. 335, q, r, s, t; H., B., and K., loc. cit.

A low branching shrub with slender twigs destitute of thorns; puberulous on new
growth, sometimes also on wood of second year, bark greenish or reddish brown,
turning to silvery or dark gray on older wood. Leaves narrowly elliptical or on
fresh shoots broadly lanceolate; base slightly produced or cuneate; margin crenately
serrate with blunt glandular or callus tipped teeth; dull green, faintly puberulous
above ; grayish green with scattered short hairs on the lower surface ; nearly glabrous
on old growth; 1.5 to 2 or 3 cm. long; petiole short, puberulous; stipules 2 to 3 mm.
long; slender attenuate, russety, hairy with glandular teeth; stomatesnot present in
upper surface of the blade.

The flowers, appearing in April or May before or with the leaves, are solitary,
minute, and dioecious by the abortion of the stamens or pistils.

Staminate flowers sessile, with glabrous campanulate calyx tube 2 to 3 mm. long;
lobes short, triangular, with expanded base and glandular ciliate margins; tube
minutely hairy within; petals white, broadly obovate, entire or with notched or erose
margins. Claw short or wanting. Stamens on filaments 1 to 2 mm. long are 10 to 15
or 18 in two or three circles, one circle opposite the petals, one opposite the calyx
lobes near the throat, and a more or less complete circle below these. (One flower
had 15 stamens and the three circles complete.) The pistil is minute, glabrous, and
abortive.

In the pistillate form the stamens, with very short filaments, are abortive; the pistil,
4 to 5 mm. long, has the ovary and lower portion of the style pubescent; stigma
expanded.

The mature fruit is 1 to 1 . 5 cm. long, oval with about equally rounded ends, apiculate
by persistence of the style, but little compressed, densely rusty pubescent; sarcocarp

1 Their description is translated as follows:

Amygdalus microphylla. Tab. DLXIV.

Amygdalus oblong, acute, mucronate, crenate-serrulate with glabrous leaves.

Grows on arid hills , between Pachuca and Moran , alt. 1,300 hex. (7,800 ft.). (Mexico) Shrub. Flowers in
May.

Shrub 3 feet high, very much branched; branches spreading divergent, reflexed, rounded, smcoth,
glabrous, blackish; twigs subangular, pubescent. Leaves sparse, petiolate, densely fasciculate on short¬
ened branches, oblong, acute and mucronate, somewhat acute at the base, crenate-serrulate, the teeth with
glandular midrib, reticulate-veined, prominent below, membranaceous, glabrous, with scattered, very
minute scurfy dots above, 5 to 6 lines long, 2 to 2% lines wide. Petioles 1 line long, canaliculate,
puberulous. Stipules linear-subulate, serrulate- glandular below, pubescent, twice as long as the petiole.
Flowers axillary, solitary, with very short peduncles, scarcely as large as the flower of Amygdalus incana ;
peduncle scarcely half a line long, thick, glabrous, subtended by several imbricate, ovate, purplish, glabrous
bracts. Calyx (figs. 1 to 3) subturbinate-campanulate, limb 5-parted, reddish, glabrous, later split around
above the base and deciduous, with ovate lacinise, denticulate-glandular at the margin, 3- veined, equal,
reflexed. Petals (fig. 5) five, inserted in the throat of the calyx, alternating with the lacinise of the latter
and twice as long, unguiculate, obovate, entire (2-parted fide Bonpl.), white, glabrous (this I saw formerly
in specimens no longer at hand), fallen from the specimens at hand. Stamens (fig. 4) about 14, slightly
shorter than the lacinise of the calyx; of these 4 inserted in a tube towards the middle; 10 around in a border
(five opposite the lacinise of the calyx, five opposite the petals). Filaments subulate, glabrous. Anthers
subrotund, affixed dors^lly, exposed (figs. 6 to 7). deeply trisulcate in front, bilocular, longitudinally
dehiscent on the inside. Ovary (figs. 8 and 9) free, sessile, oblique ovate, somewhat compressed, shorter
than the calyx tube, sericeous, unilocular (fig. 10); ovules (fig. n) two, ovate, side by side, suspended below
the apex, pendulous. Style terminal, filiform, exserted, glabrous. Stigma (fig. 12) dilated, peltate. Fruit
(not seen) globular, monospemious (fide Bonpl.).

Varies in a 6-parted calyx.

176

Journal of Agricultural Research

Vol. I, No. 2

thin and dry, probably slightly fleshy when nearly ripe, splitting tardily along the
ventral suture. Three or four pairs of shallow concentric furrows sometimes radiate
from the base. Stone rounded oval with apiculate apex, smooth, with a slight ven¬
tral ridge and a faint dorsal
furrow.

Prunus microphylla is
intermediate between P.
fasciculata and P. minu-
tifiora , but differs from
both in the glandular leaf
serrations. The absence
of stomates in the upper
surface is a noticeable
difference from P. fasci¬
culata and would ally this
species most closely with
P. minuiiflora.

HAVARD ’S ATMOND
Prunus havardii W. F. Wight,
n. comb.1 * * * (PI. XVI.)
This species, the least
known of the group, was
recently described by Mr.
William Franklin Wight,

Fig. 8. — Prunus microphylla Hems.: A, Section of staminate flower, *7

showing well-developed stamens and abortive pistil, X3; B, detail Bureau 01 Jrlailt 1I1Q11S-
of calyx from outside, X 3; C, twigs showing leaves and fruit, from £ry from Specimen No.
herbarium specimen, natural size; Z>, fecundated ovary, X 3- _ ,

138851, United States

National Herbarium, collected by Dr. V. Havard, United States Army, in
July, 1883, at Bone Springs near the Chisas Mountains. This locality is

1 “ Amygdalus harvardii W. F. Wight, sp. nov. heaves obovatetooblong-obovateorsometimesfan-shaped
on young growth, 7 to 20 mm. long, 3 to 10 mm. broad, glabrous or sometimes finely pubescent on both

surfaces, usually somewhat pale below and under a lens rather prominently reticulate veined, the margin

conspicuously dentate toward the apex, very rarely toothed below the middle, the teeth usually acute
and apparently glandless. Flowers appearing with the leaves and sessile; calyx slightly pubescent, the
tube about 2.5 mm. long, the lobes scarcely more than 1 mm. long, entire and obtuse; petals not seen. Fruit
sessile, nearly globular, the pubescent exocarp dehiscent along one edge, when dry about 9 mm. long, 7 mm.
broad, and 7.5 mm. thick; stone about 8 mm. long, 6.5 mm, broad, and 7 mm. thick, rounded at the base
and slightly pointed toward the apex, the surface smooth except for indistinct grooves near the ventral
edge.

A shrub with rather rigid branches, stout spinescent branchlets, and light gray bark. The type speci¬
men in the United States National Herbarium was collected in fruit by V. Havard in July, 18S3, in western
Texas, east of the Chisas Mountains, near Bone Springs. It was also collected by C. C. Parry, J. M. Bige¬
low, Charles Wright, and A. Schott on the Mexican Boundary Survey under the direction of Major W. H,
Emery, this specimen being labeled * chiefly in the valley of the Rio Grande, below Donana/ The species
is most closely related to Amygdalus microphylla H. B. & K. of Mexico, but is easily distinguished by its
broader, more obovate leaves as well as by their reticulate venation and eglanflular margins.5' Wight,
W. F. North American species of the genus Amygdalus. Iceland Stanford Jr. Univ., Dudley Memorial
Volume, p. 133, 1913.

The spelling of the specific name harvardii is a typographical error, as the type specimen was collected
by Dr. V. Havard.

Nov. io, 1913

Pubes cent- Fruited Species of Prunus

177

in the southern part of Brewster County, Tex., at about the southern
extremity of the bow of the Big Bend of the Rio Grande.

The description cites also one specimen from the Mexican Boundary
Survey Collections, No. 338. As both specimens show only matured
fruit, it is difficult to place this species with reference to Prunus
minutijbra and P. microphylla , to which it appears to be nearly related,
both in the character of the fruit and in the absence of stomates in the
upper epidermis of the leaves. In its abruptly angled and thorny
branchlets and nearly eglandular leaves (PI. XVI) it would seem to be
most nearly related to P. minutiflora . Whether it will agree with the
above species in the dioecious character of the flowers, small number of
stamens partly placed on the face of the calyx cup and in the finely hairy
inner surface of the cup can only be determined from complete material.
It is provisionally placed in the subgenus Emplectocladus of Prunus.

DESCRIPTION OF PLATES

Plate IX.

XI.

XII.

XIII.

XIV.
XV.

XVI.

Fig. i. — Prunus texana: Better quality of fruit. Natural size.

Fig. 2. — Prunus texana: Fruiting bush, 2 meters in diameter.

Fig. 3. — Prunus texana: Seeds; three scraped clean of pile. Natural size.
Fig. 1. — Prunus texana hybrid, hort. var. Stuart: Fruit and leaves. Nat¬
ural size.

Fig. 2. — Prunus texana hybrid, hort. var. Stuart: Tree in first leaf.

Fig. 3. — Prunus texana hybrid, hort. var. Johnson: Fruiting branch.
Natural size.

Fig. 1. — Prunus andersonii: Plant, showing taproot.

Fig. 2. — Prunus andersonii: Flowering branch. Photographed by Vin¬
cent Fulkerson.

Fig. 3. — Prunus andersonii: Types of seeds. Natural size.

Fig. 1. — Prunus andersonii: Tangled thickets, the more common form.
Fig. 2. — Prunus andersonii : Treelike specimen, 3 meters high.

Fig. 3. — Prunus eriogyna , n. sp.: Erect, large-leaved form of plant.

Fig. 1. — Prunus eriogyna , n. sp.: Common form of plant.

Fig. 2. — Prunus eriogyna , n. sp.: Variable fruits and seeds.

Fig. 3. — Prunus eriogyna , n. sp.: Fruiting branch. Natural size.

Fig. 1. — Prunus eriogyna , n. sp.: Seedlings.

Fig. 2. — Prunus fasciculata: Growth in flood-swept wash.

Prunus minutiflora: Fruiting branch. Natural size. Photographed by
S. H. Hastings.

Prunus havardii: Fruiting branch of the type specimen.

(r7®)

ADDITIONAL COPIES of this publication
■A may be procured from the Superintend¬
ent op Documents, Government Printing
Office, Washington, D. C., at 25 cents per copy

Subscription, per year, 12 numbers - - $2.50

Plate IX

Pubescent-

Plate XI

„ i

Pubescent-Fruited Spec

Plate XII

Pubescent- Fi

Plate

Agricultural

JOURNAL OF AGRHET1AL RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. I Washington, D. C., December io, 1913 No. 3

SELECTIVE ADSORPTION BY SOILS

By E. G. Parker,

Scientist, Soil Laboratory Investigations, Bureau of Soils

From the standpoint of soil chemistry the absorption of material from
the air and the soil solution by the soil is of first importance. The ab¬
sorptive power of a soil enables it to retain the soluble salts necessary
to plant life in spite of the leaching effect of rains and the movement of
the soil solution toward the surface of the soil in dry weather, and thus
to store up soluble material, either natural or applied in the form of a
so-called fertilizer, for the future needs of crops.

The absorptive properties of soils have been under investigation in the
Soil Laboratory for several years under the direction of Dr. Frank K.
Cameron, and several publications 1 describing this work have appeared
from time to time. The object of the work described in this paper was
to obtain clearer insight into the mechanism of adsorption phenomena,
particularly selective adsorption, and the characteristic effects of one
solute upon the adsorption of another.

It is a well-known fact that either by leaching or by shaking a soil with
a solution of potassium chlorid (or some neutral salt) the amount of
potassium present will be diminished, and a certain amount of the bases
of the soil (Ca, Mg, etc.) will be found in the resulting solution, while the
amount of the chlorin will remain practically unchanged. Also, the re¬
sulting solution is slightly but distinctly acid to our common indicators.

On treating kaolin with solutions of magnesium and sodium chlorids
Kohler 3 found the resulting solutions to be slightly but distinctly acid

1 Cameron, F. K., and Bell, J. M. The mineral constituents of the soil solution. U. S. Dept. Agr., Bur.
Soils, Bui. 30, 1905.

Cameron, F. K., and Patten, H. E. The distribution of solute between water and soil. Jour, of Phys.
Chem., v. 11, p. 581-593, 1907.

Patten, H. E. Some surface factors affecting distribution. Trans. Amer. Electrochem. Soc., v. 10,
p. 67-74. 1906.

Patten, H. E., and Gallagher, F. E. Absorption of vapors and gases by soils. U. S. Dept. Agr., Bur.
Soils, Bui. 51, 1908.

Patten, H. E., and Waggaman, W. H. Absorption by soils. U. S. Dept. Agr., Bur. Soils, Bui. 52, 1908.

Schreiner, Oswald, and Failyer, G. H. The absorption of phosphates and potassium by soils. U. S.
Dept. Agr., Bur. Soils, Bui. 32, 1906.

2 Kohler, Ernst. Adsorptionsprozesse als Faktoren der Lagerstattenbildung und Lithogenesis. Ztschr.
Prakt. Geol., Jahrg. 11, p. 49-59. 1903-

Journal of Agricultural Research, Vol. I, No. 3

Dept, of Agriculture, Washington, D. C. Dec. 10, 1913

H— 1

17072“— 13 - 1

(179)

i8o

Journal of Agricultural Research

Vol. I, No. 3

to litmus and attributed this to the fact that a selective concentration
of the dissolved substance — an adsorption of the base — had taken place.

E. C. Sullivan 1 2 repeated these experiments and obtained the same
result, accounting for it by an exchange of the sodium and magnesium
of these salts in part for the iron and aluminium of the kaolin, the salts
of the latter undergoing extensive hydrolysis in dilute solution.

Similarly, the acidity of a salt solution after treating a soil with it is
explained by some as a hydrolysis of aluminium and iron salts after the
replacement by the base of the salts and by others as a selective ad¬
sorption of the base of the salt.

It has been found by many experimenters that on quantitatively de¬
termining the replaced bases present in a salt solution after treating a
soil, kaolin, various silicates, etc., with the solution the replaced bases are
equivalent or very nearly equivalent, within the limits of experimental
error, to the loss of the base of salt.

Van Bemmelen 3 treated ioo grams of soil with 200 c. c. portions of
solutions containing 8 and 40 mg. equivalents of potassium chlorid.
After filtration the solutions were analyzed, and it was found that an
almost complete exchange of potassium for sodium, calcium, and mag¬
nesium had taken place. Chlorin was determined in one experiment and
had not changed.

Sullivan 3 found that by treating kaolin and various other silicates
with salt solutions a quantity of bases almost equivalent to the loss of
the base from the salt was dissolved in each case.

Wiegner4 found that on treating an artificial amorphous water-
containing (hydrated) so-called double silicate with a neutral salt solu¬
tion the cation of the neutral salt was taken in part from the solution,
and in its place the cations of the silicate-gel in nearly equivalent amounts
entered the solution. The anion of the neutral salt remained unchanged,
provided secondary reaction did not take place.

From many similar investigations with the same general result —
namely, that the bases dissolved are very nearly equivalent to the loss
of the base of the salt in solution — it would seem and is concluded by
many experimenters that an exchange of bases takes place in the soil
according to the following reaction :

KC1 (for example) + Xn silicate,^ XnCl + K silicatem

From the -standpoint of fertilizer practice, however, on applying
potassium chlorid to the soil it is very unlikely that the above reaction
takes place and that the potassium is held in the soil as a relatively
insoluble silicate and in a form highly unavailable for plants.

1 Sullivan, E. C. The interaction between minerals and water solutions. U. S. Geol. Survey, Bui. 312,
1907.

2 Bemmelen, J. M. van. Das Absorptionsvermogen der Ackererde. Landw. Vers. Stat., Bd. 21, p.
I35-I9I* 1877*

8 Sullivan, E. C. Op. cit.

4 Wiegner, Georg. Zum Basenaustausch in der Ackererde. Jour. Eandw., Bd. 60, p. 111-1:50, 197-222,
1912.

Dec. 10, 1913

Selective Adsorption by Soils

181

Certain inactive solid substances presenting large surfaces have the
power of taking salts from solution — that is, what is known as absorbing
or adsorbing them, a phenomenon most logically explained at present as
a concentrating of the solute at the surface of the adsorbing material.
Qualitatively, it is known that certain of these inactive solid substances
not only have the power of adsorbing a neutral salt from its solution as
a whole, but may adsorb one ion more than the other, or selectively
adsorb. In so doing, a partial hydrolysis of otherwise practically unhy¬
drolyzed salts is brought about, since the removal of one ion of the
salt more or at a greater rate than the other takes an equivalent number
of ions of opposite charge from the water and thus leaves an excess of
either hydrogen or hydroxyl ions in the solution. That such is the case
can be shown by the use of common indicators, after shaking solutions
of neutral salts with or percolating them through certain of these inac¬
tive solid substances.

These cases are so numerous that only a few of the best known and
more convincing ones will be here recalled.

A silver-nitrate solution shaken with animal charcoal and the super¬
natant liquid filtered and tested with methyl orange or litmus gives a
distinct color of acid reaction.

A potassium chlorid or nitrate solution shaken with cane-sugar char¬
coal and the supernatant liquid filtered and tested with phenolphthalein
gives a strong red color of alkaline reaction.

An interesting case of selective adsorption is to be found in our com¬
mon indicator, Congo red, and absorbent cotton. If the base of a column
of absorbent cotton is immersed in a solution of Congo red made very
slightly acid, in a very few minutes the cotton immediately above the
solution is colored blue (acid reaction), while above the blue color for
about an inch in height is seen the red color of neutral or alkaline reac¬
tion; above the red the cotton is wet with water.

The soil possesses all the essential properties of these adsorbing mate¬
rials; but that it has the power of selectively adsorbing to any appre¬
ciable extent has for a long time been a question of dispute. The fact
that a solution of a neutral salt after contact with a soil is as a rule dis¬
tinctly acid to indicators supports this hypothesis.

If a soil in contact with a solution of potassium chlorid adsorbs potas¬
sium ions at a much greater rate or in greater proportion than chlorin ions,
thereby (since an equivalent number of hydroxyl ions are also removed
with the potassium ions) causing a partial hydrolysis of the solution
(KC1 + H0H=(K0H) adsorbed + HC1), then free hydrochloric acid will
be left in the solution.

It is not unreasonable to assume that the uncombined acid might dis¬
solve an almost equivalent amount of bases from the soil particles. On this
assumption, by using a solution of a salt of potassium with a weaker acid
than hydrochloric, there should be a greater adsorption of potassium ions,

i82

Journal of Agricultural Research

Vol. I, No. 3

since the salt is more easily hydrolyzed than potassium chlorid, less sur¬
face energy being required to obtain potassium ions from solution, while
the quantity of anions adsorbed will depend upon the specific properties
of the anion employed. Also, if the anion of the salt is that of a weaker
add than hydrochloric and is not adsorbed to a much greater extent than
chlorin ions, a smaller amount of bases should be dissolved from the soil
and a correspondingly greater acidity of the solution should result.
Again, if a reaction is interposed so that the free acid will be used up
before it has a chance to react with the soil particles — i. e., by adding a
small amount of sodium hydroxid, yet enough to neutralize the acid
theoretically set free — little or no dissolved bases of the soil should be
found in the resulting solution.

On the assumption that certain ingredients of the soil adsorb in part
the base of a neutral salt in solution and that the free acid resulting from
the hydrolysis caused by this adsorption reacts with certain of the soil
particles and dissolves an almost equivalent amount of bases of the soil,
the following experimental work is based.

SERIES No. i

In series No. i , 500-gram portions of a Durham sandy loam were intro¬
duced into a number of bottles of 2 -liter capacity. To the first was
added 2,000 c. c. of a solution containing 7.65 grams of potassium chlorid
per liter; to the second 2,000 c. c. of a solution containing potassium
acetate equivalent to 7.47 grams of potassium chlorid per liter; to the
third 2,000 c. c. of water. The bottles were shaken frequently at room
temperature for two days. The soil was allowed to settle until the
supernatant liquid was apparently clear. Portions of the supernatant
liquid were then pipetted off, filtered, and analyzed.

The supernatant liquid from soil shaken with pure distilled water
showed no appreciable presence of material dissolved from the soil, while
the analyses of the supernatant liquids from soil shaken with the above
solutions showed soil material present. The potassium-chlorid equiva¬
lents of the various constituents determined by these analyses are given
in Table I.

Tabi^k I. — - Adsorption by Durham sandy loam of potassium from solutions of potassium

salts.

[Results stated in grams of potassium chlorid per ioo c. c. equivalent to constituents determined

by analyses.]

Constituents by analysis.

From

KC1

solution.

From [
CHaCOOK!
solution, j

Constituents by analysis. |

From

KC1

solution.

From

CHaCOOK

solution.

K before contact .

Grams.

0. 7650
.6950
. 0107

Grams . j
O. 7470
. 6560
■ 0015

Mg after contact .

Grams.

O. 0157

Grams.

O. 0167

K after contact .

Na after contact .

A1 after contact .

Free acid after contact.

. 0112

. 0402

Ca after contact .

•0353

• °3*4

Anions after contact. . .

• 7647

•7450

Dec. io, 1913

Selective Adsorption by Soils

183

In the foregoing experiments the determination of the free acid is
unreliable, considering the fact that no indicator could be used for
titrating which was sensitive enough and at the same time unaffected by
carbon dioxid. The results can be considered only as approximations.
Boiling to remove the carbon dioxid is impossible when potassium ace¬
tate is used, since it hydrolyzes on boiling, giving an alkaline reaction
to indicators. Iron and titanium were determined in several cases and
found to be present in negligible amounts in the precipitated alumina.
The amount of chlorin present in the solution was found to be practi¬
cally unchanged.

From the data obtained when the potassium chlorid is used, the amount
of potassium chlorid equivalent to loss of potassium (0.7650—0.6950=
0.0700 grams per 100 c. c.) during contact is greater than the amount of
potassium chlorid equivalent to the bases dissolved from the soil (0.0107 +
°.0353 + o.oi57 = o.°6i7 grams per 100 c. c.) by an amount (0.0700—
0.0617 = 0.0083 grams per 100 c. c.) about equal to the amount of potas¬
sium chlorid equivalent to the estimated free acid (0.0112 grams per 100
c. c.). When potassium acetate is used, the amount of potassium
chlorid equivalent to the loss of potassium (0.7470— 0.6560=0.0910
grams per 100 c. c.) during contact is again greater than the amount of
potassium chlorid equivalent to the bases dissolved from the soil (0.0015 +
0.0314+0.0167 = 0.0496 grams per 100 c. c.) by an amount (0.0910—
0.0496 = 0.0414 grams per 100 c. c.) about equal to the amount of potas¬
sium chlorid equivalent to the estimated free acid (0.0402 grams per 100
c. c.).

When potassium acetate is used, the bases dissolved from the soil are
54.5 per cent^P 0910 X IO°^ w^at they would be if a complete exchange

of bases had taken place, while, when potassium chlorid is used, this per-

(0.0617 \

o 0700 X 100 J'

SERIES No. 2

In series No. 2, 250 grams of a Norfolk sandy loam were placed in a
2-liter bottle. To this was added 1,000 c. c. of a solution containing
18.38 grams of potassium chlorid and about 1 gram of sodium hydroxid
per liter. The bottle was shaken frequently at room temperature for
two days. The soil was allowed to settle until the supernatant liquid
was apparently clear. Portions of the supernatant liquid were then
pipetted off, filtered, and analyzed.

Soil shaken with pure water showed no appreciable presence of material
dissolved from the soil in the supernatant liquid.

The above potassium-chlorid solution when shaken with soil showed
a quantity of potassium chlorid equivalent to the loss of potassium of
0.1520 grams per 100 c. c. and no appreciable loss of chlorin. The

184

Journal of Agricultural Research

Vol. I, No. 3

amount of bases of the soil (Ca, Mg, etc.) present in the resulting solution
was found to be negligible. If, however, too great an excess of sodium
hydroxid is present, the resulting solution is discolored, and iron in
appreciable amounts is found in the solution.

It was found that the addition of a small amount of sodium hydroxid
to a solution of potassium chlorid prevented the presence of dissolved
bases when the solution is shaken up in contact with a soil, and yet a loss
of potassium occurred of the same magnitude as when bases were found
in the resulting solution, the amount of chlorin remaining practically
unchanged.

Believing the assumption previously made to have been entirely justi¬
fied by the foregoing experimental work, the hope of finding the effect of
concentration, size of soil particles, and presence of other substances,
with special regard to substances commonly used in fertilizer practice,
on the selective adsorption by soils led to the following experimental
work:

SERIES No. 3

In series No. 3, 35-gram portions of a Norfolk sandy loam collected
near Laurinburg, N. C., and a Marshall silt loam collected near Edgerton,
Mo., were placed in 200 c. c. bottles with solutions of potassium chlorid
containing varying quantities of potassium chlorid and a small amount
of sodium hydroxid per liter. The bottles were then rotated in a ther¬
mostat at room temperature for two days. The soil was allowed to
settle until the supernatant liquid was apparently clear. Portions of
the supernatant liquid were then pipetted off, filtered, and analyzed,
the results of the analyses being given in Table II.

Table II. — Effect of concentration on adsorption of potassium from solutions of potas¬
sium chlorid by Norfolk sandy loam and by Marshall silt loam.

Norfolk sandy loam.

Marshall silt loam.

Quantity of KC1 equivalent
to the quantity of K per
100 c. c. of solution.

Loss.

Quantity of KCl equivalent
to the quantity of K per
100 c. c. of solution.

Loss.

Before contact.

After contact.

Per 100
c. c. of
solution.

Percent¬

age.

Before contact.

After contact.

Per 100
c. c. of
solution.

Percent¬

age.

Grams.

25-855°

14. 7700

9. 1250

6. 2580

4. 7400

3. 1120

1. 8380
. 6406
.3064
. 1283

Grams.

25. 6750
14. 6500

8. 9650

6. 1100

4- 5950

2. 9600

1. 7010
. 5640
.2650
. 0960

Grams.

O. 1800
. 1200
. 1600
. 1480
• 1450
. 152O
. I370
. 0766

.0414

*0323

O. 70
.81

Im 75
2.36

3. 06
4.89

7- 45
11. 96
I3* 5i
25. 18 j

Grams.

II. 8400
IO. 0450

6. 6950

4. 4860

2. 6700

I. 1640

Grams.

II.3500

9. 5700

6, 2450

4. 0420

2. 24OO
. 7700

Grams.

O. 4900
•4750
.4500
.4440

.4300

.3940

4. 14

4- 73

6. 72

9. 90
16. 11
33- Si

Bee. io, 1913

Selective Adsorption by Soils

185

From the data obtained in this experiment (see fig. 1) we find that
from the zero concentration of potassium chlorid, where necessarily the
adsorption of potassium is zero, the loss of potassium during contact
increases regularly with the concentration to a certain point and then
remains practically constant, the surface of the soil particles having
apparently taken up the greater part of the potassium possible at this
point. The point at which the adsorption of potassium becomes prac¬
tically constant is much lower in the case where a sandy loam is used
than when a silt loam is used. The percentage of potassium adsorbed
increases asymptotically as the concentration of potassium chlorid

Pig, i. Curves showing the effect of concentration on the selective adsorption of potassium from solutions
of potassium by Norfolk sandy loam and by Marshall silt loam.

decreases, and it may be concluded that the adsorption of potassium
becomes practically complete at very low concentrations of potassium
chlorid. Chlorin was determined in several cases and was found to have
remained unchanged.

SERIES No. 4

In series No. 4, 35-gram portions of a subsoil of Cecil clay, a subsoil of
Marshall silt loam, a subsoil of Norfolk sandy loam, a subsoil of Decatur
clay loam, and a subsoil of Carrington loam were placed in 200 c. c.
bottles with solutions of potassium chlorid of about the same concen¬
tration and treated as in series No. 3. The results are given in Table III.

Table III .—Effect of amount of surface exposed on adsorption.

Type of soil.1

Cecil clay .

Decatur clay loam .
Marshall silt loam. .
Carrington loam . . .
Norfolk sandy loam,

Quantity of KC1 equivalent to
the quantity of K per 100 c. c.
of solution.

Before contact.

After contact.

Grams.

6- 7350

6 ■ 5550

6. 6950

6. 4300

6. 2580

Grams.

6. 4100

6. 3I50

6. 2450

6. 2050

6. 1100

Difference.

Grams.

o. 3250
. 2400
.4500
.2250
. 1480

1 The soils in this table are arranged in order of the relative amount of surface exposed.

Journal of Agricultural Research

Vol. I, No. 3

1 86

As was expected, since the removal or adsorption of potassium from
a potassium-chlorid solution is undoubtedly a surface phenomenon, in
general the smaller the soil particles the greater was the adsorption of
potassium. Clay, however, in spite of the fact that the particles are
smaller than those of the other types of soil, does not show a corre¬
spondingly greater adsorptive power, the surface of the clay particles
being probably of a different nature. The classification of the different
types of soil is based entirely on their mechanical analysis.1

SERIES No. s

In series No. 5, 35-gram portions of Marshall silt loam (the same as
that used in experiment III) were placed in 200 c. c. bottles with solutions
containing varying amounts of potassium chlorid per liter. To some of the
portions 10 grams of sodium nitrate were added, while to others 10 grams of
monobasic calcium phosphate were added. These were treated as in
experiment III. A solution containing 58.25 grams of potassium chlorid
per liter in contact with calcium phosphate alone lost an amount of
potassium during contact equivalent to 0.0500 gram of potassium
chlorid per 100 c. c. The results of the analyses of the supernatant
liquids are given in Table IV.

Table IV. — Effect of the presence of other substances on adsorption.

Experiment No.

Quantity of KC1 equiva<-
lent to the quantity of K
per 100 c. c. of solution.

Loss.

Before

contact.

After

contact.

Per 100 c. c.
of solution.

Percent¬

age.

A. — With 10 grams of NaNOa present:

Grams.

I .

11. 1850

*0- 3750

0. 8100

7- 25

II .

8. 9950

8. 2650

. 7300

8. 12

Ill .

6. 2400

5. 6600

. 5800

9- 30

IV .

4. 4270

3-9470

. 4800

10. 83

V.... . . . . .

2.0450

1. 7140

•33OS

16. 15

VI .

. 8270

•5950

. 2320

28. 05

B. — With 10 grams of CaH4(P04)2 present:

I .

11. 1100

10. 5700

. 5400

4. 86

II .

9. 1300

8. 6200

* 5IO°

5- 59

Ill .

6. 3400

5.8500

. 4900

7- 73

IV .

4- 5830

4. 1200

.4630

10. 10

V .

L 993O

1. 5480

•4405

22. 10

VI .

.QIQO

. 5500

.3690

40. 15

C. — With 5 grams of NaN03 present:

I .

6- 3950

5. 8200

* 5750

9. 00

D. — With 5 grams of CaH4 (P04)2 present:

I .

6. 3850

5. 9000

. 4850

7. 60

Table IV and figure 2 show that the presence of sodium nitrate at con¬
centrations of potassium chlorid below about 37.5 grams per liter

1 Fletcher, C. C , and Bryan, H. Modification of the method of mechanical soil analysis. U. S. Dept.
Agr., Bur. Soils, Bui. 84, 1912.

Dec. io, 1913

Selective Adsorption by Soils

187

decreases the adsorption of potassium from a potassium-chlorid solution
by a soil and increases it above this concentration. They also show
that the presence of monobasic calcium phosphate does not alter the
adsorption of potassium from a potassium-chlorid solution appreciably,

Fig. 2. — Curves showing the effect of the presence of sodium nitrate and calcium phosphate on the selective
adsorption of potassium from solutions of potassium chlorid.

what change there is in the form of the curve being undoubtedly due
to the removal of potassium by the calcium phosphate not in solution,
either by a physical (adsorption) or a chemical reaction.

SUMMARY

Soils not only have the power of adsorbing dissolved salts from solu¬
tions but also of adsorbing one ion at a greater rate than the other, or
selectively adsorbing, to a marked extent.

The presence of bases of the soil (Ca, Mg, etc.) in solution after shak¬
ing certain salt solutions with or percolating through a soil is probably
not due to a direct chemical reaction of the salt in solution with the
silicates of the soil, but to a reaction of free acid, resulting from a
selective adsorption of the cation, with the mineral components of
the soil.

The rate of adsorption of chlorin ions from solution by soils is much
less than of potassium ions.

The selective adsorption of potassium from a potassium-chlorid
solution by a soil increases in amount with the concentration up to a
certain point and then remains practically constant.

Journal of Agricultural Research

Vol. I, No. 3

1 88

The percentage of potassium adsorbed from a potassium-chlorid
solution increases asymptotically as the concentration of potassium
chlorid decreases and at very low concentrations adsorption is practically
complete.

In general, the smaller the soil particles the greater the selective
adsorption of potassium from a potassium-chlorid solution by the soil.

The presence of sodium nitrate decreases the adsorption of potassium
from a solution of potassium chlorid by a soil up to a concentration of
about 37.5 grams of potassium chlorid per liter and then increases it.

The presence of monobasic calcium phosphate does not change ap¬
preciably the adsorption of potassium from a potassium-chlorid solution
by a soil.

Finally, if a mineral fertilizer be applied to a soil and exposed to the
rain and thus dissolved and carried through the soil in solution, these
substances will be adsorbed (an entirely physical phenomenon) either as
a whole or selectively from the solution by the vast surface of the soil
particles and will be held there by this same physical force until the
plant or subsequent leaching removes it.

The presence of other mineral substances added to the soil may or may
not increase or decrease the rate at which this adsorptive phenomenon
takes place.

A BACTERIUM CAUSING A DISEASE OF SUGAR-BEET
AND NASTURTIUM LEAVES

By NELLIE A. Brown, Assistant Pathologist , Laboratory of Plant Pathology, and

Clara O. Jamieson, Scientific Assistant, Office of Cotton and Truck Disease and

Sugar-Plant Investigations , Bureau of Plant Industry

INTRODUCTION

The bacterial disease described in this paper was first observed in the
spring and summer of 1908 on nasturtium leaves growing near Richmond,
Va., and on sugar-beet leaves collected from the Government plat at
Garland, Utah. The disease on both hosts was of the leaf-spot type, but
since the general appearance was not at all similar there was no thought
at the time of a possible relationship between the causal organisms.1
Investigations of the disease as it occurred on each host were at once
begun, but not until the studies had progressed for nearly two years did
it become evident that there was a striking similarity in regard to both
cultural and morphological characteristics of the bacteria isolated from
the two kinds of diseased leaves.

A comparative study of the bacteria followed, care being taken to use
the same media placed under similar conditions. As a result of studies
extending over four years, it has been found that in essential character¬
istics the bacterial organisms are so nearly identical that in the opinion
of the writers the causal organism is one and the same bacterium. Any
minor differences which occur may be attributed to individual adapta¬
tion due to host influence.

OCCURRENCE AND GENERAL APPEARANCE OF THE DISEASE ON THE

TWO HOSTS

The material furnishing the basis of this study was received during
the spring and summer of 1908. The diseased nasturtium leaves were
sent in from Richmond, Va., to Dr. C. O. Townsend, then Pathologist in
Charge of Sugar-Beet Investigations in the Bureau of Plant Industry.
The diseased leaves had been gathered from young nasturtium plants
growing in an open garden bed and when received were somewhat wilted
and discolored, showing water-soaked and brownish-colored spots from
2 to 5 mm. in diameter. Upon microscopic examination the tissue
within and surrounding these diseased spots was seen to be filled with
great numbers of active bacteria.3

1 Brown, Nellie A. A new bacterial disease of the sugar-beet leaf. Science, n. s., v. 29, no. 753, p. 915,
1909. Jamieson, Clara O. A new bacterial disease of nasturtium. Science, n. s., v. 29, no. 753, pp. 915-
916, 1909.

* [Halsted, B. D.] Nasturtium blight. New Jersey Agr. Expt. Sta., 17th Ann. Rpt., [1895] 1896, p. 410,
fig. 56, 1897.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Vol. I, No. 3
Dec. ro, 1913
G-5

190

Journal of Agricultural Research

Vol. I, No. 3

The diseased sugar-beet leaves were collected by Dr. Townsend in
Utah and California on inspection trips to the sugar-beet sections of the
West and were sent to the laboratory in Washington for examination.
Leaves similarly diseased were also received from Oregon during the
summer of 1909, but, so far as known to the writers, the trouble has not
been noticed up to the present time in any other beet-growing State.

The first leaves came from Utah and had dark-brown, often black,
irregular spots and streaks from 3 mm. to 1.5 cm. in diameter. They
occurred on the petiole, midrib, and larger veins. Occasionally the
discoloration extended along the veins for some distance, and the tissue
on either side was brown and dry; sometimes there were corklike pro¬
tuberances at the central point of the spots. In badly diseased petioles
the tissue had softened as though affected with a soft rot, but when only
a few spots occurred there was no indication of softness.

Unlike the spot diseases due to Cercospora and Phyllosticta, this
spotting did not spread through an entire beet field, but was generally
limited to small areas.

The tissue embracing the dark spots was examined with the micro¬
scope as soon as the material was received and was found to be filled
with very active bacteria; no fungous hyphae were seen. Some of the
leaves were placed in a moist chamber and carefully watched for several
days, but there was no fungous mycelium in or around the spots.

ISOLATION OF THE ORGANISM FROM THE TWO HOSTS

The method of isolating the bacterial organism from the diseased
sugar-beet and nasturtium leaves was by means of poured agar plates.
Spots from the soundest leaves were used, the tissue being immersed in
mercuric chlorid (1:1,000), washed in sterile water, and mashed in
bouillon. The plate colonies were up in 24 hours. They were round,
thin, smooth, glistening, whitish in reflected light, bluish in trans¬
mitted light, and 1 to 5 mm. in diameter. In three days the agar in the
immediate neighborhood of the colonies had changed to a yellowish-
green color. No other colonies appeared on the plates.

With young subcultures from these plate colonies needle-prick inocu¬
lations were made into sugar-beet and nasturtium plants, in order to
prove that the right organism had been isolated in either case. The
inoculations with the separate organisms from the two hosts are as fol¬
lows :

INOCULATIONS WITH ORGANISM ISOLATED FROM SUGAR-BEET LEAF

Inoculations with the organism isolated from sugar-beet leaves into
healthy sugar-beet leaves of plants growing in the greenhouse proved
that the right organism had been isolated, for in three days there were
black spots at all points of inoculation. The checks were free from

Dec. iof 1913

Disease of Sugar-Beet and Nasturtium Leaves

191

spots. Some of the inoculated leaves were taken to the laboratory, the
black spots examined, and numerous bacteria found swarming in the
cells. From these spots, produced by the first inoculations, the organ¬
ism was reisolated in pure culture, and sugar-beet leaves in the green¬
house were inoculated repeatedly, the dark spotting and streaking of the
leaves occurring in every case. Altogether, more than 100 sugar-beet
leaves were inoculated. Although the infection took readily at the
inoculated places, the disease was not observed to occur on any uninocu¬
lated beet plants except in two instances, when several beets of a neigh¬
boring row became affected. No slugs or worms were on the leaves, but
thrips were abundant, and there were also a number of grasshoppers
which had escaped capture; so possibly the infection was carried by one
of these insects.

When the petioles, midrib, and large veins were inoculated by means
of needle pricks, the infection took very rapidly, and the discoloration
often ran along the course of the veins and veinlets. When the leaf
blades were inoculated at the ends of tiny veins, there was only a dark¬
ened ring around the punctures. The infection took most rapidly on
the petiole. (PI. XVII, fig. 1.) In three days after needle- prick inocula¬
tions in young growing leaves the tissue was depressed, darkened, and often
ruptured for a distance of 5 mm. around the puncture. Young beet
leaves with blades about 8 cm. in length very readily succumbed to
needle-prick inoculation in the blade as well as in the petiole and mid¬
rib. When material from a young culture less than 2 days old was
inoculated into rapidly growing leaves, the spotting began to show in
24 hours. Old tissues were also found susceptible to the disease, but
the infection did not take so rapidly. The sugar-beet root also was inocu¬
lated and the disease was found to take hold there slightly. (PI. XVII,
fig. 2.) There was no soft-rot condition, but cavities occurred in the
roots where the inoculation pricks were made. These cavities pene¬
trated into the interior of the beet and reached a depth of 2 cm. within
two weeks after inoculation. Occasionally a cork-like condition of a
dark color followed along the immediate line of the needle prick and no
cavities were present. The discoloration, however, was not nearly so
dark as in the leaf, nor was there as much tendency to spread as in the
leaf.

So far as the writers know, this organism has not been found in the
field attacking the beet root, and as none of the field beets with affected
leaves had any root trouble, it is thought that the disease in the field
is confined strictly to the leaf.

Spraying the organism on the leaves of beets did not produce the
disease. Precautions were taken to prevent the bacteria from drying
before they had time to get into the leaves. An infection cage was
placed over beets growing in the open ground in the greenhouse, the

192

Journal of Agricultural Research

Vol. I, No. 3

plants were watered well, and the leaves were sprayed with sterile water
and left under the cage overnight, so that the stomata would open. The
following day the growth from two-day-old agar cultures was shaken up
well in sterile water and sprayed on the upper and lower surfaces of the
leaves. The plants were watched carefully for two weeks, but no trace
of the disease was ever seen. The experiment was repeated some months
later with the same result.

Some cultures were sent to Garland, Utah, and Mr. H. B. Shaw, who
had charge of the experiment station there during the season of 1909,
inoculated the leaves of sugar beets growing in the open field. There,
as well as in the greenhouse, the plants became infected very readily.
Mr. Shaw sent some of the leaves to the sugar-plant laboratory at Wash¬
ington. Upon examination swarms of bacteria were found in the black¬
ened areas. Mr. Shaw also took portions of the diseased leaves, including
the spots, and inoculated other leaves with them. Fifty per cent of the
leaves treated in this way became spotted.

The most striking feature of this affection as it occurs in the green¬
house from inoculations is the black color of the spots and streaks, for
they stand out prominently against the green of the leaves. These
leaves never become soft, but bend over at the badly sunken spots, lose
their turgidity, and finally die from drying out. If the petiole is inocu¬
lated, it frequently happens that the leaf blade will drop at a sharp angle
from the infected area in less than two weeks.

INOCULATIONS WITH ORGANISM ISOLATED FROM NASTURTIUM LEAF

Inoculations with the organism isolated from nasturtium leaves were
made into leaves of some rather old nasturtium plants growing in pots
in the greenhouse. After several days small, watery-looking areas
became visible, and the tissue within these areas became discolored and
shriveled, resembling in all particulars the original spots from which the
organism was obtained. A microscopic examination of the tissue within
the diseased areas thus produced showed the cells to be filled with many
active bacteria. Check plants having leaf surfaces pricked with a sterilized
needle presented no indication of diseased spots. From the observation
of inoculated plants it was noticed that the general appearance of the
leaf spot changed considerably during the different stages of its devel¬
opment. Leaves of a healthy young nasturtium plant showed the
effects of needle-prick inoculations within 48 hours, the tissue at first
becoming slightly darker in the infected areas and presenting a water-
soaked appearance. These spots gradually increased in size, becoming
4 to 6 mm. in diameter, while the tissue within became dry and brown¬
ish in color and often brittle enough to crack (PI. XVIII). A dropping
out of this diseased tissue frequently followed, and finally the whole
leaf turned yellow and fell from the stem.

Dec. 10, 1913

Disease of Sugar-Beet and Nasturtium Leaves

193

REISOLATION FROM INOCULATED TISSUE

Out of a small piece of tissue cut from one of the spots produced by
inoculation a bacterial organism was isolated by means of agar plates,
and by careful comparison with previous cultures was found to be
similar in all respects to the organism obtained from the original diseased
leaves. As soon as suitable cultures of this reisolated organism could
be grown, inoculations were made into healthy young plants, and again
the characteristic brown and shriveled spots were produced, with an
abundance of active bacteria in the tissue. By these and other similar
experiments it is proved beyond a doubt that the nasturtium leaf spot
is caused by a bacterial organism. The manner in which the bacteria
gain entrance to the tissue of the host has not been fully demonstrated,
but from observations made during the investigation it seems probable
that insect injuries, as well as mechanical wounds, open the way for the
entering of the parasites.

CROSS-INOCULATIONS BETWEEN HOSTS

After proving that the right organism had been isolated from either
host, inoculations into leaves of other plants were made, with the result
that the sugar-beet organism proved very infectious to nasturtium, and
likewise the nasturtium organism proved infectious to the sugar beet.
But as the two investigators were working independently, each with
one organism, this interesting fact had no particular significance at the
time. Nasturtium leaves inoculated with the sugar-beet organism
became spotted and watery-looking for some distance beyond the in¬
oculation pricks, appearing in all respects similar to spots produced by
inoculations with the nasturtium organism. Later, the watery-looking
areas turned from a yellow to a brown color, and still later these tissues
dried up and fell out (PI. XIX, fig. 2). Some leaves drooped and died.
The check leaves showed no discoloration ; nor did any part of the tissue
fall out, as in the inoculated leaves.

Three years afterward the same strain of the organism was inoculated
into young nasturtium leaves at the same season of the year and under
practically the same conditions as before, but there was a slight in¬
fection only, though young sugar-beet leaves inoculated with the same
culture were badly infected.

Although inoculations with the nasturtium organism into sugar-beet
leaves produced the disease, this strain of the organism was not so in¬
fectious as the sugar-beet strain on nasturtium. This difference in the
behavior of the organisms in cross-inoculation was considered to be one
of host influence.

194

Journal of Agricultural Research

Vol. I, No. 3

OTHER PLANTS INOCULATED WITH THE ORGANISM FROM BOTH HOSTS

That this bacterial spot is not confined to sugar-beet and nasturtium
leaves has been shown by a number of inoculations performed upon
other plants growing in the greenhouse. Both strains of the organism
were used. Diseased spots were produced with the bacteria upon leaves
of pepper, lettuce, eggplant, and upon the leaves and pods of the bean
plant. Inoculatidn experiments were also tried on potato, clover, %nd
daisy plants, but without any definite infection, although there was
slight discoloration on potato leaves.

The stems and leaves of the young pepper plants were readily in¬
fected through needle-prick inoculations. The spots were black, and the
stems seemed more susceptible than the leaves.

Lettuce leaves growing in the greenhouse blackened readily after in¬
oculation. One plant out of seven was entirely destroyed by the in¬
fection. One month later, when the temperature of the greenhouse was
not so even throughout the day and night and the plants of the same
lot had stopped growing rapidly and become toughened, the organism
failed to produce infection.

The leaves of eggplant were inoculated, and brown spotting resulted at
the punctured places; later, these areas dropped out of the leaves.

Of these various hosts the bean proved especially susceptible to the
organism, inoculations taking effect almost as readily as upon the nas¬
turtium and sugar-beet leaves. Bean plants inoculated with a young
agar culture of both strains of the organism showed the characteristic
brown spots on the leaves within three to five days. Ten days after in¬
oculation some of the diseased leaves (PI. XIX, fig. i) were examined, and
active bacteria were found in the cells. Three weeks after inoculation
the bean leaves shriveled and died. Later, inoculations which were made
upon the young pods of bean plants produced conspicuous, somewhat
sunken, brownish spots in the tissue. (PI. XIX, fig. 3.)

At the same time that the inoculation experiments were being carried
on, cultural and morphological studies were made with both strains in the
laboratory. From time to time notes and various tests were compared,
and, as a result, the identity of the two strains was established. Such
being the case, only one description will hereafter be given for the two
strains, except where marked differences occur.

DESCRIPTION OF THE ORGANISM
MORPHOLOGICAL CHARACTERS

Vegetative Cells. — The organism is a medium-sized schizomycete of
varying length when grown in different media. It is a short rod with
rounded ends, occurring singly or in pairs (fig. 1); occasionally it pccurs
in long chains of two to many elements and again in long unsegmented

Dec. io, 1913

Disease of Sugar-Beet and Nasturtium Leaves

195

(

0 0

Fig. x, — Bacterium a ptatumirom. a 2-day
beef-bouillon culture stained with
carbol fuchsin.

filaments (fig. 2, a and 6). In stained tissue of the hosts the average
measurement of a single rod is 1.2 by o ,6ft. The organism grown in a
3-day-old beef bouillon culture and stained
in carbol fuchsin has an average size of 2.1
by 0.7/*. When stained with Toeffler's
flagella stain, the average is 3.2 by 1.3/t.

Process of Ceu, Division. — Cell divi¬
sion takes place in the bacterium by
simple, transverse fission. In order to
study the process of fission, agar hang¬
ing blocks containing the organism
were made in the following manner:

Thin beef-agar plates were poured and transfers from a bacterial
culture streaked across the surface of the hardened agar. Agar blocks

a few millimeters square were
then cut out along the streak
and transferred to clean cover
slips. Care was taken to place
the upper surface of the block
next to the glass, after which
the whole was turned over a
Van Tieghem moist cell and
kept at room temperature. At
the end of 18 hours, by means
of a microscopic examination of
the agar block, with 2 mm. ob¬
jective (oil immersion) and No.
6 ocular, bacteria were selected
and their development through
several generations was observed
(%■ 3).

Feageeea. — The organism is

Fig. 2. — Filaments of Bacterium aptatum taken from
the condensation water from a 2-day-old agar
culture; stained with carbol fuchsin: o, Segmented;

b, unsegmented. _ _

motile by means of polar flagella, varying from one to several at each
pole. In general, the number is one to two, but occasionally three
occur. The best results in staining flagella were
obtained by the use of Toeffler’s stain, with acid
mordant correction. Five drops of sulphuric acid
(the acid of such dilution that i c. c. is neutralized
by the same amount of i per cent sodium
hydroxid) were added to 1 5 c. c. of mordant. The
flagella are threadlike, frequently wavy and some¬
what tapering, often forming a loop or coil at the
distal end, and are about twice as long as the body of the bacterium,
actual measurement of 10 flagella giving an average of 4 ft (fig. 4).

1 7072 ° - 13 - 2

8 8 Hi I k

abode T

Fig. 3. — Process of cell division
as seen in an x8-hour-old
hanging drop culture of Bac¬
terium aptatum. Time, a to
/, 5* minutes.

196

Journal of Agricultural Research

Vol. I, No. 3

Question of EndosporES. — No spores have been demonstrated either
by staining or testing with heat. Vacuolated forms were seen in cultures
stained with spore stains. Several tests with heat were made, bouillon
cultures 2 to 6 months old being treated as follows: Two were boiled
three minutes and two were kept at 8o° C. for 20 minutes; then transfers
were made from both sets. These transfers were watched for nearly a
month, but no trace of growth was seen. The transfers made before
heating, as checks, showed a vigorous growth of the organism in two days.
Prom these results it appears that spores are not formed by this bac¬
terium, since, if present, they would have been carried over after the
death of the vegetative cell, and growth would have been apparent in the
new transfers. The fact that the bacterium is quite easily killed by

atmospheric drying points to the same con¬
clusion in regard to the absence of spores.

Involution Forms. — Involution forms
are not common, but a few Y-shaped and
cross-shaped forms were noticed in old cul¬
tures grown in media not favorable *for the
best development of the organism, such as
beef bouillon containing 0.2 per cent of
tartaric acid, or beef bouillon containing 0.1
per cent of oxalic acid. Some were found
Fig. 4.— Bacterium aptatum showing in ordinary media which had been placed

flagella from a 2-day-old agar culture; under Unusual Conditions,
stained with Loeffler’s flagella stain.

Capsules. — No capsules have been dem¬
onstrated. The organism is viscid after growing three days on agar and
five to seven days in bouillon. Ribbert’s and Richard Muir’s capsule
stains were used.

Zooglcele. — Pseudozoogloese occur in +15 bouillon, Fermi’s solution,
bouillon containing salt, acids, and alkalies, and other liquid media in
which the growth rises in a viscid swirl when the tube is shaken. When
examined under the microscope, the viscid mass is found nearly always
to be made up of short rods held in place by a network of gelatinous
threads. Sometimes the mass is composed largely of the unsegmented
filamentous bacteria (fig. 2,6).

BEHAVIOR TOWARD STAINS

The organism stains readily and uniformly in the ordinary basic
aniline stains, such as methyl violet, gentian violet, saffranin, dahlia,
fuchsin, and carbol fuchsin. It is not acid-fast and does not stain by
Gram.

cultural characters

In general, the organism grows well upon many different kinds of
artificial media, the most favorable for rapid and prolonged growth
being +15 beef agar and bouillon, upon which it has been observed to
live from 9 to 12 months.

Dec. io, 1913

Disease of Sugar-Beet and Nasturtium Leaves

197

Agar Plates. — At a temperature of 20° to 240 C. the colonies on
peptonized beef agar (+15 on Fuller's scale) are up in 24 to 36 hours
when plates are poured from a young bouillon culture. They are round,
smooth, flat, glistening, 1 to 2 mm. in diameter, with entire edge, fish¬
scalelike markings, whitish in reflected light, bluish in transmitted light.
In three days the colonies are 4 to 5 mm. in diameter on plates thinly
sown, and the agar has changed to a faint yellowish green color. In 7
to 10 days the colonies are a deep-cream color.

Agar Stroke. — There is a moderate growth along the stroke in 24
hours. It is whitish, flat, smooth, and glistening, spreading at base. In
two days there is a heavy growth; in four days the agar has changed
to a slight yellowish green, with the growth of a viscid consistency. In
from five to seven days the bacterial growth covers nearly the entire
surface of the agar, densely clouds the condensation water, and becomes
slightly malodorous. The greatest growth of the organism occurs at the
base of the stroke and in the condensation water. The margin of the
stroke is often scalloped, with some edges of the scallop thinner than
others. The growth on old cultures is a deep-cream color, the medium
having become brown.

Tests made with the same organism, transferred at intervals for
several years to artificial media, showed that the greenish color was not
always produced in agar.

Agar Stab. — Growth is slow in stab culture, only a slight trace occur¬
ring in two days. It is best at the surface; very little along the line
of puncture. In four days the entire surface of the agar is covered with
a whitish, smooth growth, and the agar at the top has changed to a
faint yellowish green. Many crystals occur in the path of the needle.
The agar is not liquefied or softened.

Beef-Bouillon Cultures. — A slight clouding is noticeable in beef-
bouillon (+15) cultures within 18 to 22 hours at room temperature
(220 to 250 C.), increasing in density until a thick, viscid sediment forms
in the bottom of the tube. When shaken, this sediment rises in a thick
coherent, ropelike swirl. In bouillon cultures of three to five days*
growth the solution becomes slightly greened, and a thin, whitish pellicle
forms on the surface. This pellicle, which is composed of small masses
of bacteria, is easily disturbed when shaken and falls in hundreds of
tiny particles. In two weeks the medium has nearly cleared, a thick,
whitish sediment has accumulated, and the solution is apple green in
color, the fluorescence being most distinct toward the surface. In two
months the medium has changed to a dark-amber color (Ridgway's
“tawny”). Crystals may or may not occur.

Neutral Beef Bouillon. — Growth occurs in 22 to 24 hours. There
is a good growth in five days, and the medium has become a faint yel¬
lowish green color.

198

Journal of Agricultural Research

Vol. I, No. 3

Bouillon Containing Sodium Chlorid. — Growth occurred in neutral
bouillon containing 5 per cent of sodium chlorid when tests were made
with the organism soon after first isolating. Three years later these
tests were repeated. Growth then took place in bouillon containing 3
and 3X per cent of sodium chlorid, but there was no growth in bouillon
to which 4 per cent of sodium chlorid was added.

Bouillon Over Chloroform.— There is retardation of growth for two
days; then the bouillon clouds and in nine days is colored a yellowish
green tinge, as in the +15 bouillon without chloroform.

Nitrate-Bouillon Cultures. — In nitrate bouillon a thin clouding is
produced within 24 hours, and in four days the solution is distinctly
clouded, especially in its upper portion, where pseudozooglceaelike masses
are visible. In eight days the thin pellicle which forms on the surface
is easily shaken into many small particles. At this time a slight greenish
cast appears in the solution. The same ropelike sediment described in
beef bouillon was observed in a 9-week’s-old culture of nitrate bouillon.

Uschinsky’s Solution. — In plain Uschinsky’s solution and in the
peptonized solution (1 per cent) strong clouding was produced in three
to five days. In four days a thin pellicle composed of pseudozooglceae¬
like masses was observed. A greenish fluorescence became visible in
five to eight days, and in three weeks the uniformly clouded solution
had turned pale green (No. 328B, Code des Couleurs, Klincksieck et
Valette).

Fermi’s Solution. — There is a slight clouding in one day. In five
days there is a thick tenacious pellicle, and the medium has changed to
a decided pea-green color. A few fragments on the underside of the
pellicle are suspended in the medium, and these occur in long gelatinous
strings. On shaking the culture it is difficult to break up the pellicle
and cause it to sink. In one month this pellicle is from 3 to 4 mm,
thick.

Cohn’s Solution. — The organism does not grow in Cohn’s solution.

Sterile Milk. — The milk is cleared slightly in two to four days,
showing a gradual separation of whey from curd. This separation
begins on the surface as a watery band and gradually extends downward,
becoming complete in 12 days when kept at room temperature from 18
to 220 C. The medium is a yellowish cream color with a suggestion of
green. There is a slight rim, but no pellicle. In one month the medium
has become darker, and the green tinge has disappeared. It is trans¬
lucent throughout. Compared with Ridgway’s Color Chart, it is a clay
color. After two months at room temperature the cultures are dried
down 5 c. c., and are of a thick, creamy consistency. Transfers from
these cultures showed that the organism was still alive.

Litmus Milk. — In two days a blue ring appears at the surface of the
liquid, extending down about 1 cm. In four days there are three rings

Dec, io, 1913

Disease of Sugar-Beet and Nasturtium Leaves

199

of graded shades of blue, while the lowest third of the liquid remains the
color of the check tubes. Six to eight days later none of the original color
of the liquid remains. Some tubes have four or five rings of color, the
upper ring being the darkest blue. From 12 to 15 days after inoculating
a brownish color appears at the bottom of the tube and extends up¬
ward, changing the entire liquid to a muddy blue in from three to six
days. About four days later the medium begins to change to blue again
and in seven days is entirely blue, approaching Ridgway’s plum purple.
Four different tests were made in which the color changes followed in
this same manner. Room temperature, 180 to 220 C.

Gelatin Plates. — Colonies of the bacterium which appear on gelatin
( + 10) plates within 48 hours are whitish, round, and glistening, with a
smooth, flat surface having fishscalelike markings. Slight liquefaction
began in two days at a temperature of 20° to 22 0 C., causing small clear
areas around the colonies. In thickly sown plates liquefaction pro¬
ceeded rapidly, becoming complete in three to five days. In plates
thinly sown the liquefaction is only in cuplike areas about the colonies.
When liquefied, the gelatin becomes a turbid, slightly greenish fluid.

Gelatin Stab Cultures. — In gelatin ( + 10) stabs, growth was visible
in two days on the surface about the stab, extending downward about
1 cm. (temperature 20° to 22 0 C.). Craterlike depressions with fluid
contents were observed on the third day, increasing in size until a layer
of fluid was formed. In 10 days this layer had become 1 cm. in depth.
Liquefaction of the gelatin stab culture was complete in 30 days.

Steamed Potato Cylinders, — In three days growth on this medium
is abundant, flat, smooth, cream white, and glistening. The potato
changes to a gray-brown color in 3 days, and in 1 5 days is from two
to four shades darker. The bacterial slime approaches Ridgway’s wood
brown. There is no diastasic action of the starch.

Starch Jelly. — Growth is scant on starch jelly. In seven days the
medium at the surface and about 3 mm. below the streak along which
the growth of the organism has taken place has changed to a delicate
green. The test for sugar with Fettling’ s solution was negative.

LoEFRlER's Blood Serum. — The growth is moderate and slow, scarcely
a trace occurring in three days. The medium becomes gray and at the
end of 32 days has liquefied a little. The stroke is filiform, flat, glisten¬
ing, and smooth. The heaviest growth occurs in the condensation
water.

Litmus-Lactose Agar. — Copious growth developed within two
weeks in litmus-lactose agar cultures. The condensation water first
clouded, after which growth began to show at the base of the stroke. In
eight days there was growth along the entire stroke, with a spreading
at the base and a pellicle formation in the condensation water. The
medium was blued. At the end of nine weeks the growth was azure
blue in color (No. 401, Code des Couleurs, Klincksieck et Valette).

200

Journal of Agricultural Research

Vol. I, No. 3

Gentian-Violet Agar. — Growth of the bacterium on gentian-violet
agar was very slow, no growth being visible in 4 days and only a
slight growth in 18 days. When examined four weeks after inocu¬
lation, however, a thin bluish growth was observed along the stroke and
spreading from the base over the surface of the slant. The medium had
paled, some of the violet color having been extracted by the bacterium
in its growth.

OTHER CULTURAL FEATURES OF THE ORGANISM

Nitrates. — Nitrates are not reduced. Tests were made with nitrate
bouillon cultures 5 and 10 days old in the following manner: 1 c. c. of a
potato-starch solution was added to each culture, then 1 c. c. of a fresh
potassium-iodid solution (1 1250), after which 5 drops of dilute sulphuric
acid (2:1) were added. There was no change of color in any of the 5
or 10 day old cultures.

Indol* — No indol is present in cultures 1 to 10 days old. It is present,
however, in cultures n to 25 days old. The tests were made as follows:

Transfers were made from a 2-day-old bouillon culture to Uschinsky’s
solution containing 2 per cent of peptone. These cultures grew at room
temperature, 20° to 240 C., tests being made at the end of 1, 3, 5, 8, 10,
11, 12, 13, 15, and 25 days. Ten drops of concentrated sulphuric acid
were added to each culture to be tested and after standing for five min¬
utes, 1 c. c. of a 0.02 per cent solution of sodium nitrite was added. If
no pink color was visible in the cultures five minutes after adding the
nitrite, the tubes were heated to a temperature between 70° and 8o° C.
The rose color which indicates the presence of indol was not present in
any of the tests up to the tenth day.1 Indol was present in some of the
11-day cultures, but in the 15-day and 25-day cultures each one gave
the definite rose-color reaction.

TEST FOR HYDROGEN SULPHID

No hydrogen sulphid is produced. Litmus-lactose agar slants were
inoculated from a 2-day beef-agar culture. Small strips of filter paper
previously moistened in a saturated solution of lead acetate were inserted
in the tubes, being held in place by means of cotton plugs in such a man¬
ner as to prevent contact with the medium. In two days there was
growth along the entire stroke, accompanied by a bluing of the agar, but
without any discoloration of the filter paper. In six days the bacterial
growth had become abundant, spreading at the base of the stroke and
filling the condensation water. During a period of four weeks there was
no evidence of hydrogen sulphid. The test was repeated with litmus-
lactose agar, beef agar, and beef-bouillon cultures with the same result.

1 In a few instances a faint pinkish color appeared on the tenth day in tests made with the nasturtium
strain of the organism.

Dec. io, 1913

Disease of Sugar-Beet and Nasturtium Leaves

201

TEST FOR AMMONIA

The organism produces ammonia. Beef-bouillon cultures (2 to 8
weeks old) were tested with Nessler’s solution. Strips of filter paper
were moistened with the solution and suspended in the tubes to be
tested. The cultures were then heated in a water bath. A brownish
red color appeared on the filter paper and in the drops of distillate which
collected on the sides of the tube. This coloration indicated the pres¬
ence of ammonia in the cultures. A second test for ammonia was
made by placing 25 c. c. of the Nessler’s solution in large-sized tubes.
Ordinary test tubes of beef bouillon inoculated with the bacterium were
put into these larger tubes. The inner tubes were left open and the
outer tubes closed with cotton plugs. After five days a brownish pre¬
cipitate had formed in the Nessler's solution, forming a ring on the glass
tubes at the surface of the liquid. Check tubes used in both tests did
not show this precipitation.

TOLERATION OF ACIDS

Toleration of acids by the bacterium was tested in different percentages
of tartaric, oxalic, and hydrochloric acid made up in beef bouillon.
The organism was transferred from bouillon to acid cultures ranging
from 0.1 per cent to 0.3 per cent solutions. Clouding occurred in 1 day
in the tartaric acid in a 0.2 per cent solution, but there was no clouding
in 10 days in a 0.3 per cent solution. In a 0.1 per cent solution oxalic
add there was slight clouding in 1 day, moderate clouding in 2 days,
and strong clouding in 3 days, but no clouding in a 0.2 per cent solu¬
tion. In the 0.1 per cent solution of hydrochloric acid, growth was
slow in appearing; the solution became turbid in 1 to 2 weeks, and a
greenish color was produced in the medium. No growth occurred in
0.125 per cent solution of hydrochloric acid during 10 days. A final
test for acid toleration was made in beef bouillon containing hydro¬
chloric and tartaric acids (titrating on Fuller's scale from +19 to +35).
Results of this test showed heavy clouding in 5 days in + 30 beef solu¬
tion of both hydrochloric and tartaric acids, while no trace of clouding
appeared in the + 35 acid bouillons during 4 weeks.

Toleration of Sodium Hydroxid. — The toleration of sodium hydroxid
by the bacterium is moderate. Transfers from a 7-day beef-bouillon
culture clouded —15 beef bouillon in 1 to 2 days, —18 in 10 days, and
occasionally a slight growth occurred in ---20 after 2 weeks, but there
was no clouding in — 25 beef bouillon during a period of 4 weeks.

Optimum Reaction for Growth in Bouillon. — The optimum reac¬
tion for growth in beef bouillon is between +15 and +30; the organism
grows nearly as well at +25 as at +15, and the medium becomes fluores¬
cent as in +15.

202

Journal of Agricultural Research

Vol. I, No. 3

Gas Formation. — The organism is aerobic and does not form gas.
Tests were made in fermentation tubes with water containing 2 per
cent of Witte’s peptone to which was added 1 per cent of each of the
following carbon compounds: Glycerin, saccharose, mannite, maltose,
dextrose, and lactose. (Levulose and galactose were used in addition
with the strain of the organisms isolated from nasturtium.) No gas
formed in any of the tubes. Because of differences between the two
strains in regard to the clouding of solutions in the closed end of some
of the fermentation tubes, the results of the tests are given separately.

With the organism isolated from sugar beet there was a heavy growth
in the open arm of the tubes, but none in the closed ends. Dextrose
and saccharose gave an acid test with litmus after the organism had
been growing in the tubes 16 days. Glycerin, mannite, maltose, and
lactose gave an alkaline test.

From inoculations with the organism isolated from nasturtium the
following readings were made after 5, 10, and 28 days:

Table I. — Readings from fermentation tubes inoculated with the nasturtium strain of the

bacterium.

Peptonized
water with
per cent solu¬
tion of—

After 5 days.

After 10 days.

After 28 days.

Lactose. .

Levulose

Maltose. .

Solution clouded
in open end.

— do. . .

_ do .

Mannite

Glycerin

Dextrose

Clouded in open end and outer
two-thirds of U tube; sharp line
of demarcation; perfectly clear
in closed end; no pellicle; litmus
test, alkaline.

Clouded in open end and outer
two-thirds of U tube; clear in
closed end; no pellicle and no
flocculence; litmus test, alkaline.

Uniformly clouded in open end
and outer two-thirds of U tube;
sharp line of demarcation; no
pellicle; clear in closed end; lit¬
mus test, alkaline.

Clouded in open end and in U
tube; no sharp line of demarca¬
tion; no pellicle; perfectly clear
in closed end; litmus test, alka¬
line.

Uniformly clouded in open end
and outer two-thirds of U tube;
sharp line of demarcation; no
pellicle; clear in closed end; lit¬
mus test, alkaline.

Uniform clouding in open end and
whole of U tube; no pellicle;
clear in closed end; litmus test,
acid.

Galactose.

Clouded in open end and in U
tube; no distinct line of demar¬
cation; faint clouding in closed
end; no pellicle; litmus test, dis¬
tinctly acid.

Saccharose.

Uniformly clouded in open end
and in U tube; no sharp line of
demarcation; no pellicle; clear in
closed end; litmus test, feebly
acid.

Clouded in open end and outer
part of U tube; whitish pre¬
cipitate; no growth in closed
end; litmus test, alkaline.

Clouded in open end and outer
U tube; clear in closed end;
litmus test, alkaline.

Clouded in open end, with
whitish precipitate; no
growth in closed end; litmus
test, alkaline.

Clouded in open end and U
tube; clear in closed end; no
pellicle; whitish precipitate;
litmus test, alkaline.

Clouded in open end, with
whitish precipitate; clear in
closed end; litmus test, alka¬
line.

Well clouded in open end, with
numerous small particles in
suspension; closed end clear,
except a slight clouding in
lower end; no pellicle; litmus
test, distinctly acid.

Well clouded in open end, with
many small particles in sus¬
pension; clouded in two-
thirds of closed end; no pel¬
licle; considerable precipi¬
tate; litmus test, distinctly
acid.

Thinly and uniformly clouded
in open end and outer two-
thirds of U tube; sharp line of
demarcation; clear in closed
end ; no pellicle; whitish pre¬
cipitate; litmus test, dis¬
tinctly add.

Dec. IO, 1913 Disease of Sugar-Beet and Nasturtium Leaves .

203

From Table I it may be seen that growth occurs in the open end of the
fermentation tube in each of the nine solutions tried, while in the closed
end there is slight clouding in dextrose and a distinct clouding in presence
of galactose. In the test for alkaline and acid reactions neutral litmus
paper was used. As a result of this test six of the sugar solutions showed
an alkaline reaction and three (dextrose, galactose, and saccharose)
showed a distinctly acid reaction. No gas formation was observed in the
closed arm of any of the solutions during a period of 30 days.

TEST FOR ANAEROBISM

The organism will not grow in an atmosphere deprived of oxygen. The
test was made as follows:

Fresh transfers were made to beef bouillon from a 24-hour bouillon
culture and placed in a Novy jar containing a solution of pyrogallic acid
and sodium hydroxid (1 gram of pyrogallic acid to 10 c. c. of a 10 per cent
solution of sodium hydroxid for each 100 c. c. of air space).

The control cultures were kept under normal conditions at room tem¬
perature.

The Novy jar was waxed and clamped tightly and connected on one
side to a series of wash bottles containing pyrogallic acid and sodium
hydroxid and on the other side to the exhaust. There were stopcocks to
regulate the passing of the gasses through the jar. In the jar with the
cultures was a fermentation tube which had its closed arm filled with
water except for a bubble of air at the top. This bubble was noted as
an indicator of pressure within the jar. As the oxygen was absorbed by
the solution within the jar, air was allowed to pass in from the wash
bottles until the bubble in the fermentation tube indicated the normal
pressure. The exhaust was used to draw off the gases from the jar.

The operation was repeated several times during a period of three hours,
after which the Novy jar was sealed and set aside. The atmosphere in
the jar was then practically one of nitrogen. At the end of six days the
cultures were taken from the jar and examined. There was no trace of
clouding in the bouillon. The controls, however, showed heavy growth;
in fact they were heavily clouded within two days.

This test was made a second time, the Novy jar being set up in the same
way and the bouillon transfers made from a 24-hour culture as before.
This time the jar was sealed for two weeks. When it was opened no
growth could be detected in any of the bouillon cultures, while the con¬
trols showed the usual heavy growth after two days. The cultures which
had been kept in the Novy jar were clouded heavily five days after they
were removed.

TEMPERATURE RELATIONS

Thermal Death Point. — The thermal death point is 47.50 to 48° C.
when transfers are made from a 24-hour bouillon culture and the inocu¬
lated tubes are kept at that temperature in the water bath for 10 minutes,

204

Journal of Agricultural Research

Vol. I, No. 3

readings being taken at half -minute intervals during that time. Many
tests were made, using for transfers +15 bouillon cultures 18 hours to 6
days old. When 3 to 6 day old cultures were used and kept in the water
bath for 10 minutes at 510, the organism was not killed; nor was it killed
at 530 C. for the same length of time.

Maximum Temperature. — The maximum temperature for the organ¬
ism isolated from sugar beet is 35 0 C., while the maximum temperature
for the organism from nasturtium is 330 to 340 C.

Minimum Temperature. — The minimum temperature is between o°
and — 1 0 C. When kept at a temperature of — 20 to — 50 C. for five days
by means of an ice and salt mixture, the organism remains alive and
begins to grow after being restored to room temperature. A good growth
of the organism occurs in both agar and bouillon at 11.50 C. A fair
growth occurs in bouillon at 8° C.

Optimum Temperature. — The optimum temperature is 270 to 28° C.

relation to light

The organism is not especially sensitive to sunlight. Thinly sown
agar poured plates were exposed in bright sunlight at midday in mid¬
winter on bags of crushed ice out of doors, half of each plate being cov¬
ered with black paper to serve as a check. The test with the organism
isolated from sugar beet was as follows:

Fifty minutes exposure did not kill the organism, for colonies appeared
on the exposed side of these plates in two days, but no colonies appeared
on those plates exposed 60 minutes. Three different tests were made.
The organism isolated from nasturtium proved more resistant to sunlight,
since a few scattered colonies appeared on the agar plates even after an
exposure of 80 minutes.

RELATION TO MOISTURE

The beet organism is killed very readily by drying, even at a moderate
or low temperature. When drops of a 1 -day-old, well-clouded bouillon
culture are placed on sterile cover glasses and kept in the dark at a tem¬
perature of 21 0 to 2 50 C. from four to five hours, growth occurs in bouillon
tubes into which these covers are dropped. When kept six hours, all
the organisms are dead. With 3 to 6 day old cultures treated in the
same way the organism was able to withstand drying from one to three
days.

VITALITY IN CULTURE MEDIA

This organism lives from 10 to 12 months in liquid media, such as beef
bouillon, sterile milk, and Fermi’s solution, when kept at temperatures
varying from 1 1 0 to 20° C. Bouillon cultures may die in four months and
less when the plugs in the tubes are loose and such rapid evaporation
occurs that the culture dries down. This usually takes place in the

Dec. 10, 1913

Disease of Sugar-Beet and Nasturtium Leaves

205

summer at room temperature, 2 40 to 30° C. Beef-agar cultures live
from 4 to 10 months, depending upon the temperature under which they
are grown. Those cultures which die in from four to five months are
grown at temperatures of 240 to 30° C.

LOSS OF VIRULENCE

No loss of virulence was noticed in the organism isolated from nastur¬
tium until April, 1910 (two years after the first isolation), when inocula¬
tions were made into nasturtium and bean plants growing in the green¬
house. Five days after inoculation no apparent discoloration of the
tissue could be observed. This result was unusual, since in all past inoc¬
ulations the diseased spots had been readily produced. After repeated
inoculations had been made from cultures of the bacterium grown in
beef bouillon upon agar slants and potato cylinders it became evident
that the organism, which had been growing on artificial media for two
years, had lost its virulence.

In the case of the organism isolated from the sugar-beet leaf, no loss of
virulence was noticed until about three years after obtaining the organism,
and up to that time practically every needle-prick inoculation into sugar-
beet leaves proved infectious. After three years the percentage of posi¬
tive results from inoculations fell off considerably, as only the youngest
leaves, growing under the proper conditions of moisture and temperature,
became diseased. Efforts were made in the summer of 1911 to obtain a
new strain of the organism from the field, but they were unsuccessful.
Later, string-bean agar was tried and proved to be a rejuvenator of the
organism isolated from both hosts. After growing on this medium, the
organism was almost as infectious to sugar-beet leaves and nasturtium
leaves as when it was first isolated. This virulence, however, was not
permanent, for in the course of a year it became much reduced.

BACTERIA IN CELL TISSUE

Diseased tissue produced in both hosts by inoculation was fixed, em¬
bedded in paraffin, sectioned, and stained in carbol fuchsin. Microscopic
examinations of these sections showed the presence of bacteria in large
quantities within the cells of the diseased tissue (fig. 5). In sections cut
through the central portion of the diseased spots the walls appeared
ruptured or collapsed. The cells at the margins of these ruptured places
show that the bacteria are in the cells, although most of the bacteria
were seen in the broken-down tissues adjacent to the sound cells.

NATURAL INFECTION AND CONTROL

Since practically all of the work has been done under laboratory and
greenhouse conditions, there has been no opportunity to investigate the
complete life cycle of this organism or to follow out the natural means of

206

Journal of Agricultural Research

Vol. I, No. 3

infection in the field. This being the case, no practical methods of
control have been undertaken, but in order to determine if possible some¬
thing in regard to the way in which the organism gains an entrance into
the tissue of its hosts, young plants were placed in infection cages in the
greenhouse and the foliage sprayed with a bacterial solution until it was
thoroughly wet. This solution was prepared from 5-day-old cultures of
the organism. Check plants were placed in a control-infection cage and
sprayed with distilled water. Examination was made at intervals of
several days, but no diseased spots appeared on either the nasturtium
or sugar-beet leaves during a period of 20 days. The result of the experi¬
ment suggests that infection takes place only in bruised or wounded

tissue, due to insects or to mechanical
injury.

TECHNICAL DESCRIPTION OF THE
ORGANISM

Bacterium aptatum, n. sp.

According to the numerical designations
adopted by the Society of American Bacte¬
riologists, the group number of Bacterium apta¬
tum is 211. 2322133.

Form, a short motile rod with rounded ends;
flagella, bipolar; involution forms rare; no
spores or capsules observed; pseudozoogloeae
occur; aerobic ; smooth whitish colonies on agar
plate with fishscalelike markings; clouds beef
bouillon in 18 to 24 hours; produces alkaline
reaction in litmus milk, with a gradual sepa¬
ration of whey from curd; liquefies gelatin;
produces ammonia; no reduction of nitrates;
fluorescence greenish; no diastasic action on
potato starch ; grows in Uschinsky’s and Fer¬
mi’s solutions; indol produced after 10 days;
optimum temperature 270 to 28° C.; maximum
340 to 350 C.; minimum — i°C.; thermal death point 47. 50 to 48° C.; vitality 4
to 10 months in beef agar, 10 to 12 months in bouillon, depending on temperature;
good growth on litmus-lactose agar; growth much retarded on gentian-violet agar; stains
readily with basic anilin dyes; not acid-fast; not stained by Gram ; tolerates acids;
oxalic, 0.1 per cent; tartaric, .0.2 per cent; hydrochloric, 0.1 percent; tolerates sodium
hydroxid in beef bouillon, —18 Fuller’s scale; no growth in Cohn’s solution; killed
readily by drying; not very sensitive to sunlight; retains its virulence 2 to 3 years;
pathogenic to nasturtium, sugar-beet, and several other plants.

COMPARISON OF PSEUDOMONAS TENUIS WITH BACTERIUM APTATUM

While the work on Bacterium aptatum was being prepared for publica¬
tion, Bulletin No. 167 of the Vermont Experiment Station was received,1
part 3 of which contains a description of green fluorescent bacteria

Fig. 5. — Camera-lucida drawing of a portion
of a cross section of sugar-beet leaf inocu¬
lated with Bacterium aptatum . The cells
containing bacteria were next to many
collapsed cells.

1 Hdson, H. A., Jones, C. H., and Carpenter, C. W. Micro-organisms of maple sap. Vermont Agr. Exp.
Sta. Bui. 167, p. 321-610, 14 fig., 16 pi., 1912.

Dec. 10, 1913

Disease of Sugar-Beet and Nasturtium Leaves

207

occurring in maple sap. Results of a comparative study of seven repre¬
sentative strains of the green fluorescent sap bacteria and six known
fluorescent species are given, and the group numbers of these organisms
determined. Since one of these numbers, that of Pseudomonas tenuis
is identical with the group number of Bacterium aptatum , it was found
necessary to make cultural comparisons. A culture of Pseudomonas
tenuis was obtained by Dr. Erwin F. Smith from Mr. C. E. A. Winslow,
American Museum of Natural History, who stated that he had received it
from Mr. Edson.

Table II shows the results of comparative tests made with Pseudo -
monas tenuis and Bacterium aptatum .

Table II. — Comparison of the characteristics of Pseudomonas tenuis and Bacterium

aptatum.

Media, etc.

Pseudomonas tenuis.

Bacterium aptatum.

1. Bed bouillon .

2. Bed-agar stroke .

3. Uschinsky's solution.

4. Nitrate reduction .

5. Indol test .

6. Hydrogen - sulphid

test.

7. Gelatin plates .

Rapid clouding; green fluorescence;
distinct pellicle.

Smooth, thin, whitish growth;
medium greened.

Strong clouding with fluorescence;
pellicle formed.

None .

No indol in 10-day cultures, but
present in 16-day cultures.

Hydrogen sulphid produced . .

A trace of liqudaction in 3 weeks on
thickly sown plates.

3. Gelatin stabs . .

9. Sterilized milk

10. Litmus milk.

No liqudaction in 3 weeks .

Gradual thickening in 6 weeks with¬
out clearing.

Alkalin reaction; color uniform
throughout during 7 weeks.

11. Ammonia test . Ammonia produced . .

12. Pathogenicity . Nonpathogenic to sugar-beet and nas¬

turtium leaves.

Clouding with green fluorescence;
distinct pellicle.

Smooth, thin, whitish growth;
medium greened.

Strong clouding with fluorescence;
pellicle formed.

None.

Indol present in 10 to 12 days.

No hydrogen sulphid.

Liquefaction begins on second day and
is complete in 5 days in thickly sown
plates.

Liqudaction begins in 2 to 3 days.

Clearing begins in 2 to 3 days and is
completed in 2 weeks.

Alkalin reaction; banded appearance
resulting in clearing and a uniformly
blue color in 3 to 4 weeks.

Ammonia produced.

Pathogenic to sugar-beet and nastur¬
tium leaves.

From results given in Table II it is evident that Pseudomonas tenuis
and Bacterium aptatum} although closely related in the green fluorescent
group of bacteria, do not belong to the same species. Similarity of
growth occurs and was especially noticed in beef bouillon, on beef agar,
and in Uschinsky's solution. Pseudomonas tenuis , however, clouds bou¬
illon and Uschinsky’s solution more quickly than Bacterium aptatum.
Both organisms produce indol and ammonia. Neither reduces nitrates.
Pseudomonas tenuis has a strong putrefactive odor not present in cultures
of Bacterium aptatum. Pseudomonas tenuis produces hydrogen sulphid,
while Bacterium aptatum does not. In sterilized-milk cultures, Bacterium
aptatum gradually separates whey from curd, and in litmus milk this
process is accompanied by changes of color, giving a distinctly banded
appearance during the first week’s growth. Neither the separation of

1 Zimmermann, O. E. R. Die Bakterien unserer Trink- und Nutzwasser . . . Reihe 1, Chemnitz, 1890.
106 p. Also in 11. Bericht, Naturwissenschaftliche Gesellschaft, Chemnitz, 1887, 1889, p. 53-154. 1&90.

Thumm, Karl. Beitrage zur Biologie der fluoresderenden Bakterien. Arb. Bakt. Inst. Karlsruhe, Bd.
1, Heft 3; p. 291-377. [1895J

208

Journal of Agricultural Research

Vol. I, No. 3

whey from curd nor the color changes were apparent in cultures of
Pseudomonas tenuis during a period of seven weeks. One of the most
important cultural differences between these two organisms appeared on
gelatin plates. Bacterium aptatum is a rapid liquefier, while Pseudomonas
tenuis showed only a trace of liquefaction in three weeks, this slight
liquefaction occurring only on thickly sown plates and not at all in stab
cultures. The essential difference, however, between Bacterium aptatum
and Pseudomonas tenuis is not so much a cultural as a physiological one.
This is shown in the ability of Bacterium aptatum to produce diseased
spots on sugar-beet, nasturtium, and bean leaves, while Pseudomonas
tenuis is nonpathogenic to these hosts.

COMPARISON OF BACTERIUM PHASEOLI WITH BACTERIUM APTATUM

When it was observed that Bacterium aptatum 1 produced diseased
spots so readily on leaves of the bean plants, the question at once sug¬
gested itself as to the relation between this organism and Bacterium
phaseoli , the cause of the well-known bacterial blight cff bean, as de¬
scribed and worked out by Dr. Erwin F. Smith.1 2 The cultural charac¬
teristics of Bacterium aptatum were, therefore, compared with those of
Bacterium phaseoli. As a result of this comparison it is evident that
the two organisms are entirely different.

Some of the characteristic differences between the two organisms are
shown in Table III.

Table III. — Comparison of the cultural characteristics of Bacterium aptatum and

Bacterium phaseoli.

Media, etc.

Bacterium aptatum.

Bacterium phaseoli.

Beef agar (plate) . Whitish colonies, slightly bluish in

diffused light; medium greened.
Agar slant . Whitish, smooth, faintly blue in trans¬

mitted light; medium greened.

Potato slant

Litmus milk .

Thermal death point ....

Flagella .

Pathogenic to —

Resistance to dry air. . . .
Resistance to sunlight. . .
Color in mass .

Cream white to wood-brown; viscid;
medium browned; no diastasic
action.

Alkalin reaction; slow clearing during
seven weeks.

47-5° to 48° C .

Bipolar; one to several .

Nasturtium, sugar beet, bean, and
other plants.

Few hours to several days .

80+ minutes .

Whitish .

Yellow colonies, smooth, wet-shining;
thin, distinct margins.

Smooth, translucent, yellow; slimy
consistency; growth without retarda¬
tion.

Copious yellow slimy growth, medium
grayed; diastasic action powerful.

Slow alkalinity and separation of casein
from whey.

49-5 °C.

Polar; one.

Bean and lupine.

27 days.

30 to 45 minutes.

Yellow.

1 This comparison was made with Bacterium aptatum isolated from nasturtium.

2 Smith, B. F. Description of Bacillus phaseoli n. sp., with some remarks on related species. Proc,
Amer. Assoc. Adv. Sci., 46th meeting, 1897, p. 288-290, 1898.

- The cultural characters of Pseudomonas hyacinthi, Ps. campestris, Ps. phaseoli, and Ps. stew.

arti — four one-flagellate yellow bacteria parasitic on plants. U. S. Dept, of Agr., Div. Veg. Physiol, and
Path., Bui. 28, 153 p., illus., 1901.

- Bacteria in Relation to Plant Diseases, v. 2, Washington, D. C., 1911, p. 62. (Carnegie Inst.

Washington, Pub. 27, v. 2.)

Dec. io. 1913

Disease of Sugar-Beet and Nasturtium Leaves

209

COMPARISON OF BACTERIUM XANTHOCHEORUM WITH BACTERIUM

APTATUM

While investigations with Bacterium aptatum were in progress, atten¬
tion was called to the recent work of Dr. Julius Schuster upon a bacterial
decay of the potato tuber caused by Bacterium xanthocktorum } From
Dr. Schuster's description it was observed that in morphological and
certain cultural characters this potato bacterium resembled quite closely
Bacterium aptatum. Since both belong to the green fluorescent group
of bacteria, it seemed worth while to take up a comparative study of
the two organisms. Fortunately a culture of Dr. Schuster's Bacterium
xanthocktorum was at hand, having been brought to our laboratory by
Dr. H. W. Wollenweber in November, 1911. Accordingly a series of
cultural tests was begun at once and continued for a period of about
three months.1 2 As a result of these tests it is evident that Bacterium
aptatum and Bacterium xanthocktorum are not identical, although their
appearance is quite similar upon some kinds of culture media. Table
IV gives a partial record of the results obtained and will be sufficient
to show the differences.

Table IV. — Comparison of the cultural characteristics of Bacterium aptatum and

Bacterium xanthochlorum .

Media.

Bacterium aptatum.

Bacterium xanthochlorum.

+ 15 beef-agar plates .

+15 beef-agar stroke .

+ 15 beef-agar stab .

-f 10 gelatin plates .

-f is beef bouillon .

Potato cylinders .

Nitrate bouillon .

Sterile milk .

Litmus milk . .

Uschinsky’s solution ....

Litmus-lactose agar .

Growth less rapid than Bacterium
xanthochlorum; fishscalelike mark¬
ings on surface colonies pronounced.

Growth less rapid than Bacterium
xanthochlorum and greenish fluor¬
escence not so marked.

Growth whitish to drab color in center
of nail head.

Growth slower than Bacterium xan¬
thochlorum and liquefaction does
not begin so early; medium only
slightly greened.

Thin pellicle of pseudozoogloeselike
masses; sediment a ropelike viscid
swirl; fluorescense appears slowly.

Appearance similar to Bacterium xan¬
thochlorum.

Less rapid growth than Bacterium
xanthochlorum; pellicle easily break¬
ing into small particles; fluorescence
weak.

Slow separation of whey from curd;
no distinct fluorescence; pellicle of
floating islands.

Color of whey blue with whitish rim
formed around tube above solution;
pellicle not complete.

Clouding less dense than Bacterium
xanthochlorum; fluorescence mod¬
erate; pellicle composed of pseudo-
zoogloeae-like masses.

Growth less rapid than Bacterium
xanthochlorum; blue in color; me¬
dium blued; precipitate lead colored.

Growth more rapid and appearance of
colonies more compact than those of
Bacterium aptatum.

Growth rapid and fluorescence marked.

Growth pinkish colored in center of
nail head.

Growth and liquefaction rapid; me¬
dium distinctly greened.

Growth rapid; pellicle membranous
and falling entire; green fluorescence
striking.

Growth gradual; at first creamy white,
later brownish; starch not broken
down.

Growth rapid; pellicle membranous
and breaking into fragments; fluor¬
escence much greater than Bacterium
aptatum.

Separation of whey from curd more
rapid than in Bacterium aptatum;
pellicle more distinct; greenish fluo¬
rescence marked.

Color of whey grayish; rim above solu¬
tion pink to purplish; pellicle dis¬
tinct.

Clouding dense; pure green fluores¬
cence; membranous pellicle.

Growth rapid and dense; color of
growth, greenish blue; medium
blued; precipitate brownish.

1 Schuster, Julius. Zur Kenntnis der Bakterienfaule der KartofTel. Arb. K. Biol. Anst. Land- u.
Forstw., Bd. 8, Heft 4, p. 452-492, 13 fig., pi. 5, 1912.

2 The bacterium isolated from nasturtium leaves was used in these tests.

210

Journal of Agricultural Research

Vol. I, No. 3

Table; IV. — Comparison of the cultural characteristics of Bacterium aptatum and
Bacterium xanthochlorum — Continued .

Media.

Bacterium aptatum.

Bacterium xanthochlorum.

Gentian-violet agar .

Growth of streak much retarded; no
growth during first 4 days; after 18
days, moderate growth; medium
paled.

Acid reaction in peptonized saccha¬
rose, in peptonized galactose, and in
peptonized dextrose solutions.

No retardation; copious growth in two
days; blue in color; medium greened.

Alkaline reaction in peptonized saccha¬
rose solution; acid reaction in pep¬
tonized galactose and in peptonized
dextrose solutions.

Fermentation tubes .

SUMMARY

1. The leaf -spot diseases of sugar beet and nasturtium described in
this paper are due to a bacterial organism.

2. The two diseases occurred during the same summer. The causal
organism was isolated in pure cultures from both hosts and proved
infectious to sugar-beet and nasturtium leaves interchangeably.

3. It is proved from cultural, morphological, and inoculation tests that
the organisms causing these leaf-spot diseases on both hosts are identical.

4. The organism is also infectious to bean leaves and pods, lettuce,
pepper, and eggplant.

5. It probably enters the plant through wounds or by means of insect
injuries and may be spread by insects.

6. The organism is a bacterium belonging to the green fluorescent
group. It is proved to be different from Bacterium xanthochlorum ,
which is pathogenic to potato, and from Pseudomonas tenuis , which has
been given the same group number.

7. It is also different from Bacterium phaseoli , although both organ¬
isms produce spotting of bean leaves and pods.

8. The name Bacterium aptatum , n. sp., is suggested.

DESCRIPTION OF PLATES

Plate XVII. Fig. 1. — Sugar-beet leaves inoculated with Bacterium aptatum .
Photographed eight days after inoculation.

Fig. 2. — Sugar-beet root inoculated with Bacterium aptatum. Photo¬
graphed two weeks after inoculation.

XVIII (colored). Nasturtium leaves showing bacterial leaf spots 10 days after inocula¬
tion with Bacterium aptatum. (May, 1909.)

XIX. Fig. 1. — Bean leaves inoculated with Bacterium aptatum from leaf -
spot of sugar beet.

Fig. 2. — Nasturtium leaves inoculated with Bacterium aptatum from
leaf-spot of sugar beet.

Fig. 3 . — Bean pods inoculated with Bacterium aptatum from leaf-spot
of sugar beet.

(Inoculated Nov. 12, 1908; photographed Nov. 25, 1908.)

Disease of Sugar-Beet and Nasturtium Leaves

Plate XVIII

Journal of Agricultural Research

Vol I, No. 3

THE CALLIEPHIALTES PARASITE OF THE CODLING

MOTH

By R. A. Cushman

Entomological Assistant, Deciduous Fruit Insect Investigations, Bureau of Entomology

INTRODUCTION

The notes and observations on which the present paper is based
were obtained at Vienna, Va., under the direction of Prof. A. L. Quain-
tance, in Charge of Deciduous Fruit Insect Investigations, Bureau of
Entomology, the writer having been assigned to work on the parasites of
deciduous fruit insects at the Vienna laboratory in the spring of 1911.

So much has been published concerning the Calliephialtes parasite of
the codling moth, under the names Calliephialtes messor Grav. and
Ephialtes carbonarius Christ, since its introduction into California that it
seemed advisable to begin the work on the project with a study of this
species and its liberation on a large scale. The specimens with which
the start was made were obtained from two lots of parasitized codling-
moth larvae secured in 1911 from the California State Insectary. The
propagation from the first lot was unsuccessful, only three diminutive
males being reared. The second lot was received in the late summer.
These were reared to maturity, 15 females and a larger number of males
being secured. After these had mated they were given access to codling-
moth larvae that had been compelled to spin their cocoons in strips of
strawboard. The parasites oviposited very readily in the codling-moth
cocoons. The progeny of these individuals did not emerge until the fol¬
lowing spring. A large majority were lost in an attempt to force them
through to early maturity in a greenhouse, where, in spite of daily soak-
ings with water, the pupae dried up. A few females forced to maturity
in this way deposited eggs, but only males came from them. However,
21 females and 52 males were reared later from unforced material, and it
was with these that the real start in the work was made in the spring of
1912.

* During the season of 1912 several hundred individuals of both sexes
were reared under observation from egg to maturity. The results of
these observations are recorded in the following pages.

While the major part of the work was performed by the writer, it was
greatly facilitated by the work of Mr. J. D. Luckett, half of whose time
during the period from June 15 to September 15, 1912, was spent in
assisting in this work.

Journal of Agricultural Research, Vol. I, No. 3

Dept, of Agriculture, Washington, D. C. Dec. 10, 1913

K-3

(211)

17072 —13 — 3

212

Journal of Agricultural Research

Vol. I, No. 3

IDENTITY AND INTRODUCTION OF THE SPECIES

When the California State Horticultural Commission began its work of
introducing this parasite into California in an attempt to control the
codling moth, specimens were submitted to Dr. William H. Ashmead for
determination. Dr. Ashmead determined them as the Calliephialtes
mess or of Gravenhorst, a species inadequately described from a single
female specimen from Russia. Up to the time of the introduction into
California, C. messor had been mentioned in literature only once since its
description. This was by Taschenberg, who in 1863 recorded it as hav¬
ing been reared as a parasite of {Tinea) Galleria mellonella , the wax
moth.

When the writer took up the work on the species, specimens reared
from the codling moth in material sent to the Bureau of Entomology
from Sachsen, Germany, were submitted to Mr. H. L. Viereck, who
determined them as Calliephialtes comstockii Cress., a species described
from the United States. Later, specimens reared by the writer as prog¬
eny of the specimens received from California were sent to Dr. A. Roman,
of the Stockholm Museum. Dr. Roman reported that the museum had
no specimens of C. messor , but that those sent were identical with a
specimen determined for the museum by Dr. Ashmead as C. pusio
Walsh, another species described from America. The specimen in the
Stockholm Museum bears only the label “Long I.” Dr. Ashmead there¬
fore evidently determined the same thing under two specific names, one
European and the other American.

INTRODUCTION INTO CALIFORNIA

Late in 1904 Mr. George Compere, acting as an agent of the State
Horticultural Commission of California, found this species attacking the
codling moth in Spain. Living specimens were sent by him to Cali¬
fornia, where they were propagated and their progeny released in infested
orchards. At this time the species was supposed to be Ephialtes car-
bonarius Christ, and references to it under that name have appeared in
literature, but specimens from California were determined by Dr. William
H. Ashmead as messor Grav. and the species placed in his genus Cal¬
liephialtes. That it is not Calliephialtes carbonarius is firmly established
by the well-known habit of that species of attacking wood-boring insects.

In view of the uncertainty as to the specific identity of the parasite,
the writer has avoided the use of any specific name in the present paper.

INTRODUCTION INTO SOUTH AFRICA

Prom California specimens of the species were sent to the Cape of
Good Hope in 1907, where they were propagated and released by the
Government Entomologist, Prof. C. P. Lounsbury. Reports of the re¬
sults of this introduction indicate that it is of doubtful success.

Dec. io, 1913

Calliephialtes Parasite of Codling Moth

213

DESCRIPTION OF THE SPECIES
GENERAL description

The adult female is normally about half an inch long, exclusive of the
ovipositor, which about equals the body in length. It is of the char¬
acteristic pimpline appearance, long and slender, black in color, with the
legs red and the membranous portions of the venter white. The ovi¬
positor is straight for most of its length, but toward the tip curves some¬
what ventrally. The male is somewhat shorter and more slender than
the female, as is commonly the case in this group.

VARIATION IN SIZE

There is considerable variation in size, depending upon the abundance
of suitable larval food, a few individuals of each sex of not more than
half the normal dimensions having been reared. However, extremely
diminutive individuals are usually males.

TECHNICAL DESCRIPTION

Female. — Length 11 mm. ; ovipositor n mm., curving slightly ventrally at the tip;
abdomen about twice as long as thorax. Head and abdomen black; tegula and a
small triangular spot on the dorso-posterior angle of the mesonotum pale yellow, and
a very small spot on the dorsal border of the mesopleurum dark brown ; thorax other¬
wise black; palpi pale; antennae with two basal segments black, remaining segments
dark brown ; all legs uniform dark fulvous ; wings slightly brownish ; veins and stigma
brown. Thorax finely and sparsely punctate; propodeum more coarsely and densely
punctate, with a shining, impunctate, median depression; abdominal segments
coarsely and densely punctate; segments 2 to 5 with a smooth, shining, impressed
area on the posterior lateral angle. Sheath of ovipositor black, densely hairy; ovi¬
positor proper brown, shining.

Male. — Length 9.5 mm. ; more slender; otherwise, except in sexual characters, like
female.

DESCRIPTIONS OF THE THREE SPECIES TO WHICH THIS SPECIES HAS BEEN

REFERRED

Calliephialtes messor (Grav.).

Calliephialtes messor Gravenhorst was originally described in the genus
Ephialtes in 1 82 1 (1 )* from a unique female from Russia. Dalla Torre (5)
credits Gravenhorst with having recorded Tinea mellonella as a host of
this species, but this should be accredited to Taschenberg (2).

E . messor n. — Pedibus rufo-fulvis, tibiis posticis arcuatis. f. (aculeo longitudine
corporis).

Statura, imprimis proportione et tuberculis segmentorum, haec species medium
tenet inter antecedentem et sequentem; tibiis posticis arcuatis ab utraque differt.

Longitudo fere 7 linearum. Caput palpis fulvis. Thorax puncto parvo testaceo
ad radicem alarum. Alae testaceo-hyalinae, stigmate et radio fulvis, radice et
squamula stramineis, areola triangulari sessili. Pedes rufofulvi, postici tarsis fuscis,

1 Figures in parentheses refer to "Literature cited,” p. 235-237.

214

Journal of Agricultural Research

Vol. I, No. 3

tibiis arcuatis, supra fuscentibus. Abdomen thorace triplo longuis, eoque paulo
angustius, cylindricum, segmentis 3 et 4 latitudine paulo longioribus, 5-7 quadratis,
omnibus tuberculis lateralibus subprominentibus. Aculeus longitudine corporis,
terebra badia.

Unicam feminam Besser e Volhynia transmisit.

A translation of this description is given below.1

Calliephialtes comstockii (Cress.).

The only reference to Calliephialtes comstockii Cresson is the original
description published in 1880 (4). The type was reared as a parasite of
Retinia comstockiana Femald. It was referred to the genus Ephialtes.

Ephialtes comstockii Cresson, n. sp.

Female. — Black, shining; thorax smooth, very feebly punctured; metathorax
smooth, rounded, with two abbreviated, longitudinal, feebly developed elevated
lines on disk, slightly divergent posteriorly; tegulae white; wings hyaline, subiri-
descent, nervures and stigma fuscous, the latter with a pale spot at base, areolet as
usual; legs including coxae bright; posterior tibiae and tarsi black; abdomen about
twice the length of the thorax, distinctly punctured; sides of the second and following
segments tuberculated ; first segment a little longer than broad, broadly excavated at
base and slightly grooved on disk above ; second segment longer than broad, widened
posteriorly; third and fourth segments quadrate; remainder transverse; ovipositor as
long as the body; length of body .35 inch.

Habitat. — Ithaca, N. Y. Parasitic upon Retinia comstockiana Femald.

Calliephialtes pusio (Walsh.).

Calliephialtes pusio Walsh was originally described in 1873 (3) in the
genus Ephialtes without host record, this constituting the only reference
to the species in literature.

Ephialtes pusio , n. sp. — ? . Differs from gigas 9 as follows:

I 1. The size is 1/2 smaller. 2. The face is highly polished and scarcely punctate.
3. The metathoracic carinse are obsolete, being represented only by a slightly im¬
pressed stria extending 1 Is of the way to the tip . 4. The carinae of the first abdominal
joint are entirely obsolete. 5. The relative proportions of the first 5 abdominal
joints are quite different, 2-4 being equal in length and each twice as long as wide,
and 1 about 1/4 shorter, and 5 a trifle shorter than 2-4. 6. The usual tubercles are

obvious only on 3 and 4, and are much less prominent and round, not elongated. 7.
The ovipositor is rather piceous than black. 8. The legs are pale rufous, all the
sutures a little darker, but both trochanters of the front leg, and the outermost one in
the middle and hind leg, are whitish ; and in the front leg the tarsal tip, in the middle
leg the exterior face of the tibia and the whole tarsus, and in the hind leg the extreme
tip of the femur and the whole tibia and tarsus, are pale fuscous. 9. The wings are
subhyaline. Length 9 .60 inch; front wing 9 .36 inch; length abdomen 9 42 inch;
width abdomen 9 *06 inch; ovipositor .85 inch.

1 E[phialtes ] messor , n. sp. — Feet rufo-fulvous, posterior tibiae arcuate, female with the ovipositor as long
as the body.

In habitus, especially in proportions and in the tubercles of the segments, this species stands midway
between the preceding [i. e. , E. tubercuiatus ] and the following [i. e. , E. manifestator ] ; in its arcuate posterior
tibiae it differs from both.

Length about 7 lines. Head with the palpi fulvous. Thorax with a small testaceous spot at the base
of the wing; wings testaceo-hyaline, stigma and radius fulvous, base and tegulae stramineous, areolet
triangular and sessile; legs rufo-fulvous; posterior tarsi fuscous; tibiae arcuate, shading to fuscous above;
abdomen three times as long as the thorax, and slightly narrower, cylindrical, segments 3 and 4 slightly
longer than broad, s to 7 quadrate, all lateral tubercles subprominent; ovipositor as long as the body,
terebra brown.

A single female sent by Besser from Volhynia.

Dec. 10, 1913

Calliephialtes Parasite of Codling Moth

215

METHODS AND APPARATUS USED IN PROPAGATION

The most convenient and successful cage devised, the one in use at
present, is constructed as follows:

A glass cylinder about 6 inches in diameter and 10 inches long is laid
on its side in a baseboard constructed to keep the cylinder from rolling.
The back end is covered with cheesecloth held in place by rubber bands.
The front is a frame about 1 2 inches square, over which is tightly stretched
a piece of cheesecloth. This is held against the front of the cylinder by
means of rubber bands stretched between nails at the side of the frame
and the side of the baseboard, permitting access to the cage without
actually removing the front frame, by simply pulling the frame down,
as the rubber bands will stretch sufficiently to admit the hands.

The cage is almost equally lighted from all sides, and the cheesecloth
at each end permits good circulation. It is very easy of construction
and management and very easily cleaned. In addition, a parasite either
dropping or crawling from the top of the cage almost invariably reaches
the rack of codling-moth cocoons at the bottom. About 15 adult female
parasites can be placed in one cage.

The racks in which the codling-moth larvae were placed for spinning
were of two kinds, depending on the use to which the larvae were to be
put. For ordinary propagation the common corrugated strawboard used
in packing glassware was used. This was cut across the corrugations
into strips about three inches long and five-eighths of an inch in width.
This gives comfortable quarters in each cell for a single worm. These
were placed on edge in small wooden boxes 3 inches long hy 2% inches
wide and three-fourths of an inch deep. Worms placed on the racks
crawled almost immediately into the cells and shortly spun up. One
box at a time was placed in a cage with the adult Calliephialtes for
parasitization.

For the detailed study of the life history of the parasite double slides
of transparent celluloid were constructed. The celluloid was cut into
strips three inches by five-eighths of an inch. These were held apart and
the space between divided into seven cells of the proper size by small
slips of cardboard one-tenth of an inch thick and held in place by being
fastened with shellac to one of the celluloid strips. The whole was held
together by small gummed labels pasted over the ends. Each cell was
numbered on the cardboard slip preceding it. Each slide was also given
a number, and the slides used in each experiment were grouped under a
Roman numeral. In this way notes on the contents of any given cell
could be definitely associated with the subject without any chance of
confusion. With this device it was only rarely that accurate observa¬
tions on the development and activities of the insects within the cells
could not be readily made by transmitted light.

2l6

Journal of Agricultural Research

Vol. I, No. 3

When not under observation, each slide was placed in a folder of dark
paper which left only one edge exposed, and was filed with others of the
same experiment in a shallow box constructed for the purpose.

Observations were as a rule made twice daily, in the early morning
and in the late afternoon, the intervening time being considered, for the
purposes of the notes made, as half a day.

It was found that a living worm within its cocoon would respond
immediately to the stimulus if a needle was thrust through the bottom
of the cocoon. This aided materially in the determination of the time
at which oviposition of the parasite took place, since, with but one excep¬
tion, the parasite was never known to deposit an egg without first killing
the host larva.

The food supplied the parasites consisted of sweet liquids, such as
sugar solution, dilute molasses, and strained honey. All of these sub¬
stances were lapped up greedily by the parasites of both sexes.

REPRODUCTION

THE EXTERN AT SEXUAE APPARATUS

Ovipositor. — The ovipositor (figs, i, 2, and 3) is composed of five

long slender pieces. The two outer ones are black and hairy, grooved

longitudinally within, and form a tube or

sheath surrounding the ovipositor proper.

Next inside of this is a smooth chitinized

piece, deeply grooved on the ventral side and

terminating in a prowlike point. At its base

it is forked, indicating that it is formed of

two opposed pieces fused along their dorsal

edges. Within this is a pair of very slender

flattened pieces barbed at their tips.

The outside pair together form the sheath.

This has no part in the act of oviposition,

but is merely a protection for the ovipositor

proper, which is composed of the three other

pieces. The single piece may be called the

“lance,” since it is with this that the host

larva is pierced. The inner pair have been

variously termed “lancets,” “stylets,” etc.

In oviposition the egg passes down the channel fig. 2.—Caihephiaiies <sP.: lateral

formed by the three parts of the ovipositor view of terminal abdominal seg¬
ments, showing relative position

proper. of elements of ovipositor, a, Valves

On each side and slightly above the base of sheath: b> lance= lancets; d,

cerci.

of the sheath is a small tuberclelike appendage

bearing a number of long, stiff hairs. These are the cerci.

GeniTaeia of MaeE. — The male external sexual organs (figs. 4 and 5)
consist of two sets of paired pieces and the penis. The outer pair are

Fig. 1— Collie phialtes sp.: Ventral
view of terminal 'abdominal seg¬
ments, showing relative position of
elements of ovipositor, a, Valves
of sheath; b , lance; c, lancets; d,
cerci.

Dec. 10, 1913

Calliephialtes Parasite of Codling Moth

217

Fig. 3. — Calliephialtes sp.: Lateral
view of tips of elements of oviposi¬
tor. a. Sheath; b, lance; c, lancet.

broad, tapering toward the tip, concave within, and, except during
copulation, fit together like the two valves of a mussel shell, forming a
sheath inclosing the other organs. They are
homologous with the parts of the ovipositor
sheath, and, like those, probably have no other
function than that of protection for the more
essential organs. The penis is probably homol¬
ogous with the lance of the ovipositor, since
its position in relation
to the other organs cor¬
responds to that of the

lance in relation to the other portions of the
ovipositor. It is a fleshy, flattened organ, termi¬
nating ventrally in two lobes contiguous at their
apices. Immediately in front of these on the
ventral side is an opening leading into the cavity
of the organ. Immediately below the penis and
on each side is a 2 -jointed appendage correspond¬
ing to the lancets of the ovipositor. The basal
joint of this organ is thick and muscular and on
the dorsolateral side is prolonged into a blunt pro¬
jection bearing at its tip a number of
stiff hairs. It is probably a tactile
organ, and may be called the genital
palpus. The second joint is a large
blunt tooth which curves laterad. It probably serves the
double purpose of clasper and dilator. The genitalia, as
described above, are surrounded at the base by a more
or less cup-shaped chitinized piece, the cardo.

COPULATION

Fig. 4. — Calliephialtes sp.:
Ventral view of male geni¬
talia. a, Sheath; b, penis;
c, clasper; d, genital palpus;
e, cardo.

Copulation occurs shortly after the emergence of the
female and may evidently be repeated. The attraction
between the sexes seems to be rather weak and is somewhat
stronger in the female than in the male, as evidenced by the
excited movement of the antennae and wings in that sex
on the approach of the male. The male apparently must
be within about an inch of the female before he becomes
conscious of her proximity. Of courtship there is none,
the male simply jumping to the back of the female as soon as he per¬
ceives her. If she is not ready for his attentions a lively encounter
ensues, the female using her hind legs and wings in freeing herself from
the male. The act of copulation is short, no case having been observed
in which the sexes were together more than five minutes. In copulation

Fig. 5. — Calhe-
phialtes sp.:
Ventral view
of clasping
organ of male
genitalia. af
Basal portion;
b, clasper; c,
genital pal¬
pus.

218

Journal of Agricultural Research

Vol. I, No. 3

the tip of the abdomen of the male is curved down at one side of the
abdomen of the female while he clings to her wings and body.

oviposition

Oviposition began in the cages about nine days after the emergence of
the female. The stage of the host selected is the full-grown larva in its
cocoon. In no case was any other stage attacked.

The act of oviposition (PL XX, figs. 2 and 3) was observed many times.
The insect first explores the surface of the cocoon carefully with her
antennae. Then standing “on tiptoe” directly over the cocoon she
raises the abdomen to a perpendicular position, at the same time lowering
the ovipositor. Sometimes the ovipositor is lowered the entire distance
free from the sheath, the latter remaining in line with the abdomen;
but more frequently it is not released until it is at or below the horizontal,
in which case the sheath bends downward, only the tip clasping the
ovipositor. The sheath finally snaps back into position in line with the
abdomen.

When the lowering of the ovipositor is completed it lies along the
ventral surface of the abdomen and extends down between the legs, while
the tip of the abdomen is bent downward over the base of the ovipositor.
The tip of the ovipositor, guided by the antennae, is placed against the
surface of the cocoon. The antennae are then extended in front of the
head and almost parallel with the surface on which the insect is stand¬
ing. The insect is now exactly analogous to a machine drill, the body
and legs representing the machine and the ovipositor the drill. The
bent-over tip of the abdomen is pressed against the base of the ovipositor,
which bends forward against the ventral surface of the abdomen. With
a more or less augurlike motion the ovipositor is forced through the
cocoon. A few rapid jabs stir up the prospective host larva and it
begins a desperate attack upon the ovipositor of its enemy, biting it and
sometimes holding on with bulldog tenacity. In a number of cases the
defense of the larva was so determined and powerful that the parasite
was defeated and left the field minus a portion of her ovipositor, which
had been bitten off by the larva. Usually, however, the parasite is suc¬
cessful in her efforts and finally thrusts her ovipositor into the larva, sting¬
ing it into insensibility. The stinging is usually repeated one or more
times after intervals of rest. The subjugation of the host accomplished,
the ovisitor is withdrawn from the host and thrust its entire length into
the cocoon; then the parasite rests quietly for several minutes. In this
position the abdomen is bent downward so that the tip is close to the
base. The ovipositor sheath during all this time has retained its vertical
position and is now in contact with the dorsal surface of the abdomen for
about one-third of its length. In a few moments there begins a pulsation

Dec, io, 1913

CalliephiaMes Parasite of Codling Moth

219

of the membranous portion of the venter at the base of the ovipositor,
at which time the egg is being forced into the ovipositor. The egg slips
rathfcr quickly down the ovipositor, becoming visible at a point just inside
the cocoon and remaining visible during the remainder of its passage.
It leaves the ovipositor, caudal pole first, at a point about 1 millimeter
from the end on the ventral surface. It is placed at almost any point in
the cocoon, not necessarily on the host larva.

Her egg having been deposited, the parasite usually gives a parting thrust
or two and withdraws the ovipositor, which springs back into its sheath.

The duration of the act of oviposition is very variable, depending on
the length of time required to locate and kill the larva. The shortest
time observed was n minutes and the longest fully 45 minutes. The
essential portions of the operation, however, probably do not require
more than 4 or 5 minutes in the aggregate.

Only one egg is deposited at a time, and normally only one parasite
develops on a single host. However, in a considerable number of
instances superparasitism took place, and in a few cases under observa¬
tion two parasites developed on a single codling-moth larva. This tend¬
ency was undoubtedly encouraged by the confinement of the cages, and
as many as seven eggs were deposited in one cocoon.

No data were kept on the exact number of eggs deposited by individual
parasites nor on the number deposited daily by individuals, since in each
of the life-history cages from five to nine females were used. But the
results in these cages indicate that the total individual oviposition was
in the neighborhood of 75 eggs and the average daily oviposition about
2 eggs.

THE EGG

The egg (fig. 6) is opaque white, smooth, 1.5 mm. long, and about one-
fifth as wide at the widest part. It is rounded at the cephalic end and
tapers to a long point at the caudal end; in one plane
it is considerably curved. The surface is without

sculpture. Fig. b.—Calliepkialtes

As the embryo develops, it draws away from the sp" Bgg*
poles, and the chorion appears transparent and shriveled. Hatching
takes place through a slit on one side near the cephalic pole, the larva
freeing itself by a series of contortions which finally throw off the egg¬
shell, which is very tough and persistent.

The incubation period for 825 eggs was determined. It varied from
one to seven days, depending on weather conditions. Table I shows
the incubation periods by months, the number of eggs hatching in each
period, and the weighted average mean temperature for each period and
for the season.

220

Journal of Agricultural Research

Vol. I, No. 3

Table? I. Incubation periods of eggs of Calliephialtes sp . and the relation between incu¬
bation period and temperature at Vienna , Va., IQI2 .

Incubation period.

Number of eggs hatching in —

Total.

Average
mean tem¬
perature.

Apr.

May.

June.

July.

Aug.

Sept.

Oct.

1 day .

1.5 days .

2 days .

2. 5 days .

3 days .

3.5 days .

169

250

255

°F.

78. 0

74.4

70. 2

67. 0

62. 7

58. 7
58.9

55-3

55-6

57-2

54- 8

53-2

4 days .

4.5 days .

5 days. .

5.5 days .

6 days .

6.5 days .

7 days .

Total .

Average .

5- 43

330

2. 15.

1.74

171

1. 60

2. 64

109

3. 18

825

2. 14

69. 96

The relation of incubation period to the average mean temperature
based on the figures of Table I is shown in graphic form in figure 7. Ref¬
erence to this dia¬
gram will show that
with a fair degree of
constancy the dura¬
tion of the incuba¬
tion period varied
inversely as the aver¬
age mean tempera¬
ture. The tempera¬
tures that are far¬
thest from the curve
(those of 3.5, 4.5,
and 5.5 days) are
based on the incuba¬
tion periods of few
eggs (16, 12, and 4,
respectively) , and
the possibility of
error was therefore
greater than had the
number been larger.

Eig. 7* — Diagram showing relation between incubation period of eggs of
Calliephialtes sp. and average mean temperature at Vienna, Va., 1912.

THE TARVA

larva (fig. 8) is
across the head.

The newly hatched
yellowish, slightly shorter than the egg, and widest
The head is distinctly separated from the rest of the

Dec. io, 1913

Calliephialtes Parasite of Codling Moth

2 21

Fig. 8. — Calliephialtes sp.: Dorsal view
of newly hatched larva.

body. The body is about three and one-half times as long as the head
and is composed of 13 segments, tapering in size toward the caudal end.
The head of the newly hatched larva is shown in ventral view in figure 9.

The form of the larva changes after the first molt to thick spindle
shape; it is curved dorso-ventrally and is
without a definite head. When full grown
(fig. 10), it varies much in size, depending
on the condition and abundance of food.

Normally it is about three-eighths of an
inch long and slightly less than a third as thick in its greatest diameter.
It is pinkish white in color, the body contents showing through the
transparent skin, while the adipose tissue appears as opaque-white
granules. Larvae that later develop into females average somewhat
larger than those that develop into males. The face
of the full-grown larva is shown, much enlarged, in
figure 10, b.

The larva begins feeding very shortly after hatching
and may attack its host at almost any point, although
it is more likely to attack the dorsum or sides than
the venter. As feeding continues, it may change its
position occasionally. In most cases the point of
attack is finally shifted to a point near the posterior
end of the host, the parasite pushing the collapsing
skin up toward the head until there is nothing left of
the host but a pellet consisting of skin and head shield. This is finally
pushed to one end of the cocoon.

Calliephialtes is normally a solitary parasite, but as indicated in the
foregoing discussion of oviposition,
more than one egg was deposited on
numerous occasions on a single host ;
though on only a few occasions did
more than one live beyond the first
stage. Usually the extra eggs did
not hatch, owing probably to their
being destroyed by the first larva to
hatch. The actual destruction of eggs
in this way was observed on a few
occasions. However, in a very few
instances, two larvae developed on a
single host. In such cases neither of
the larvae attained normal size and
all produced dwarf adults. In only one instance of double parasitism was
an adult female produced, and then the other individual was a male.

As a rule, the cocoon was started very shortly after the larva finished
feeding, and for the purpose of this paper the beginning of the cocoon
is taken as the end of the feeding period. However, in a considerable

head of newly hatched
larva.

Fig. io. — Calltephialtes sp.: o, Full-grown larva;
6, face.

222

Journal of Agricultural Research

Vol. I, No. $

number of cases some time elapsed after the larva had finished feeding
before it began its cocoon, and in a few instances in which the insect was
reared to maturity no cocoon was made. But such cases as that last
mentioned resulted in diminutive adults. In most cases in which the
cocoon making was delayed the supply of food had been small.

The feeding period — determined, as indicated above, from the hatching
of the egg to the beginning of the cocoon — varied, in a total of 579 cases
observed, from 3^ to 1 8% days, with an average of about 7X days.
In Table II all of the larvae carried through to the spinning of the cocoon
are recorded, the months in which they spun their cocoons and their
feeding periods being indicated. The weighted average feeding periods
for each month and for the entire period are also shown. This is un¬
doubtedly higher in each case than the normal average, because, while
the conditions of nature were imitated so far as possible in the cages,
abnormal influences affected some of the larvae so that not only was their
feeding period protracted, but some time passed after they had finished
feeding before they started their cocoons. However, it is impossible to
tell at what point to begin eliminating such larvae from the averages,
so all are included.

Table II. — Actual and weighted average feeding periods of larvce of Calliephialtes sp.for
the period from May to October and the average for the season at Vienna Va. , IQ12.

Feeding period.

Number of larvae in

May.

June.

July.

Aug.

Sept.

J Oct.

Total.

3*5 days .

4 days .

4*5 days .

c days .

c.c days .

6 days .

6. c days .

7 davs .

7.5 days .

8 days .

8. < days .

. 1

o days .

2 !

0. < days .

10 days . .

10. 5 days .

11 days .

1 1 . 5 days .

""2"1

12 days .

1 j

12. c days .

13 days .

. 1

12.C days .

I !

14 days .

14.5 days .

15 days .

15.5 days .

16 days .

16. 5 days .

18.5 days .

Total .

143

6.8s

157

j 5*54

1 — 4. -

5-*i |

8. 42

i1* 53

579

7- °7

Average feeding
period (days) . . .

6. 98

Dec. io, 1913

Calliephialtes Parasite of Codling Moth

223

A considerable portion of the larval life of Calliephialtes is passed in
the cocoon. This period was determined for 116 female larvae and 404
male larvae. The females, after spinning their cocoons, required, on the
average, about 2)4 days longer to attain the pupal stage than did the
males. This is probably somewhat less than the difference that would
exist under natural conditions, inasmuch as the males under observa¬
tion were somewhat more inclined to extend this portion of their de¬
velopment beyond the normal than were the females.

In Table III are brought together the data on that portion of the
larval life passed within the cocoon. The figures include the prepupal
period, which, not being a definite stage in the development of the insect
but a transition stage, it is impossible to determine exactly. From
this table are eliminated the data on 8 females and 11 males that re¬
mained in this condition for an abnormally long time. The actual
maximum period recorded for females was 24 days and for males 3 6)4
days.

Table III. — Larval period of both sexes of Calliephialtes sp. in cocoon in various months ,
weighted average period for each month and for the season , and weighted average mean
temperature for each period and for the season at Vienna, Va., IQI2.

Average
mean tem¬
perature
for period.

7.
78. o
78. 2

76. s

74.8
74.2

72.7

71. O

72.4

71.7

67.7
69. 6

68.5
71. 1

70.9
72.4

72. 1

Larval period in
cocoon.

4 days. . .

4.5 days.

5 days . . .

5.5 days.

6 days. . .

6.5 days.

7 days . . .
7.< days. .

8 days . . .

8.5 days.

9 days . . .

9.5 days,
xo days. .

10.5 days.

1 1 days . .

11. 5 day's.

12 days. .

12.5 days.

13 days. .

13.5 days.

14 days . .

May

and

June.

Total. . . .
Weighted av¬
erage pe¬
riod, days. .
Average tem¬
perature,

O T>

Females: Number of
larvae pupating in-

IO. 2

July.

Aug.

Sept.

8.0

9- 7

10. 4

Total
num¬
ber
of fe¬
males.

2
4

108

9.9

Average
mean tem¬
perature
for period.

May

and

June.

a F.

77.8
77-7

72- 3
74-4
72. 2

73- 9

70.9
69. 6

68.4
67- 5
68.8
69. o
69. 1

67. 7
66.8
67. 6

70. 1

Males: Number of
larvae pupating in —

i3S

7-9

July.

2
1
1

3
1

125

6.6

Aug.

2
24

13
20
II

14
5

2
2

108

7.2

Sept.

Total
num¬
ber of
males.

4
7
7

25 393

8.7

7-4

224

Journal of Agricultural Research

Vol. I, No. 3

The figures of Table III are expressed in graphic form in the diagram
(fig. n), which shows the relation between temperature and the larval
period, in the cocoon. From the curve for males it is evident that indi¬
viduals which took more than 8l/i days between the spinning of the cocoon
and pupation were more largely influenced by external conditions other
than temperature than were those that required less time. The same is

true of the females
after io}4 days,
although these
showed the effect
to a less marked
degree than did the
males.

It will be seen
from the figures
given for the feed¬
ing period and the
larval period in the
cocoon that the
minimum and max¬
imum possible to¬
tal larval periods

would be for females 9.5 and 42.5 days, respectively, and for males 7.5
and 55 days. The actual minimums and maximums were for females 12
and 27 days, respectively, and for males 7.5 and 51 days.

In Table IV are summarized the data obtained on the total larval
period, with the exception of those on 13 females and 16 males in
which this portion of the life cycle was unduly protracted. The total
number for which the duration of this period was determined was 99
females and 344 males. The females required, in the average, nearly
three days more to complete their larval life than did the males.

Fig. 11. — Diagram showing relation between temperature and larval period
of males and females of Calliephialtes sp. in the cocoon at Vienna, Va. ,1912.

Tabus IV. — Summary of data on total larval period of Calliephialtes sp. at Vienna, Va.,

IQ12 .

! Total larval
| period.

Females: Number of lar¬
vae pupating in —

Total

Males: Number pupat¬
ing in —

Total

May

and

June.

July.

Au¬

gust.

Sep¬

tem¬

ber.

number
of females.

May

and

June.

July.

Au¬

gust.

Sep¬

tem¬

ber.

number
of males.

7.c days .

8. < days .

9 days .

0. c days .

10 days .

10.5 days _

11 days .

11,5 days _

Dec. 10, 1913

Calliephialtes Parasite of Codling Moth

225

Table IV. — Summary of data on total larval period of Calliephialtes.; at Vienna , Va. ,

IQ12 — Continued.

Total larval
period.

Females: Number of lar
vse pupating ih —

Total

Males: Number pupat¬
ing in —

Total

May

and

June.

July.

Au¬

gust.

Sep¬

tem¬

ber.

number
of females.

May

and

June.

July.

Au¬

gust.

Sep¬

tem¬

ber.

number
of males.

12 days .

12,5 days _

13 davs .

13.5 days....
14 days .

14.5 days....
it days .

15.5 days....
16 days .

16.5 days....

17 days .

17.5 days....
t A days

18.5 days....

m davs

....

19.5 days....

or\ davs

Total .

Average pe-

129

i°5

328

riod, days.

16.5

13-3

15-4

j I5-9

16. 0

I4-3

11.9

12.8

I3.6

13.2

THE PREPUPA

A few days before pupation the larva begins to show the constriction
between the thorax and the abdomen, the eyes become discernible as
distinct red spots, and before
pupation actually takes place
the appendages can be indis¬
tinctly seen through the delicate
larval skin. The antennse are
coiled under the head instead of
being extended along the venter,
as in the pupa. In the prepupal
stage (fig. 12) the sex of the insect can with certainty be determined
for the first time. In *he female prepupa the tip of the abdomen is bent
slightly backward, indicating the developing ovipositor, while in the
male the caudal segment is straight.

THE PUPA

When pupation takes place, the larval skin splits along the median
dorsal line over the top of the head and for a short distance down the
back, and through this opening the pupa makes its exit. Figure 13
shows the beginning of pupation of a female Calliephialtes. The rent in
the exuvium, through which the antennae are shown to extend, was

226

Journal of Agricultural Research

Vol. I, No. 3

Pig. 13. — Calliephialtes sp.: Beginning of exuviation of female
pupa.

probably caused accidentally in the preparation of the specimen. By
a series of twisting contortions the exuvium is gradually worked back¬
ward to the tip of the abdomen, where it is thrown off. It is very delicate
and transparent, but as it is pushed back and becomes wrinkled it
gradually appears darker until, when it is entirely shed, it is light grayish
brown and is a mere shred.

In the male this is the end of the act of pupation, but it leaves the
female with the ovipositor only a small fraction of its ultimate length
and very thick.

The extension of the ovipositor is accompanied by a series of rythmical
movements, about seven to the minute, during which the organ is repeatedly

pressed against the dorsum of
the abdomen. Whether the
pressure thus exerted is the
cause of the lengthening of
the ovipositor or the effect of
pressure from within the
body and merely incidental
could not be determined.

The act of exuviation required about 15 minutes, but where the exten¬
sion of the ovipositor was observed and timed the extension consumed
from 35 to 41 minutes. The pupation of the male
therefore required about 15 minutes, while the
female required from 50 to 56 minutes to complete
the process.

The newly formed pupa is entirely white, with
the exception of the eyes, which are red. The legs
and antennae lie fully extended along the sides and
venter, and in the female the ovipositor lies along
the dorsum, extending the whole length of the
body and curving somewhat at its tip over the
head.

Gradually the eyes darken, becoming very dark
before the adult color begins to appear over the rest
of the body. The head and thorax are the next to
begin to assume color, then the dorsal and ventral
plates of the abdomen, the antennae, the legs, and
finally the ovipositor. When the coloring is com¬
plete (see fig. 14), the head, thorax, and antennae are
black, the eyes dark reddish brown, the wing pads
gray, the chitinized portions of the abdomen and
ovipositor nearly black, the legs yellowish, and the unchitinized portions
white.

The pupal periods of 109 females and 366 males were determined. The
average female spent 1.66 days longer in this stage than did the average
male. This difference would, however, probably be somewhat greater

Fig. 14. — Calliephialles sp.:
Pupa of female and tip of
abdomen of male pupa.

Dec. io, 1913

Calliephialtes Parasite of Codling Moth

227

under natural conditions, as the males under observation were consider¬
ably more likely to extend this period beyond the normal than were the
females. The actual difference is probably more closely indicated by
the shortest pupal period for each sex, which gives a difference of two
days.

In Table V the data on the pupal period are summarized and the aver¬
age mean temperature for the various periods given.

TablK V. — Summary of data on duration of pupal period of Calliephialtes sp. and aver¬
age mean temperature at Vienna, Va., 1912,

Pupal period.

Females: Number
transforming in —

Total

number

Average

mean

tempera¬

ture.

Males: Number
transforming in —

Total
num¬
ber of
males.

Aver¬

age

mean

tem¬

pera¬

ture.

1— i

of fe¬
males.

3
*— >

4-!

6 days .

°F.

. . . .

°F.

77. 2

79. 0

78. 2

74. 7

74.4

73*4

68.4

68.3
68.0
<55*3
66.9 j

66.4

6.C days .

. . . .

7 days .

....

7.5 days .

17
22

8 days .

8.5 days .

9 days .

9-5 days .

iodays .

10.5 days. . . .

11 days .

11. 5 days ....

4
18
18
13
17
15

78.7

72.4

76.9
72. 6

71.9
71. 6

69.9
68.6

68.9
67*5
66. 7

12 days .

12. *; days. . . .

12 days*

Total ....

Average pu¬
pal period,

days .

Average tem¬
perature ,°F

109

i37

107

366

11. 50

9. 90

9- 57

9- 73

10. 78

70.9

9.94

7-83

9. 19

8. 36

9. 12

7°- 3

The August column for males in Table V includes the data on 49
pupae which were reared from unfertilized eggs. Whether the parthenoge-
netic character of these eggs had any effect in lengthening the pupal
period is a question, but a comparison of the pupal periods of these with
those of the 34 males that were developing at the same time from ferti¬
lized eggs shows that the pupae from parthenogenetic eggs required a
somewhat longer time. This is shown in Table VI. If these 49 indi¬
viduals were eliminated in Table V, the total and average in the August
column would be 58 and 8.61, respectively, and the grand total and grand
average would be 317 and 9.0, respectively.

17072 13 - 4

228

Journal of Agricultural Research

Vol. X, No. 3

Table) VI. — Relative length of pupal stage of males of Calliephialtes sp. from fertilized
eggs and those from partheno genetic eggs at Vienna , Va., IQI2.

Number of pupae from —

Pupal period.

Fertilized

Parthenoge-

eggs.

netic eggs.

7.5; days .

8 days .

8.5 days .

9 days .

9-5 days .

10 days .

10.5 days .

11 days .

Total .

Average pupal period, days .

9. 44

9.87

The truth of the relation between the pupal period and tempera¬
ture is in all probability not nearly so closely shown by the figures

as is that between
the incubation period
and temperature by
the figures in Table
I, since the impor¬
tant factor of amount
and condition of
food has had oppor¬
tunity to have its
full effect. This fac¬
tor, next to temper¬
ature, is probably
the most important
single factor influ¬
encing the duration
of the stage, espe¬
cially under the unnatural condition of the breeding cage.

The figures of Table V are expressed in graphic form in figure 15.

THE COCOON

As stated, the larva usually begins its cocoon shortly after having
finished feeding. Before starting its spinning it pushes the remains of
its host to one end of the host cocoon and then accommodates its own
cocoon to the size and shape of the space remaining within that of the
host. The parasite cocoon therefore varies considerably in shape. It is
usually, however, about one-half inch in length, about a third as broad.

PUPAL PEP/OD - DArS

Pig. 15.— Diagram showing relation between pupal period of Calliepht-
altes sp. and temperature. The dot-and-dash line is the curve of
average temperature, while the dotted line represents the female
curve superimposed on that ot the males at Vienna, Va,, 1912. The
greater tendency of the males to delay transformation to the adult
stage is shown by reierring the male and female to the line of average
temperatures.

Dec. iot 1913

Calliephialtes Parasite of Codling Moth

229

and of the depth of the host cocoon. The upper and lower sides and
the end next the remains of the host, most frequently the cephalic end,
are flattened, while the edges and the other end are more rounded.

The cocoon is of a pale pinkish brown color, semitransparent, and
is composed of a thin tissuelike material containing but few threads.

It was found very easy to observe by transmitted light the devel¬
opment of the parasite in its cocoon.

The period in the cocoon includes a part of the larval life, all of the
pupal period, and a small portion of the adult life. By adding the
minimums and maximums for each of these phases of development the
total possible minimum period would be for females 14.5 days and for
males 10.5 days and the total possible maximum period for females 39
days and for males 50 days. The actual minimum and maximum for
females were 15.5 and 37.5 days, respectively, and for males 11.5 and
36 days.

The duration of this period was determined for hi females and 396
males. The weighted average duration of this portion of the life his¬
tory indicates that the females remain in the cocoon about four days
longer than do the males. Table VII summarizes the data obtained.
Eight females and five males, the recorded periods of which were far
in excess of the normal for the month in which they emerged, are
omitted from the table.

Table VII. — Summary of period spent by Calliephialtes sp. in cocoon , showing number
for each period by months , total for each period and each month, and average period for
each month and for the season at Vienna, Va., IQ12.

Period in cocoon.

Number of females emerging
in —

Total
num¬
ber of
fe¬
males.

Number of males emerging
an —

Total
num¬
ber of
males.

June.

July.

Aug.

Sept.

June.

July.

Aug.

Sept.

11. 5 days ....

12 days. . . .

12.5 days ....

13 days .

1 1

13.5 days. . . .

14 days .

14.5 days ....

15 days .

2 1

ic.c days ....

16 days .

16. *; days ....

1 1

3°

17 days .

17.5 days. . . .

18 days .

17 |

18.5 days. . . .

19 days .

3°

iq. c days ....

■2

20 days .

T-

20. 5 days ....

4 j

21 days .

' 1

8 !

230

Journal of Agricultural Research

Vol. I, No. 3

Table; VII. — Summary of period spent by Calliephialtes sp. in cocoon, showing number
for each period by months , total for each period and each month , and average period for
each month and for the season at Vienna, Va 1912 — Continued.

Period in cocoon.

Number of females emerging
in —

Total
num¬
ber of

Number of males emerging
in —

Total
num¬
ber of
males.

June.

July.

Aug.

Sept.

fe¬

males.

June.

July.

Aug.

Sept.

21.5 days ....
22 days .

22. < days ....

23 days .

23-5 dVs- • • •
24 days .

24.5 days -

2 c days .

-2

26 days .

::::::

26.5 days. . . .
27 days. .....

Total num¬
ber .

19.4

20. 0

20. 5

103

21- 5

133

I9.4

*5-3

130

17. 2

17. 1

39r

17. 7

Average pe¬
riod, days. .

22. 6

THE ADULT

Transformation from the pupa to the adult within the cocoon takes
place one or two days before the emergence of the adult, depending
largely on the difficulty encountered by the insect in biting its way by
the remains of the host and through the two cocoons. The female
effects her escape in a somewhat shorter time than the male.

In the spring the males appear some time ahead of the females, as
indicated by the emergence of unforced material in the spring of 1912.
From this material the first males appeared on April 23 and the first
females 10 days later. In fact all but a few belated males appeared
before the first female.

The males far outnumbered the females throughout the period covered
by the observations, and it was found that the proportion of males
increased with each succeeding brood. It appears that the effect of the
unavoidably unnatural conditions of the artificial propagation tended to
the production of males and that this effect was cumulative. Of the
528 individuals reared from mated females in the regular life-history
experiments 396, or exactly three-fourths, were males. Table VIII
summarizes the data on this point.

Dec. 10, 1913

CalliephiaUes Parasite of Codling Moth

231

Table VIII. — Proportion of sexes of CalliephiaUes sp. from bisexual reproduction at

Viennat Va., IQI2.

Brood.

Number

females.

Number

males.

Ratio of
females
to males.

Hibernating .

i : 2. 48

First .

153

1: 1. 87

Second .

1 12

1:5. 60

Third .

1:8.79

Total .

132

396

1:3.00

Of the 57 individuals reared from parthenogenetic eggs all were males.

No definite data were obtained on the longevity of the females, for the
reason that it was necessary to use all such in propagation experiments,
and the individuals could not be distinguished. Some information on
this point can, however, be obtained from the notes on the propagation
cages. All females were fed, and hence there are no data on longevity
without food.

Of the unforced hibernating females the first emerged on May 3 and the
last on May 13. The latter was a weak individual and lived only 10
days. The last to emerge previous to it appeared on May 7. The
earliest death, with the exception mentioned above, occurred on June 4
* and the last on June 22. This gives a maximum longevity of 50 days, a
minimum of 22 days, and an average of 36 days.

Females emerging from June 13 to 17 died from July 9 to August 7.
The maximum longevity was 55 days, the minimum 22 days, and average

38.5 days.

Females emerging from June 24 to 26 died from July 19 to August 9.
The maximum longevity was 46 days, the minimum 23 days, and the
average 34.5.

Females emerging from June 27 to July 1 died from July 9 to 30. The
maximum longevity was 33 days, the minimum 8 days, and the average

20.5 days.

The females surviving on August 9 in all first-generation cages were
assembled in one cage on that date. Of these, 4 were from a lot emerging
from June 18 to 20, an average of 51 days previous to the transfer; 3 from
a lot emerging from June 22 to 23, an average of 47.5 days previously; 3
emerging from July 3 to 10, 33.5 days previously. The 10 females, after
being placed together, died August 13 to 19, an average of 7 days later.
The average longevity of the females from the earliest of the three lots
was therefore 58 days, of those from the second lot 54.5 days, and of
those from the third lot 40.5 days.

The average longevity of all females listed above was 51 days.

232

Journal of Agricultural Research

Vol. I, No. 3

A number of surplus males emerging from June 14 to 22 were used in
an experiment to determine the longevity with and without food. Of
the 51 males used in the experiment 22 were fed and 29 unfed. For the
fed males the maximum longevity was 51.5 days, the minimum 8.5 days,
and the weighted average 32.5 days. The longest lived unfed male lived
10 days, the shortest lived 3 days, and the average lived 5.4 days. The
average fed male therefore lived almost exactly six times as long as the
average unfed male.

The adult Calliephialtes were very easily handled on account of their
great docility. . On many occasions while photographing the females in
the act of oviposition the writer has carried a transparent slide on which
a female was perched from the insectary to a greenhouse 20 feet distant,
set it up in front of the camera, and made one or more exposures without
the insect withdrawing her ovipositor; and in no case was the insect
sufficiently disturbed to cause her to fly away.

The adults fed greedily at all times on the sweet liquids supplied them,
and the males confined their feeding to this sort of diet. But the females
very frequently fed on the juices of the codling-moth larvae. This food
they secured by repeatedly jabbing with their ovipositors the larvae in
the cocoons and licking up the juices that saturated the cocoon. Fre¬
quently a half or more of a larva would be consumed in this way, the
parasite continuing to feed for an hour or more, alternately pumping the
juices of the larvae out with her ovipositor and licking them up. On one #
occasion a female Calliephialtes was observed to have killed and partially
eaten a larva that had left its cocoon and was at large in the cage.

The total developmental period from oviposition to emergence was
determined for 112 females and 399 males. For females it ranged from
23.5 days to 44.5 days and for males from 18 to 44 days. Both of the
maximums as well as a considerable number of other records are based
on individuals which, for some cause — usually inadequate food supply —
were unable to go through their development in as short a time as they
would have done under normal conditions. The records for 12 such
females and 22 males are omitted from Table IX, which summarizes the
data on the 100 other females and 377 other males. This table indicates
that the average female required about 5 days longer to complete devel¬
opment than did the average male, the shortest period for females being
SH days longer than the shortest for males.

Dec. io, 1913

Calliephialtes Parasite of Codling Moth

233

TablB IX. — Total developmental period of Calliephialtes sp.; summary of duration of
period by months , sexes, and for the season at Vienna , Va., IQI2.

Total develop¬
mental period.

Number of females emerging
in —

Total
number
of fe¬
males.

Number of males emerging
in —

Total

number

males.

June.

July.

Aug.

Sept.

June.

July.

Aug.

Sept.

Days.

18 .

18. c .

IQ . . . , #

7 .

io.e .

20 . . . .

20. < .

21 .

21. K .

22 .

22. c.. .

27 .

27. C . . * , .

24 .

24. z .

2 C .

2 c . c; .

26 .

26. K .

27 .

27#C .

28.“. .

2S.Z .

20 .

2Q. C .

70 .

70. Z . . .

'll . .

71. K . * . . . .

72 .

72. < .

OO • . .

77. C. . . .

OO 0

7A .

7Z .

Od .

K.C .

OO O

26 .

Total ....

Average de¬
velopmen¬
tal period,
days .

3i* 1

27.4

27. 1

27. 2

100

29. 6

126

27. 2

21. 7

130

24. 1

23- 5

377

24.7

No definite experiments were conducted in experimental control of the
development, but during the warm weather many strawboard slips of
parasitized larvae were placed in cold storage to retard the development
of the parasites. In so far as it was possible to determine, they were
placed in storage after the spinning of the parasite cocoon. This retar¬
dation of development had no apparent effect on the further development
after removal from cold storage. It did seem, however, to reduce the
activity and vitality of the resulting adults. L. J. Newman (18) records

234

Journal of Agricultural Research

Vol. I, No. 3

the keeping of immature specimens of CalHephialtes messor in cold storage
for a period of 14 months, after which they emerged without having suf¬
fered in the least.

SEASONAL HISTORY

The first females to emerge from hibernation in the spring of 1912
appeared on May 3 and the last on May 15. These were placed with
males in propagation cages. The first egg was deposited on May 13, ten
days after the first emergence.

In order to determine the maximum and minimum number of genera¬
tions in a season, the five earliest and five latest appearing female progeny
of the hibernating brood were used in the life-history cages, a separate
cage being used for each group. The same plan was followed out with
each succeeding generation. From the earliest female progeny three
complete generations were reared, and from the latest group two genera¬
tions were bred. With the hibernating brood this gives a maximum of
four generations in the year and a minimum of three generations. Table
X summarizes the data on the number of generations. It is interesting
to note that the total time consumed by the three generations is only
one day longer than that consumed by the two.

Table X. — Number of generations of CalHephialtes sp. reared at Vienna, Va 1912.

Generation.

Maximum number of
generations.

Minimum number of j
generations. ;

Bate of first
female.

Total cycle.

Date of last
female.

Total cycle.

Days.

Days .

Hibernating .

May 2

Mav

First .

June 13

July 13

Second .

July 18

Sept. 12

Third .

Sept. 3

Total peri¬

od, days . .

123

122

Average cycle,

days .

Development ceased at about 50° F., although oviposition was fre¬
quently carried on actively at that temperature. After the middle of
October very few eggs hatched, although the last eggs of the season were
not deposited until November 1. All but a very few of the larvae that
hatched at this season passed through the feeding stage and constructed
their cocoons.

Calhephialtes sp. hibernates as a full-grown larva in its cocoon. In
this stage it is capable of withstanding a very low temperature. The
mortality among hibernating larvae during the winter of 1911-12 was
very slight, if not nil, in spite of the fact that a temperature of —6°

Dec. zo, 1913

Calliepkialtes Parasite of Codling Moth

235

Fahrenheit was recorded in the insectary. This is an unusually low
record for the locality and indicates that the species would have no
difficulty in acclimating itself were it liberated in the region.

ALTERNATE HOSTS

The female parasites appeared in the spring a few days in advance of
the first adult codling moth, or somewhere about 40 days before they
could, under natural conditions, attack the first brood of larvae of the
codling moth. The hibernating brood of parasites would therefore have
passed the greater portion of their adult life before an abundance of
codling-moth larvae could be found. This would necessitate a very small
first generation of the parasites unless they would attack some other host.

To determine if Calliephialtes would attack other species of insects,
larvae of Enarmonia prunivora Walsh, Euzophera semifuneralis Walk.,
and Gnorimoschema gallaesolidaginis (Riley) were placed in the propa¬
gating cages with actively ovipositing female parasites. The larvae of
the first two species were placed in transparent cells, and those of the
last were allowed to remain in their galls. Only a single Enarmonia
larva was available, and this was parasitized within 2 days, a diminutive
male Calliephialtes emerging from the cocoon 22 days later. This species
is, however, much smaller than the normal full-grown larva of the para¬
site, and it is doubtful if it would serve in the long run as an alternate
host.

Of the two other species of larvae neither was apparently given the
least attention by the parasites, although those of Euzophera were left
in the cage for several weeks.

Codling-moth larvae containing the internally parasitic larvae of As -
cogaster carpocapsae were readily attacked and parasitized by Calli¬
ephialtes. This always resulted in the death of the earlier parasite and
the production of a diminutive adult Calliephialtes.

On one occasion a Calliephialtes larva that had already spun its cocoon
was attacked and killed by an adult of the same species. When the fact
was discovered, a small living larva was feeding on the dead parasite
larva. This parasite larva died without spinning.

LITERATURE CITED

1. GravEnhorst, J. L. C. Ichneumonologia Europsea. v. 3, Vratislaviae, 1829.

f* Ephialtes messor, n.,” p, 232. Original description.

2. Taschenberg, E. L. Die Schlupfwespenfamilie Pimplarise der deutschen Fauna,

mit besonderer Riicksicht auf die Umgegend von Halle. Ztschr. Ges.

Naturw., Bd. 21, p. 245-305.

" Ephialtes messor Gr.,” p. 254. Included in synoptic table of genus and recorded as reared from

the wax moth. {Tinea) Galleria mellonella .

3. Walsh, B. D. Descriptions of North American Hymenoptera. Trans. Acad. Sci.

St. Louis, v. 3, p. 65-166, 1873.

“ Ephialtes pusio , n. sp,,” p. m-112. Original description.

236

Journal of Agricultural Research

Vol. I, No. 3

4. Cresson, E. T. “ Ephialtes comstockii Cresson (n. sp.).” U. S. Comr. Agr. Rpt.,

1879, p. 235, 1880.

Original description. Type reared as parasite of Retinia comstockiana Femald.

5. Ball a Torre, K. W. von. Catalogus Hymenopterorum. v. 3, Eipsiae, 1901-2.

" Ephialtes messor Grav.,” p. 47 s'/* Ephialtes comstocki Cress.," p. 471; t( Ephialtes pusio Walsh.,”
p. 476. Credits Gravenhorst with having recorded Tinea mellonella as host of E. messor Grav.
This record should be accredited to Taschenberg (1863).

6. Cooper, El wood. The codling-moth parasite. 2d Bien. Rpt. Comr. Hort.

Gal., 1905/6. p. 231-235, pi. 10. 1907.

"The codling-moth parasite ( Caltephialtes messer Grav.)," p. *31-235, pi. 10. Short general ac¬
count of introduction into California, together with description, life history, habits, and letters
from fruit growers regarding success of introduction.

7. Eounsbury, C. P. Report of the Government Entomologist [Cape of Good Hope],

1905. 1906.

" Ephialtes messor Gravenhorst," p. 98-99. Mentions introduction into California and expresses
doubt as to probable success.

8. Froggatt, W. W. Codling-moth parasites. Agr. Gaz. N. S. Wales, v. 17, pt. 4,

p. 387“395> APr- 2> *9°6-

"The Spanish parasite ( Ephialtes carbonarius >,” p. 393~394- Mentioned in list of codling-moth
parasites.

9. Eounsbury, C. P. Report of the Government Entomologist [Cape of Good Hope],

1906. 1907.

(Spanish parasite), p. 86. Mentions introduction into California late in 1904. Doubts value.

10. - Report of the Government Entomologist [Cape of Good Hope], 1907.

1908.

" Caltephialtes messer, " p. 55. Records introduction into Cape of Good Hope from California.
Comments on introduction into California and expresses belief that as yet the species has not proved
of any practical value or given evidence that it will.

11. Schreiner, J. T. Zwei neue interessante Parasiten der Apfelmade Carpocapsa

pomonella E. Ztschr. Wiss. Insektenbiol, Bd. 3, Heft 7, p. 217-220, 1 fig.,
Dez. 9, 1907.

" Ephialtes carbonarius Christ.," p. 218. Records rearing from codling-moth larva in Europe.

t2. Quaint ance, A. E. The codling moth or apple worm. U. S. Dept. Agr. Year¬
book, 1907, p. 432-450, 1908.

" Calliephialtes messor Grav.," p. 443. Mentioned in list of parasites of codling moth as having
been introduced into California to prey upon that insect.

13. Froggatt, W. W. Insect pests in foreign lands. Second progress report. Jour.

Dept. Agr. Victoria, v. 5, pt. 12, p. 716-720.

‘Ephialtes carbonarius, ' ' p. 717. States that in visit to California he was unable to find any in¬
stance in which the parasite had been found in any orchard.

14. Eounsbury, C. P. Report of the Government Entomologist [Cape of Good

Hope], 1908. 1909.

'Spanish codling-moth parasite. (Caliephialtes messer)," p. 64. Records experience in rearing
parasite.

1 5 . Theobald , F. V. The insect and other allied pests of orchard , bush , and hothouse

fruits and their prevention and treatment. Wye, 1909.

"The codling moth ichneumon ( Ephialtes carbonarius Zach.)," p. 77-78. Mentions introduction
into California. Brief life history.

16. Froggatt, W. W. Report on Parasitic and Injurious Insects, 1907-8, Sydney,

1909.

" Calliephialtes messor ," p. 5-7. Doubts efficiency of species in California.

Dec. xo, 1913

Calliephialtes Parasite of Codling Moth

237

17. Lounsbury, C. P. Report of the Government Entomologist [Cape of Good Hope],

1909. 1910.

“ Calliephialtes vies set," p. 85-86. Reports further liberations. No hope held out that species
will prove of importance.

18. Newman, L. J. Long-lived parasites. Jour. Dept. Agr. West Aust., v. 18, pt. 4,

p. 297, Apr., 1909.

“ Caliephialtes messer .” Records keeping moth larvae parasitized by this species in cold storage
14 months, after which the parasites emerged, apparently not having suffered from the long cold.

19. Essig, E. O. Injurious and beneficial insects of California. Mo. Bui. Cal. State

Com. Hort., v, 2, no. 1-2, Jan. /Feb., 1913.

Brief description and biologic remarks on Calliephialtes messor Grav., p. 265-266, fig. 264.

DESCRIPTION OF PLATE

Plate; XX. Calliephialtes sp. Fig. i. — Female. Figs. 2 and 3. — Characteristic posi¬
tions assumed by the insect in oviposition. Fig. 4. — Male. Figures 1 and 4 are
enlarged about 2^2 times. Figures 2 and 3 are retouched photographs from life;
enlarged about 3 times.

(23S)

Plate XX

POLYPORUS DRYADEUS, A ROOT PARASITE
ON THE OAK

By W. H. Long,

Forest Pathologist , Investigations in Forest Pathology, Bureau of Plant Industry

Bulliard (1789, 1791) 1 figured and described under the name Boletus
pseudo-igniarius a fungus which most European mycologists believe is
the plant now called Polyporus dryadeus. Apparently the next record
of this fungus is by Persoon (1799), where it is described as Boletus
dryadeus . Again it is described by the same writer in his Synopsis
Fungorum (1801), where Bulliard’s fungus is listed as a synonym. It
is first named Polyporus dryadeus by Fries (1821), who describes the
plant and cites as synonyms the names given by Bulliard and Persoon.
Hussey in Illustrations of British Mycology (1849) gives a fairly good
figure of the sporophore and a most excellent mycological description
of the fungus, with its habitat.

Since 1849, repeated references to this fungus are found in European
mycological literature, but nothing was written concerning the rot
produced by it in the oak until Robert Hartig in his epoch-making
work on the true nature of the rots of woods (1878) described a heart
rot of the oak which he attributed to Polyporus dryadeus . A careful
study of Hartig’s figures and the description of the sporophore which
he found associated with the white heart-rot so accurately described
by him is sufficient to convince anyone who is familiar with the true
P. dryadeus that Hartig’s fungus was not P. dryadeus. It is undoubt¬
edly identical with the heart-rotting fungus known in America as P.
dryophUus and found by Hedgcock (1910 and 1912) to be associated with
a whitish piped rot in oak.

Polyporus dryophUus has one character, a hard, granular, sandstone¬
like core, that is unique and not possessed by any other polypore known
to the writer. The sporophore of this plant, represented by numerous
specimens collected by Hedgcock and the writer in western and south¬
western United States, shows this hard, granular core exactly as figured
and described by Hartig in his article on P. dryadeus. This core extends
back some distance into the tree in oaks; it is usually irregularly cylindrical
while in the tree, but on its emergence from the tree it swells into a
tuberous or spheroid mass and finally occupies the central and rear
part of the sporophore. (PI. XXI, fig. 1.) If the sporophore is formed
from a large branch hole, it is usually of the applanate type, with a
small core, but when the sporophore forms directly on the body of the

1 Bibliographic citations in parentheses in the text of this article refer to ” Literature cited,” p. 248.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(*39)

Vol. I, No. 3
Dec. 10, 1013
G— 6

240

Journal of Agricultural Research

Vol. I, No. 3

tree, as it usually does, the shape is tuberous, ungulate, or even sub-
globular (PI. XXI, figs. 2 and 3), with the bulk of the sporophore com¬
posed of a hard, granular core. This core usually has white mycelial
strands. (PI. XXI, fig. 3.) The sporophore of P. dryophilus , therefore,
has normally three distinct kinds of structures: (1) The hard, granular
core, (2) the fibrous layer which surrounds this core except at the rear,
and (3) the layer of tubes on the lower surface. Specimens are often
found, however, especially from the western part of the United States,
in which this fibrous layer may be entirely absent between the tubes
and the granular core. (PI. XXI, fig. 3.)

The sporophore of Poly poms dryadeus never has this granular core, but
its context is fairly homogeneous and of a fibrous-corky structure. (PI.
XXI, fig. 4.) Another very important difference between the two spe¬
cies is the location of the sporophores on the host tree. In P, dryadeus
the sporophores are always on the exposed roots or on the trunks at or
very close to the ground. The reason for this is explained later in this
article. In P. dryophilus the sporophores are higher on the trunk of the
tree, and in some cases are on the branches.

The rot described and figured by Hartig is identical with the rot pro¬
duced by P. dryophilus , but does not resemble in the least the rot pro¬
duced by the real P. dryadeus . Since Hartig’s time European mycolo¬
gists have followed him in descriptions of the rot wrongly ascribed to P.
dryadeus , but most of them have described the sporophores of the true
P. dryadeus both as to its character and location on the tree — i. e., at the
base of oaks. For instance, Von Tubeuf, in his Disease of Plants (1897),
describes fairly well the sporophore of P. dryadeus , while his photograph
of the rot is that of P. dryophilus . Massee, in his Diseases of Cultivated
Plants and Trees (1910), states that “the largest specimens usually occur
near the ground line, but it also springs from points where branches have
died or been broken off.” The latter statement, so far as can be ascer¬
tained by the writer, is incorrect as to the location of the sporophores of
P. dryadeus , but is correct for P. dryophilus. Massee also quotes Hartig
as to the character of the rot produced.

Polyporus dryophilus is known in Europe under at least three different
names : Polyporus fulvus Fries (PI. XXI, fig. 5) , P. friesii Bresadola, and P.
vulpinus Fries. (Pl.XXI,fig. 6.) According to Eloyd (1913), not only are
P. fulvus Fries and P. friesii Bresadola synonyms for P. dryophilus , but the
P. corruscans of Fries is also the same plant.1 Polyphorus vulpinus is the
name given to the form of P. dryophilus found on species of Populus,
authentic specimens of which were seen by the writer at the New York
Botanical Garden in collections from Finland and Sweden and also from

1 Since this article was written, the writer, through the courtesy of Mr. C. G. Lloyd, has examined the
specimens of Polyporus corruscans and of P. rheades deposited in the Lloyd Herbarium at Cincinnati, Ohio.
Both of these plants as represented in this herbarium are Polyporus dryophilus. the former being the usual
form found on oak and the latter the one occurring on poplar. According to Mr. Lloyd, the type of P.
rheades , found by him in Persoon’s Herbarium, is undoubtedly the plant called “P. vulpinus ” by Pries.

Dec. io, 1913

Polyporus Dryadeus

241

Maine. In the Cryptogamic Herbarium of Harvard University there is
a collection on PopuLus grandidentata Michx. from New Hampshire, while
in the laboratory of forest pathology of the Department of Agriculture at
Washington, D. C., there is a fine collection on Populus tremuloides Michx.
from near Steamboat Springs, Colo. (Hedgcock, 1913).

This fungus on Populus agrees in all essential characters with the form
of Polyporus dryophilus found on oak. The sporophores are, however,
somewhat smaller than those usually found on oak and approach the
applanate type. (PI. XXI, fig. 7.) The hard granular core is always
present, but is formed between the sap wood and bark (PI. XXI, fig. 8),
as the fungus is able to rot the sapwood as well as the heart of this host.
It therefore does not have to depend on branch holes or other openings
through the sapwood in order to form its sporophores, as it does in the
oak.

Through the kindness of Von Tubeuf the writer obtained a European
specimen of Hartig’s so-called rot of Polyporus dryadeus in oaks. (PI.
XXII, fig. i.)1 It is unquestionably the rot produced by P. dryophilus .

(PI. XXII, fig. 3.)

The following discussion of the rot caused by Polyporus dryadeus
embodies the results obtained from extensive field studies made in
the forests of Arkansas, eastern Texas, Oklahoma, Maryland, and
Virginia.

The sporophores of P. dryadeus are always found at or very near
the ground at the base of the host. This first suggested to the writer
that the fungus might be a true root-rotting organism. Trees with
sporophores at their bases and wind- thrown oaks with and without
the sporophores attached were carefully studied. Sections of the
trees were cut, roots dug up and examined, and every effort made to
determine exactly the character of the rot produced. The roots and
stools of 20 trees attacked by this disease were examined, and sections
of the various stages of the rot were studied.

The microscopic characters of the rot from each tree were found to
be identical, although of the 20 trees examined 5 were in Arkansas,
3 in Texas, 2 in Oklahoma, 4 in Maryland, and 6 in Virginia. In every
instance the trees were found to have a white rot which attacks first
the sap and finally the heartwood of the roots. The rot originates in
the lower portion of the roots and spreads in them toward the base of
the tree.

The first evidence of the disease is a reddish brown discoloration of
the inner bark and cambium. If the diseased roots are exposed in a
damp chamber at this stage, white floccose spots of mycelium will
appear on the outside of the bark, but the rot has not yet become
evident in the wood. As the rot progresses, discolored, watery, reddish

1 Figure i on Plate XXII was made from a photograph of a piece of the original type material used by
Hartig in his description of the rot of Polyporus dryadeus (1878).

242

Journal of Agricultural Research

Vol. I, No. 3

brown areas, which become hazel in color when the wood is dried,
appear on the surface of the sap wood and in its outer layers. At this
stage a cross section of the root has a mottled appearance, and this
discoloration gradually spreads till the root is affected to its center.
The earliest discolored spots have by this time turned white. (PI.
XXII, fig. 2.) Tater, as the rot ages, especially in the larger roots
which lie near the surface of the ground, this white changes to a cream
and finally to a straw color. The lower portion of the smaller diseased
roots, those 2 inches or less in diameter, become completely rotted and
white throughout before the advancing rot has reached the stool of the
tree. On these small rotted roots the bark separates easily from the
wood, since much of the living bark has been destroyed. The bast fibers,
however, remain intact, which gives the inner bark a loose, shredded
appearance. The rot gradually moves up the roots till the stool is
reached. This is also attacked by the fungus, but the rotted area ends
abruptly at the surface of the ground.

A radial-longitudinal section of the rot in a fresh state has a sodden,
watery appearance, with white longitudinal and transverse lines some¬
what like the rot produced by Polyporus hispidus in oaks. These white
lines or bands are not cellulose, however, but are spaces filled with air
and the mycelium of the fungus in the region of the large vessels. When
the rotted wood is thoroughly dry, these white lines disappear, and the
uniformly creamy-white rot is left. The rot in all the trees examined
did not extend for any appreciable distance into the heartwood of the
trunk proper above the collar of the tree, even when the large, completely
buried roots, 6 to 12 inches in diameter, were rotted throughout.

The thoroughly rotted wood when dry is very light in weight and,
superficially, looks and feels like pith. If a freshly dug root in the
advanced stage of the rot is twisted, it will split into concentric layers and
also into longitudinal blocks, giving the broken end of the root a coarse,
fibrous appearance. The lower ends of the diseased roots may be in a
thoroughly rotted condition, easily splitting into these concentric layers
and rough, fibrous masses, while that portion of the root next to the
base of the tree remains comparatively sound. The roots of several of
the trees overthrown by the wind were thus affected. The presence of
this rot is often indicated by irregular white mycelial patches on the
outside of the bark of the root or of the stool of the tree.

In a radial-longitudinal section through the heartwood of a diseased
root the advancing line of the rot first appears as a watery dark-brown
zone 1 to 3 inches wide. This dark area terminates rather abruptly in
the ultimate cream-colored rot on one side and in the sound heartwood
on the other. A microscopic examination of the diseased wood shows
that the starch and other cell contents of the roots are first extracted;
then the walls of the wood elements are gradually destroyed, especially
the walls of the tracheids and vessels adjacent to the large medullary

Dec. io, 1913

Polyporus Dryadeus

243

rays. The bordered pits in the vessels are usually reduced to long,
elliptical openings running transversely across the walls, and the bor¬
dered pits of the tracheids become large, round holes, which often coa¬
lesce, thus splitting the tracheids longitudinally. The pits of both
large and small medullary rays are somewhat enlarged, while their radial
and tangential walls are perforated with holes.

Even in the early stages of the rot, when the discolored spots are
beginning to show in the sapwood of the roots, the vessels have color¬
less hyphae in them, while in the later stages many of the vessels become
filled with a mass of colorless hyphae having filaments 4/i or less in diam¬
eter. The wood-parenchyma fibers show enlarged pits and perforated
radial walls, and the pits in the wood fibers are also enlarged. The walls
of the medullary rays are much corroded and often disappear entirely.

Only very slight evidence of delignification is shown by the chloriodid
of zinc test. After standing 24 hours in this reagent there is a slight
cellulose reaction in the walls of the vessels, tracheids, and wood fibers
but none in the medullary rays. In making free-hand sections of the
diseased wood the medullary rays and vessels are easily ruptured, owing
to the thinning and weakening of the walls by the solvent action of the
fungus.

The concentric splitting of the rotted wood usually occurs in the zone
of the larger vessels, which are weakened by the corrosion of their bor¬
dered pits and walls. The longitudinal splitting is caused by the coa¬
lescence of the enlarged bordered pits of the tracheids and the thinned
walls of the medullary rays. The discolored areas seen in the earlier
stages of the rot are due to the presence in the cells of the medullary
rays, wood parenchyma fibers, and sometimes in the lumen of the wood
fibers of a brownish liquid, which disappears before the white stage of
the rot is reached. In the final stage of the rot the wood is somewhat
spongy in texture and when dry is easily crushed between the fingers.

Old sporophores were often found at different places on the collar of
the diseased tree, due probably to the gradual rotting of the roots upward
toward the stool of the tree and the formation of sporophores whenever
a rotted area reached the collar of the tree or the underside of a root
whose upper surface was exposed to the air. The sporophores are
usually attached to the trunk of the tree at the surface of the ground,
but they were also found on the exposed roots or even in rare cases on
the ground, having been produced from hyphae issuing through the soil
from diseased roots lying a short distance below. Only one sporophore
was found on the trunk at a distance above the collar of the tree, and in
this case two trees had grown together at the butts for a distance of
12 inches. The rot had extended from the diseased roots upward in the
injured sapwood of the oak along the juncture of the two trunks, and a
small sporophore had formed 10 inches from the ground.

17072 13— 5

244

Journal of Agricultural Research

Vol. I, No. 3

The sporophores when old and mature usually have a hard fibrous-
corky to corky-woody context and a very rough, uneven, tuberculate
upper surface, owing to the leaves, twigs, and other foreign substances
falling on the upper surface of the growing pileus. (PI. XXII, fig. 4.)
After weathering for some months, the color of the pileus is a chestnut
brown or sometimes becomes almost black and rimose. The old spor¬
ophores as a rule are partially destroyed by insects, especially the sub-
hymenial layer and the adjacent ends of the pores. Portions of the
outer pore surface, the central part of the context, and the base of the
sporophores usually persist and can be found attached to the bases of
the diseased trees for several years after maturity.

The mouths of the pores in the weathered sporophores are stuffed to
a depth of 0.5 to 1 millimeter with a firm, brown mycelial mass, thus
completely hiding all trace of the pores from a surface view. This stuffed
pore layer becomes hard and brittle and gradually cracks in weathering
and peels off from the deeper and more insect-eaten portion. Immature
specimens shipped before being thoroughly desiccated have the tubes
loosely stuffed with a delicate, white arachnoid mycelium, which appears
on the spore surface as a thin creamy layer about 0.5 of a millimeter
thick. This condition is probably due to a growth developed in the
sporophore while in transit in a damp state. The stuffed mouths of the
pores in old weathered sporophores is apparently a normal state of old
specimens from certain sections of the United States. However, this
stuffed condition of the pores in old sporophores is not always present,
as several specimens both from America and Europe were seen by the
writer in which the mouths were entirely free and open.

The tubes in all the specimens examined — both American and Euro¬
pean — contain characteristic setse. They are dark chestnut brown,
thick walled, curved, cat’s claw to hawk beaked in shape, giving them a
somewhat bulbous-shaped base when seen in side view. They are 7 to
12 pi thick at base, 15 to 24/jt long, and usually project 10 to 20 pt beyond
the hymenial surface into the tube cavity.

The sporophores vary greatly in shape and size, ranging from 9 cm.
long, 5 cm. wide, and \% cm. thick to 20 cm. long, 15 cm. wide, and 10
cm. thick, and may be simple or imbricated, depending to a great extent
on the environment and food supply. In many of the thick sporophores
growing from the collar of the tree the pore surface is borne at an angle
of 40° to 6o° to a horizontal plane. In the thinner and broader speci¬
mens the pore surface approaches more nearly the normal angle of other
dimidiate sporophores. The margin is very thick and rounded in most
of the specimens. The cavities left in the upper surface of the pileus by
the drops of water which exude during the rapidly growing period of
the sporophore are plainly discernible even in many of the old sporo¬
phores. The pore surface extends entirely to the point of attachment
to the substratum even when the sporophore has a rounded substipe,
as is often the case when it forms on the upper surface of exposed roots.

Dec. io, 1913

Polyporus Dryadeus

245

When sporophores are developed at the collar of trees growing in sandy
land, the soil for 4 to 6 inches wide and 2 to 3 inches deep immediately
at the base of the sporophore is often cemented into a hard, compact,
bricklike mass, apparently by hyphae, as many colorless fungous threads
were found ramifying through it.

Polyporus dryadeus has been found attacking the roots of Quercus
texana Buckl. and Q. nigra L. in eastern Texas. Some of the diseased
trees were dying, while others were evidently in poor health. It has
been found on Q. alba L. and Q. veluiina Lam. in the Ozark National
Forest, of Arkansas. The majority of the trees in the Ozarks affected
with the disease caused by P. dryadeus were growing on sandy ridges and
southern slopes where the soil was thin and conditions were unfavorable
to rapid, vigorous growth. Two trees of Q. minor (Marsh) Sarg. were
found with this disease in Oklahoma; one was dead and the other in
apparently fair health.

Polyporus dryadeus also occurs in Q. alba L. , Q. rubra L. , and Q. prinus
L. in Virginia, where seven trees were found with this rot ; five were grow¬
ing in crowded, unfavorable conditions, while one was standing at some
distance from other trees and was apparently in good health. Yet at
least two large roots of this lone tree — a white oak — were thoroughly
rotted, while sporophores were found on three sides of the tree, one
growing from the top of an exposed root. This sporophore was over 1
foot tall and at least as wide, judging from the old weathered remains.
It was from this root that figure 5 of Plate XXII was taken. Of the
five crowded trees one was much suppressed and would probably have died
in a year or two. This tree was dug up, and studies were made of its
roots, stool, and trunk. All of its roots, except three large lateral ones
which ran near the surface of the ground, were completely rotted by P.
dryadeus . The three living roots were partially rotted on the lower side
and at the ends, but were still alive and strong enough to hold the tree
in the ground. Old sporophores were found on all sides of this tree at
the ground line.

The trees of Quercus prinus which were attacked by this root rot were
found by Mr. G. F. Gravatt, of the Office of Investigations in Forest Path¬
ology, who made the following statement concerning the diseased trees :

Early in July at Bluemont, Va. , three small trees of Quercus prinus were found which
had been killed while in full leaf and which from a distance were mistaken for chest¬
nut trees that had been girdled by the chestnut bark disease (Endothia parasitica ).
Whitish spots of mycelium were found on the bark of nearly every root, while the lower
portions of the roots were so thoroughly rotted that the two smaller trees were easily
pulled up by hand. The two small trees were somewhat suppressed, but the largest
(sH inches in diameter) was situated in an open space in the woods. These three
trees were about 100 yards distant from each other.

The writer examined the rot from the roots of these diseased trees and
found that it was caused by Polyporus dryadeus .

246

Journal of Agricultural Research

Vol. I, No. 3

Four trees of Quercus alba were found affected with this disease in
Maryland. All had been uprooted by the wind, two very recently, so
that the character of the earlier stages of the rot and its progress in the
roots was easily observed. In both of these trees the rot was only in
the lower ends of the roots and had not reached the stool nor formed
sporophores. Three of these uprooted trees were growing in dense
stands and were much suppressed.

Oaks which have been uprooted by the wind may be separated into two
classes: (1) Those whose root system has been weakened by insect or
fungous attack and (2) those with a very shallow root system, due to the
presence of impermeable layers of rock in the subsoil or to the ground-
water being constantly near the surface of the ground (within 1 to 2 feet) .
Trees uprooted by wind owing to rotten roots have very little soil adher¬
ing to the upturned stool of the tree, as most of the roots break off within
1 to 2 feet of the base of the tree. On the other hand a tree with a sound
root system brings with it when uprooted a large mass of earth several
cubic yards in size. Bearing this in mind one can often distinguish, even
at a distance, wind- thrown trees with sound roots from those overthrown
on account of root-rot.

In every instance where the sporophores of Polyporus dryadeus were
found on trees the roots were diseased with the same type of root-rot. In
wind-thrown trees where the disease was not far enough advanced to pro¬
duce sporophores the rot was identical with that obtained from the roots
of trees which had sporophores of P. dryadeus . The rot in such uprooted
trees evidently began at some point on the lower end of the roots and
advanced up the roots toward the base of the tree, stopping, however,
when it reached the surface of the ground. Roots lying very near the
surface of the soil, especially large ones with their upper surfaces exposed
to the air, are not entirely rotted or even killed by this fungus. Many
instances of such superficial roots were found in which the part under¬
ground was rotted while the upper portion remained alive. The cross
section of the root illustrated in Plate XXII, figure 6, shows the upper
part alive, while the lower and more deeply buried portion is rotted. This
root forked some 2 feet from the tree; one root, 10 inches in diameter,
went down deep in the soil and was thoroughly rotten and dead; the other
fork was 2 to 4 inches deep and was perfectly sound 2 feet from where the
rotted root joined it.

The inability of the fungus to rot exposed roots and the trunk above
the surface of the soil, coupled with the further fact that the sporophores
usually are attached to what superficially appears to be sound wood,
probably explains why the connection between this rot and the fungus
causing it has not been previously noted. Trees in all stages of this
disease were seen; some were already dead, others dying, others on the
decline, while some showed no evidence of the disease until they were
overthrown by the wind and the decayed roots were exposed. Some of

Dec. 10, 1913

Polyporus Dryadeus

247

the trees bearing sporophores were apparently in a healthy condition,
yet an examination of the root system showed in every case one or more
large roots completely rotted Two stumps of Quercus alba were found
with sporophores of Polyporus dryadeus springing from the rotted roots.
In no instance were trees which were attacked by this fungus found in
groups or even adjacent to each other The majority of the trees with
this disease in their roots were growing under unfavorable environments.
The boles of some of them were also attacked by various heart-rotting
fungi, while others were perfectly sound above the collar, although they
bore sporophores of P. dryadeus at the ground line.

No rhizomorphs of any kind were found associated with this rot,
either beneath the bark, on the surface of the roots, or ramifying in the
adjacent soil. How the lower part of the smaller roots became infected
is not known.

The identity of the fungus causing this root-rot with the European
fungus known as Polyporus dryadeus may be questioned. Through the
courtesy of the officials in charge, the writer was permitted to examine
all the American and European specimens of P. dryadeus in the fol¬
lowing herbaria:

Pathological and Mycological Collections of the Department of Agri¬
culture, at Washington, D. C., Herbarium of the New York Botanical
Garden, and the Cryptogamic Herbarium of Harvard University.

Authentic specimens of Polyporus dryadeus from America, England,
France, Germany, and Austria were examined, and a careful comparison
of each with the material used as the basis of this article showed that the
American plant under discussion is undoubtedly identical with the Euro¬
pean fungus known as P. dryadeus .

There are three collections in the laboratory of the Office of Investi¬
gations in Forest Pathology, at Washington, D. C., of a Polyporus on
Tsuga heterophylla from three widely separated localities in the State of
Washington. These specimens were collected by C. J. Humphrey, of
this office, and the legends accompanying them indicate that the sporo¬
phores were attached to the host at or near the surface of the ground
and that the plant is a true parasite that kills the trees it attacks.
These specimens agree in all essential characters, both gross and micro¬
scopic, with Polyporus dryadeus , and although the writer has not seen
the rot produced in this host, he believes the fungus is this plant.

SUMMARY

(1) Polyporus dryadeus is a root parasite of the oak, producing a white
sap rot and a heart rot in the roots.

(2) In all the trees examined this rot did not extend upward into the
tree as a true heart or sap rot of the trunk, but was limited to the under¬
ground parts of the tree.

248

Journal of Agricultural Research

Vol. I. No. 3

(3) The rot and sporophore described and figured by Robert Hartig
do not belong to Poly poms dryadeus, but to Poly poms dryophilus.

(4) In the majority of cases only old or much suppressed trees or trees
growing under very unfavorable conditions were found attacked by this
disease.

(5) The disease does not seem to spread readily to adjacent trees.

(6) The disease is widely distributed both in America and in Europe
and is probably found in these countries throughout the range of the oak.

LITERATURE CITED

1789. Bueeiard, Pierre;. Herbier de la France. Paris, pi. 458.

1791. - Histoire des Champignons de la France, t. 1, Paris, p. 356, pi. 458.

1799. Persoon, C. H. Observationes Mycologicae. pars 2, Lipsiae. p. 3.

1801. - Synopsis Methodica Fungorum. pars 2, Gottingae. p. 537.

1821. Fries, Eeias. Systema Mycologicum. v. 1, Gryphiswaldiae. p. 374.

1849. Hussey, Mrs. T. J. Illustrations of British Mycology. London, [ser. 1], pi. 21.
1878. Hartig, Robert. Die Zersetzungserscheinungen des Holzes der Nadelholz-
baume und der Eiche. Berlin, p. 124-128, pi. 17.

1897. Tubeuf, Kare von. Diseases of Plants Induced by Cryptogamic Parasites.

English edition by W. G. Smith. London, p. 440-442, fig. 272-274.

1910. Massee, George. Diseases of Cultivated Plants and Trees. London, p.
380-381.

1910. Hedgcock, George G. Notes on some diseases of trees in our national forests.
[Not published.] Abstract in Science, n. s., v. 31, p. 751.

1912. - Notes on some diseases of trees in our national forests. — II. Phyto¬

pathology, v. 2, no. 2, p. 73-74.

1913. - Notes on some diseases of trees in our national forests. — III. Phy¬

topathology, v. 3, no. 2, p. m-114.

1913. Leoyd, C. G. Letter No. 44. p. 8, note 47.

DESCRIPTION OF PLATES

Plats XXI. Fig. i. — Polyporus dryophilus : A median-longitudinal section of a
sporophore on Quercus alba from Arkansas, showing the granular
core and the white mycelial lines in the central and rear portion.

Fig. 2. — Polyporus dryophilus: Side view of the ungulate type of sporo¬
phore on Quercus calif ornica from California.

Fig. 3. — Polyporus dryophilus: Median-longitudinal section of the glo¬
bose type of sporophore on Quercus garryana from California, showing
the large granular core and prominent white mycelial lines.

Fig. 4. — Polyporus dryadeus: Median-longitudinal view of a young
sporophore on Quercus texana from Texas, showing the fibrous, non-
granular nature of the context.

Fig. 5. — Polyporus fulvus Fries: Median-longitudinal view of a sporo¬
phore on Quercus sp. from Sweden, showing the granular core char¬
acteristic of P. dryophilus .

Fig. 6. — Polyporus vulpinus: Median-longitudinal view of sporophore
on Populus sp. from Sweden, showing the granular core character¬
istic of P. dryophilus.

Fig. 7. — Polyporus dryophilus: Front view of the applanate type of a
sporophore on Populus tremuloides from Colorado, showing the faint
zones on the pileus where the hairs have disappeared.

Fig. 8. — Polyporus dryophilus: Median-longitudinal view of sporo¬
phore on Populus tremuloides from Colorado, showing the granular
core originating between the sapwood and bark and extending into
the center of the sporophore.

XXII, Fig. 1. — Polyporus dryophilus: Radial-longitudinal view of the rot
occurring in Quercus sp. from Europe and said to be the rot pro¬
duced by P. dryadeus.

Fig. 2. — Polyporus dryadeus: Cross section of a small root of Quercus
alba from Maryland, showing the mottled appearance of the diseased
wood in the middle stages of the rot.

Fig. 3. — Polyporus dryophilus: Radial -longitudinal view of the rot ap¬
pearing in Quercus alba from Arkansas, showing the advancing line
of rot in a branch.

Fig. 4. — Polyporus dryadeus: Upper surface of a sporophore on roots of
Quercus texana from Texas, showing the rough tuberculate pileus.

Fig. 5. — Polyporus dryadeus: Rot occurring in an apparently sound
root of Quercus alba from Virginia, showing cross section of a dis¬
eased root, immediately adjacent to the point of attachment of a
large sporophore of P. dryadeus , 1 foot high and 1 foot wide. Some
sound, living wood is still present.

Fig. 6. — Polyporus dryadeus: Cross section of diseased root of Quercus
alba from Virginia, showing the nearly sound, living upper half of
the root and the badly diseased lower half.

(25°)

Plate XXI

Polyporus Dryad(

Plate XXII

THE EOOT-ROT OF THE SWEET POTATO

By L. L. Harter,

Pathologist , Cotton and Truck Disease and Sugar-Plant Investigations ,

Bureau of Plant Industry

INTRODUCTION

On August 9, 1912, Mr. O. H. Weiss sent the writer some diseased
sweet-potato ( Ipomoea batatas) vines from the vicinity of the Dismal
Swamp, Va., with a request for information regarding the nature of the
trouble. The stems for a short distance above the ground were covered
with black fruiting bodies of a fungus, and suggested macroscopically
the conidial stage of Diaporthe batatatis , the cause of the sweet-potato
dry-rot. Careful examination of the material showed that in structure
these fruiting bodies differed from those of the dry-rot organism, although
it was apparent that both fungi belonged to the same general group.
The organism was isolated in pure cultures from material taken from
diseased sweet-potato stems and its parasitic habits and growth in arti¬
ficial cultures compared with the dry-rot organism.

On August 22, 1912, the writer visited the sweet-potato fields near the
Dismal Swamp in order to observe the disease under natural conditions
and to ascertain the extent of the loss. The disease was found in prac¬
tically every field, causing a loss of from 10 to 50 per cent of the crop,
and in exceptional cases even more.

During August, 1913, the disease was found for the first time in many
fields near Cape Charles and Keller, Va. Whether this is the first appear¬
ance of the disease in this part of the State is not known. The writer
had inspected many fields in this section for several summers previous to
1913 and never observed the disease. It seems likely, therefore, that the
disease is either new to these places or has heretofore occurred only to
a very limited extent. The organism was isolated from specimens col¬
lected at both Cape Charles and Keller, and it was found to be identical
with the one obtained during 1912 and 1913 from the vicinity of the
Dismal Swamp.

Inquiry among the farmers in the vicinity failed to give a definite idea
as to how long the disease has been prevalent. It was learned, however,
that the disease has increased in severity in the last few years, and if not
checked is likely to prove a serious handicap to the growing of a crop
that would otherwise be a profitable industry.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(231)

Vol. 1, No. 3
Dec. io, 1913
G — 7

252

Journal of Agricultural Research

Vol. I, No. 3

DIAGNOSIS OF FOOT-ROT

The foot-rot organism is a slow-growing parasite, especially during
the earlier stages of infection. During about the first three weeks after
inoculation, only a slight enlargement of the wound in all directions takes
place. About three weeks seems to be required for the fungus to over
come the plant sufficiently to cause any marked reduction in its vitality
or vigor. As soon, however, as the fungus gets the upper hand, it de¬
velops very rapidly and in about one week more completely girdles and
extends along the stem from 2 to 5 inches, killing the plant by the de¬
struction of the cortex. At the end of about another week wilting of
the leaves is first observed, the plants beginning to die soon afterwards.
There is considerable variation in the length of time a plant will live
after becoming infected, especially under greenhouse conditions, some of
the plants dying in three or four weeks, while others may survive for one
to four weeks longer. It is also interesting to note in this connection
that those plants appearing to be the strongest when inoculated are likely
to be the first to succumb to the disease. An explanation of this may
be that a vigorously growing plant may stimulate the fungus to more
rapid development.

The first sign of the disease of inoculated plants is a blackening of the
cortex of the stem at the point of inoculation. When inoculated at the
soil line, the fungus seldom grows more than half an inch below the sur¬
face of the ground, but it extends up the stem several inches. The leaves
near the point of inoculation are invaded and soon turn yellow and fall
off. Under greenhouse conditions numerous black pycnidia break
through the epidermis of the stem (PI. XXIII, fig. A) along the black¬
ened area about the time the foliage begins to wilt. Under natural con¬
ditions in the field, on the other hand, the pycnidia form on the invaded
tissue before the wilting of the plant. It was observed also that diseased
plants will survive under field conditions much longer than in pots in
the greenhouse, where they are naturally handicapped by artificial con¬
ditions. Many diseased plants in the field with fruiting bodies abund¬
antly formed on the stem are often sustained by the roots which are
thrown out at the nodes along the stems, although the main stem may
be nearly destroyed by the fungus. If not supported by roots at the
nodes, the diseased plants readily succumb.

As a rule, the disease is confined to the stem of the plant from the soil
line to 4 or 5 inches above it. However, at Cape Charles, Va., in some
of the low, rather wet fields, where there was a rank vegetative growth,
vines were found diseased several feet from the hill. In such cases in¬
fection evidently took place at the node and spread in each direction
(PI. XXIV), the vine on each side of the diseased area remaining healthy.
The organism isolated from pycnidia on such diseased spots was identical
with the one obtained from the stem.

Dec. io, 1913

Foot-Rot of the Sweet Potato

253

CAUSE OF THE FOOT-ROT

The organism causing the foot-rot of the sweet potato has been de¬
scribed as Plenodomus destruens.1 2 It has also been pointed out that the
fungus does not fit well into this genus or into any of the present-known
genera. At the time, however, it was thought better to describe it as a
new species of the genus Plenodomus rather than to create a new genus
in a group where there are already a great many genera. It is probable
that this organism is the conidial stage of an ascomycete which will
eventually be discovered, and in view of that fact its generic position
can only be temporary. It falls naturally in the order Sphaeropsidales
and is more closely related to Phoma, Phomopsis, and Phyllostycta than
to any of the other genera in the order.

The diagnosis of the genus Plenodomus as found in Saccardo’s Sylloge
Fungorum is somewhat brief. In 191 1 Diedicke3 * * worked over this genus,
describing it more fully and pointing out the characteristics which distin¬
guish it from Phomopsis, the genus with which it is most likely to be con¬
fused.

Since it is quite evident that the foot-rot fungus is not a Phoma,
differing from that genus (1) in having more irregularly shaped pycnidia
(PI. XXV, B) and (2) in having a well-defined beak (PI. XXV, A)y atten¬
tion will be given only to the characteristics which distinguish the foot-rot
organism from Phomopsis, the conidial stage of the sweet-potato dry-rot.

According to Diedicke, Plenodomus is characterized by having only
two walls composing the pycnidium — a dark outer wall and a hyaline
one within. The outer wall completely surrounds the pycnidium and is
of uniform thickness at the top and base. The inner hyaline layer is
composed of several layers of cells and is somewhat thicker than the
outer wall. The conidiophores are short, fragile, and inconspicuous.
The spores are rounded at both ends.8

On the other hand, the pycnidium of Phomopsis, according to the
same author, is composed of four walls. The upper portion of the
pycnidium, especially about the beak, is composed of thick black cells.
The dark color of this layer of cells becomes less conspicuous in the
lower portion and practically disappears at the base of the pycnidium.
Phomopsis is further characterized by the development of a stroma and
chambering of the pycnidium. The conidiophores are long, conspicu¬
ous, and awl-shaped, and the spores are spindle-shaped. Because of
the variation in the shape of the spores this latter character is of less
importance than some of the others in separating the genus from Pleno¬
domus. Stylospores are found in some species of Phomopsis.

1 Harter, L. I*. Foot rot, a new disease of the sweet potato. Phytopathology, v. 3, no. 4, p. 243-245,

2 fig., 1913.

2 Diedicke, H. Die Gattung Plenodomus Preuss. Ann. Mycol., Jahrg. 9, No. 2, p. 137-141, pi. 8, 1911.

3 This last character is perhaps of the least importance, since it is well known that the spores vary

greatly within the genus and even in the same species. In fact, the spores of some species of Phomopsis

have rounded ends.

254

Journal of Agricultural Research

Vol. I, No. 3

It will be seen, therefore, that the following characteristics belonging
to the dry-rot fungus are not found in the foot-rot organism : (i) Stroma;
(2) chambering of the pycnidium; (3) conidiophores conspicuously long
and awl-shaped; and (4) long, filiform, hook-shaped stylospores.

What is believed to be even more significant than the differences in
morphological characters between these two organisms is the difference
in parasitic habits and growth in artificial cultures. It has been pointed
out in a previous bulletin1 that the dry-rot fungus does not kill the
plant but lives in apparent harmony with it without injury. The
pycnidia appear on the stem only after the plant has been lifted and
kept in a damp chamber for- 10 days or 2 weeks, this being the first
evidence that the plant was infected. The organism occurs on the
petioles and leaves of dead plants and often develops on apparently
sound roots after a period of time in storage. Stylospores are frequently
found on the roots and stems.

The foot-rot disease, on the other hand, kills the plant in three to
eight weeks after infection by the destruction of the cortex of the stem
for several inches above and a little distance below the surface of the
soil. Pycnidia are formed on the diseased portion of the stem about
the time the foliage begins to wilt (PI. XXVI, fig. A)> and under field
conditions even earlier.

The growth of the organism on several kinds of the commonly used
artificial media and especially on synthetic agar 2 and on com meal 3 * 5 fur¬
nishes additional means of distinguishing the two diseases.

On synthetic agar the foot-rot fungus grows slowly and under normal
conditions forms a very compact growth, at first irregular in outline with
a slightly darker center, attaining a diameter of not more than 2 or 3 mm.
at the end of a week or 10 days. (PI. XXVII, fig. B .) On the same
culture medium the dry-rot fungus grows much faster, forming a loose,
flaky growth of uniformly white hyphae having an irregular outline.
(See PI. XXVII, fig. A .) The growth of the dry-rot fungus is so loose
and inconspicuous that it is scarcely visible until it has attained a diam¬
eter of 2 or 3 mm.

1 Harter, I,. L., and Field, Ethel C. A dry rot of sweet potatoes caused by Diaporthe batatatis. U. S.
Dept, of Agr., Bur. Plant Indus., Bui. 281,38 p., 4 pi., 1913.

2 Synthetic agar is prepared as follows:

Distilled water .

Dextrose .

Peptone (Witte's) .

Ammonium nitrate .

Potassium nitrate .

Magnesium sulphate .

Calcium chlorid .

Agar agar .

Grams.

1,000

200

2-5
o. I
20

Place the water in the beaker first; then add other ingredients in the order given. Stir and let stand
till the agar agar is moist. Steam 1 hour. Tube with constant stirring. Plug and autoclave for 15 minutes
at no0 C. Agar of high purity only should be used.

5 Corn-meal flasks are prepared as follows: Place 5 grams of com meal in a 100 c. c. flask. Add 45 c. c. of
distilled water and steam for 15 minutes. Plug and autoclave at 11 pounds pressure for 20 minutes.

Dec. io, 1913

Foot-Rot of the Sweet Potato

255

On corn meal the dry-rot organism forms a black stroma composed of
several pycnidia with long exserted beaks. The stroma is % to 1 or more
mm. in diameter and is preceded by a profuse growth of mycelia. The
foot-rot organism, on the other hand, forms no stroma on com meal.
The pycnidia stand separately and are very numerous, while the mycelial
growth is slight and inconspicuous. The pycnidia follow closely after the
growth of hyphae, the pycnidial zone increasing with the increase in
diameter of the mycelial growth. Spores are exuded in great quantities,
forming a yellowish transparent liquid over the surface of the medium.

ISOLATION OF THE FUNGUS

Pure cultures of the foot-rot organism were particularly easy to secure
by the poured-plate method. Stems on which the pycnidia were present
were thoroughly washed in hydrant water or, preferably, disinfected with
mercuric chlorid for about 40 seconds and then rinsed in sterile water. A
few of the pycnidia were then macerated in a watch glass in sterile water
and one or two loopfuls transferred to tubes of synthetic agar and plates
poured. The fungus grows very slowly on agars, particularly on syn¬
thetic agar. The colonies are not visible in the plates for three days and
often not until five or six days after they are made. Because of the char¬
acteristic growth on synthetic agar the organism can easily be picked out
from other fungi when the appearance of the colony is once known.

DESCRIPTION OF THE FUNGUS

Mycelium. — The appearance of the mycelium varies so markedly on
different culture media and according to the age of the culture that it
would be difficult to give a simple, characteristic, general description. In
young cultures and for the most part in old cultures it is nearly always
hyaline, although occasionally browned hyphae may be found. Oil glob¬
ules are found in the mycelia at all ages (PI. XXV, C). Hyaline, spherical
and oval, thick-walled bodies 8 to 13// in diameter, generally filled with oil
globules, intercalated or, rarely, terminally, in chains or singly (PI. XXV,
D ), occur in most media and at nearly all ages. Browned bodies morpho¬
logically similar to the hyaline ones but occurring mostly at the end of the
hyphae (PI. XXV, E ) are frequently found in older cultures. In 7-months-
old corn-meal cultures which were quite well dried out the brown bodies
were abundant, especially where the media came in contact with the glass.
In these cultures the hyaline forms were few. In 4-months-old cultures
of string beans brown and hyaline bodies and brown hyphae were present.
The brown hyphae were filled with numerous beadlike swellings. On the
other hand, in a rice culture of the same age only hyaline hyphae and
hyaline spherical or oval bodies were found.

Pycnidia. — The pycnidia are at first buried, but later break through
the epidermis, appearing as black dots scattered over the surface. They
stand close together on the stem and roots, but they are not confluent

256

Journal of Agricultural Research

Vol. I, No. 3

or only rarely so. (PI. XXIII, fig. A.) They are irregular in form and
vary greatly in size, averaging about 300/1 through their greatest diam¬
eter.

In cross section the pycnidia from the stem and roots show somewhat
different structures. From either source they are completely inclosed
by a dark, almost black, outer wall (PL XXV, A and B).

The pycnidia on the roots have a well-defined inner hyaline layer
almost equal in thickness to the outer wall (PI. XXV, A). On the stem
the dark wall is more conspicuous, being better developed than on the
root, and the inner hyaline layer is completely lacking (PI. XXV, B).

The basidia are short, fragile, somewhat inconspicuous, and arise from
the inner hyaline layer or from the dark wall in pycnidia where the hya¬
line layer is absent. They are 6 to 13/* in length and very narrow.

The spores are discharged through a beak varying somewhat in length,
which may arise from any part of the upper surface of the pycnidium.
In old dried specimens the upper portion of the pycnidium may fall
away.

Pycnospores. — The pycnospores are oblong, rounded at both ends,
6.8 to io.o(u long by 3.4 to 4.1 p. wide, with two large oil droplets. They
are hyaline, 1 -celled, and sometimes slightly curved (PI. XXV, F ).

In the same pycnidium on the host and occasionally on rice and on
sweet-potato-stem cultures are found in addition to the pycnospores hya¬
line curved or straight bodies 6 to 15 pi in length. These bodies are
somewhat cylindrical in shape and rounded or tapering at the ends (PI.
XXV, G). The function of these bodies is not known. Several attempts
have been made to germinate them, and while there have been some rea¬
sons to believe that a germ tube was developed, this point was not defi¬
nitely settled. These bodies were formed so sparingly in artificial media
that it was necessary to use those from the host in order to test their
germination in hanging-drop cultures in Van Tieghem cells. Because
of the difficulty in sterilizing this material, bacteria completely overran
the cultures in about 24 hours, thus terminating the experiment.

PARASITISM OF THE ORGANISM
INOCULATION EXPERIMENTS

The details of inoculations with Plenodomus are found in the following
pages. For convenience, the experiments are numbered and arranged
according to dates of inoculation and under the heading to which they
belong. The organisms used to make the inoculations are also desig¬
nated by numbers.1

1 For convenience and ready reference, separate numbers (100, ioi, 102, 108, and no) were given to the
different isolations where they or subcultures from them were used for inoculations. No. 100 was
given the organism obtained from specimens sent the writer Aug. 9, 1912, and No. 101 from specimens
collected Aug. 22 from the same locality. The other numbers used, 102, 108, and no, were given to the
organism reisolated from inoculated plants. A new number was given the fungus only when it was the
source from which other plants were to be inoculated. However, it should be kept in mind that these
different numbers represent only different isolations of the same organism ( Plenodomus destruens).

Dec. 10, 1913

Foot-Rot of the Sweet Potato

257

Most of the inoculations were made in the greenhouse, principally
because they were performed in the winter. One set, however, which
was conducted in the field, gave results so similar to those in the green¬
house that it was not possible to distinguish between them in any essen¬
tial details. The plants for inoculation were obtained from sound pota¬
toes carefully selected for the purpose. They were grown in pots of
sterilized soil and kept far enough apart to prevent accidental infection
from watering and overlapping of the vines. Only strong, vigorously
growing plants were inoculated, all others being thrown out. That
there was probably no accidental infection is shown by the fact that
not a single check in the whole series of inoculations became diseased.

Inoculations in the Field

Experiment No. i. — On August 26, 10 sweet-potato plants, the vines being about
3 feet long, were inoculated 1 on the Potomac Flats near Washington, D. C., by insert¬
ing pycnospores and hyphae of organism No. 102 (culture No. 1 of Aug. 15) into the
lower part of the stem. Ten plants pricked with a sterile needle were used as checks.

Results. — On September 18 all the inoculated plants were infected,2 the plants
turning yellow, and the lower leaves dropping off. The periphery of the stem for 3
to 5 inches above the ground was black, and pycnidia were abundantly formed thereon.
The stems were blackened throughout, but attempts to isolate the fungus from the
fibro- vascular bundles gave negative results. None of the checks were diseased. The
infected plants were all lifted on October 10, taken to the laboratory, and examined.
Pycnidia were present on all. On October 12 cultures were made from seven of these
plants, and the organism recovered 3 in each case.

Inoculations in the Greenhouse

Experiment No. 2. — On August 26, 1912, 10 young sweet-potato plants in pots
were inoculated with organism No. 100 (culture No. 8 of Aug. 15) by inserting
pycnospores and hyphae into the stem at the soil line. Five plants pricked with a
sterile needle were left as checks.

Results. — On September 16 four plants, on November 14 one, and on November
25 three, or a total of eight plants, were infected. None of the checks were diseased.
Pycnidia were formed on all the diseased plants and the organism recovered from
three. The experiment was terminated December 2, 1912.

Experiment No. 3. — On November 13 ten young sweet-potato plants in pots were
inoculated as in experiment No. 2 with organism No. 101 (culture No. 9 of Oct. 31).
Six plants pricked with a sterile needle were left as checks.

Results. — On December 14 one plant, on December 18 three, on December 21 one,
on December 26 one, on December 30 two, and on January 10 two, or a total of ten
plants, were infected. Pycnidia were present on eight plants when lifted and devel¬
oped on the other two after two days in a moist chamber. All the checks remained
healthy. The experiment was terminated January 17, 1913.

1 All inoculations recorded in this article, unless otherwise stated, have been made from cultures grown
on sterile moistened corn meal and only when spores were exuding from the pycnidia.

2 By “ infected” is to be understood the stage when the plant began to wilt and die. It was generally
quite evident some days earlier that the plants were infected, although they were not so recorded until
this stage was reached.

3 No attempt has been made to recover the organism from all diseased plants. Occasionally, however,
the fungus was recovered from infected plants in order to compare it with the original strain, or for the
purpose of inoculating it into other plants.

258

Journal of Agricultural Research

Vol. I, No. 3

Experiments Nos. 4, 5, 6, and 7. — On November 18 four sets of inoculations were
made of 10 plants each (40 plants in all) with organism No. 101 (culture No. 8 of Oct.
31), as follows: (No. 4) By smearing pycnospores on the leaves and spraying the
foliage with spores suspended in sterile water and covering the plants with bell jars
for 24 hours, (No. 5) by smearing pycnospores on the base of the stem, (No. 6) by pour¬
ing pycnospores suspended in sterile water about the plants, and (No. 7) by insert-
ing pycnospores and hyphae into the base of the stem. Six plants were left as checks.

Results. — (No. 4) No infection. (No. 5) On December 30 one plant, on January
8 one, on January 13 two, on January 15 one, on January 23 one, and on January
30 one, or a total of seven plants, were infected. Pycnidia were abundant on all
when lifted. (No. 6) On December 26 one plant, on December 30 one, on January
4 two, on January 6 one, on January 10 one, and on January 13 one, or a total of seven
plants, were infected. The infected plants were lifted on January 24 and pycnidia
were present on all. (No. 7) On December 21 two plants, on December 26 one, on
December 28 two, on January 6 one, on January n one, on January 13 one, on January
14 one, and on January 17 one, a total of ten plants, or all of those inoculated, were
infected. Pycnidia were present on nine of these plants when lifted and developed
on the other one after three days in a moist chamber. None of the checks were
diseased. The experiment was terminated on February 27.

Experiment No. 8. — On December 9 six 5-weeks-old sweet-potato plants in pots
were sprayed with pycnospores and hyphae of organism No. 101 (culture No. 1 of Nov.
12) suspended in sterile water. The plants were covered with bell jars and shaded
with paper for 24 hours. Six plants were left as checks.

Results. — No infection. The experiment was terminated February 27, 1913.

Experiment No. 9. — On December 28 eight 4-months-old sweet-potato plants
grown in pots were inoculated by inserting pycnospores and hyphae of organism No.
100 (culture No. 2 of Dec. 10) into the base of the stem. Six plants pricked with a
sterile needle were left as checks.

Results. — On January 23 one plant, on February 4 one, on February 7 three, and
on March 8 one, or a total of six plants, were infected. The checks remained healthy.
Pycnidia were present on all the infected plants when lifted. The organism was
recovered from two of the infected plants. The experiment was terminated March 27.

Only young plants were used in the first eight experiments. Experiment No. 9
was made with old plants (as compared with those used in experiment No. 8) for the
purpose of determining whether they were as susceptible as young ones to the foot rot.
The results indicate that they are.

Experiment No. 10. — On January 23, 1913, six sweet-potato plants (three old and
three young) grown in pots were sprayed with pycnospores of organism No. 100
(culture No. 3 of Dec. 28) suspended in sterile water. All the plants were making
a good growth. As soon as the plants were sprayed, they were covered with bell jars
and manila paper for 48 hours. Six plants were left as checks.

Results. — None of the plants were infected. The experiment was terminated
March 27, 1913.

Experiments Nos. ii and 12. — On January 17 ten young plants, each of Ipomoea
purpurea (E.) Roth, and Ipomoea hederacea Jacq. were inoculated with organism No.
100 (culture No. 4 of Dec. 28). Seven plants were left as checks.

Results. — No infection.

Experiment No. 13. — On December 2 five young plants of Ipomoea coccinea L.
in pots were inoculated at the base of the stem with organism No. 101 (culture No. 2
of Nov. 12). Five plants were left as checks.

Results.— On February 28, 1913, three plants were infected. The organism from
two of the plants was recovered by pouring plates from the pycnidia and from the
third plant by planting bits of diseased tissue in plates of synthetic agar.

Dec. 10, 1913

Foot-Rot of the Sweet Potato

259

None of the checks became diseased. The experiment was terminated February

28, 1913.

Experiment No. 14. — On May 9 seven sweet-potato plants in pots in the green¬
house were inoculated by inserting the hyphae (no pycnidia in the culture ) of organism
No. 101 (culture No. 2 of May 5) into the lower part of the stem. Six plants were left
as checks.

Results. — On May 31 six plants, and on June 4 one, or a total of seven plants, were
infected. None of the checks were diseased. When the experiment was terminated
on June 5, pycnidia were abundant on the stems of all diseased plants.

Experiment No. 15. — On September 3 six sweet-potato plants in pots in the
greenhouse were inoculated by inserting spores and hyphse of organism No. 101 (cul¬
ture No. 13 of Aug. 14) into the vine at the node 3 to 4 feet from the hill. Five other
vines were wounded with a sterile needle and left as checks.

Results. — On September 25 five of the vines were infected at the point of inoculation.
The organism had spread 2 inches or more each way from the point of inoculation.
None of the checks were diseased. The experiment was terminated October 5, 1913.

Experiment No. 16. — On September 3 five sweet-potato plants in pots in the green¬
house were inoculated by inserting spores and hyphae of organism No. 108 (culture
No. 17 of Aug. 18) into a vine at the node 3 to 4 feet from the hill. The checks were
the same as those used in experiment No. 15.

Results. — On September 25 all the vines were infected at the point of inoculation,
the organism spreading as in experiment No. 15. The experiment was terminated
October 5, 1913.

INOCULATIONS PROM REISOLATIONS

Experiment No. 17. — On October 5 twelve young sweet-potato plants in pots were
inoculated by inserting pycnospores and hyphae of organism No. 102 1 (culture No. 2
of Sept. 25) into the lower part of the stem. Ten plants pricked with a sterile needle
were left as checks.

Results.— On November 5 five plants, on November 11 one, on November 13 one,
on November 15 one, on November 25 one, and on December 9 three, or a total
of all 12 plants, were infected. None of the checks were infected. Pycnidia were
found on ten of these plants when lifted and developed on the other two after three
days in a moist chamber. The organism was recovered in pure cultures from seven
plants. The experiment was terminated on December 9, 1913.

Experiment No. 18. — On January 23 eight young sweet-potato plants in pots were
inoculated by inserting pycnospores and hyphae of organism No. 1082 (culture No. 2
of Jan. 11) into the lower part of the stem. Six plants were left as checks.

Results. — On February 28 one plant, on March 8 two, on March 13 four, and on
March 28 one , or a total of eight plants, were infected . Pycnidia were present on seven
of the diseased plants when lifted and developed on the other one after 10 days in a
moist chamber. None of the checks were diseased. Experiment terminated March

29, 1913.

Experiment No. 19. — On February 19 ten young sweet-potato plants in pots were
inoculated by inserting pycnospores and hyphae of organism No. 108 (culture No. 3 of
Jan. n) into the stem. Seven plants were left as checks.

Results. — On March 21 four plants, on March 24 one, on March 31 one, on April 4
one, on April 18 one, and on April 26 one, or a total of nine plants, were infected.

1 The organism recovered from plants inoculated in the greenhouse on Aug. 26 with No. 100 is known
as No. 102.

2 When the plants, inoculated on the Potomac Flats on Aug. 26, 1912, were dug, they were placed
with the roots attached in moist chambers in the laboratory. After several weeks the fungus grew from
the stem into the roots (PI. XXIII, B), from which it was recovered. This organism was numbered “108."

17072 — 13 - 6

26o

Journal of Agricultural Research

Vol. I, No. 3

Pycnidia were abundant on eight of the diseased plants when lifted on April 22. The
one remaining diseased plant was lifted on April 26, and pycnidia were then present.
The experiment was terminated April 26, 1913.

Experiment No. 20. — On March 13 six young sweet-potato plants were inoculated
by inserting spores and hyphae into the lower part of the stem with organism No. no 1
(culture No. 3 of Mar. 5). Five plants were left as checks.

Results. — On April 18 two plants, on April 22 two, and on April 23 one, or a total
of five plants, were infected. The diseased plants were lifted on April 26, and pyc¬
nidia were present on all. None of the check plants were diseased. The experiment
was terminated April 26, 1913.

Inoculations in the Laboratory

Experiment No. 21. — On November 18 eight mature sweet potatoes (not plants)
were inoculated by inserting pycnospores and hyphae of organism No. 101 (culture
No. 8 of Oct. 31) into the end of the potatoes. They were placed in cloth bags and
stored in the laboratory. Four potatoes pricked with a sterile needle were used as
checks.

Results. — No infection. The experiment was terminated January 31, 1913.

Experiment No. 22. — On April 4, 1913, six sound sweet potatoes were prepared
for inoculation by cutting away the ends of each so as to leave nothing but healthy
tissue. They were then thoroughly washed and disinfected by treating with mer¬
curic chlorid (1:1,000) for five minutes. They were afterwards rinsed in sterile
water and placed in a moist chamber on filter paper disinfected with corrosive sub¬
limate. Three of the potatoes were inoculated at the end and three at the side by
inserting spores and hyphae of organism No. 108 (culture No. 1 of Mar. 8). Four
other potatoes pricked with a sterile needle were used as checks.

Results. — On April 15 no signs of decay had started at the point of inoculation.
The filter paper appeared a little dry, and sterile water was added. After April 15
the rot developed and progressed rapidly in all the potatoes from the point of inocula¬
tion until by May 1 one potato was completely decayed and the others about one-third.
Plate XXVIII, figure A, shows a sweet potato inoculated at the end and figure C,
one inoculated at the side. Figures B and D are sections of figures A and C, respec¬
tively, showing the extent of the rot. The potatoes' inoculated at the side decayed
more rapidly than those inoculated at the end. Mature pycnidia and spores were
formed on the surface on May 1. The organism was recovered from the pycnidia and
from the diseased brown tissue of two potatoes.

The organism causes a chocolate-brown to almost black discoloration of the tissue,
but leaves it rather firm, even in the later stages. This is not a distinctive character¬
istic, since there are a number of rots of the sweet potato, nearly all of which produce
some shade of brown in the tissue and are in general so similar that it is practically
impossible to separate them by their macroscopic appearances. All of the check
potatoes remained sound.

1 This organism was obtained from plants of Ipomoea coccinea which were inoculated with organism No.

ioi.

Dec. io, 1913

Foot-Rot of the Sweet Potato

261

Table I. — Summary of results of inoculations with Plenodomus destruens .

Organ¬

ism

No.

Host.

Place of
inocula¬
tion.

Method of inocula¬
tion.

Number —

Num¬
ber of
checks
in¬
fected.

Ex¬

peri¬

ment

No.1 2 * * * * * * * 10

In¬

ocu¬

lated.

In¬

fect¬

ed.

Checks.

100. . .

Ipom 0 e a

Pot omac

By inserting spores

batatas.

Flats.

and hyphae into

the lower part of

the stem.

_ . .do .

Green-

. do. . .

house.

. . .do .

. . . do .

. do. . . . .

. . .do .

. do .

...do .

. . .do .

. do . . .

Ipom 0 e a

. . .do .

. do . . .

purpurea.

Ipom 0 e a

. . do .

. do .

hedera-

cea.

101. . .

Ipom 0 e a

. . .do .

. do .

coccinea.

. . .do .

. do .

. . .do .

. do .

. . .do .

. . .do . .

. do .

no. . .

...do .

. . . do .

. do .

...do .

. . .do .

By spraying foliage

2 4

with spores sus-

pended in water.

. . do .

. do . . .

. do.... . .

do .

. do .

101. . .

. . .do .

By smearing coni-

dia on lower part

of stem.

IOI. . .

. . .do .

By pouring spores

in water around

the plant.

101. . .

Ipomoea

. . .do .

By inserting

batatas.

hyphae into the

lower part of stem.

. . .do .

By inserting spores

and hyphae into

the node of vine

several feet from

the hill.

108 ,

.do .

. . .do .

. do .

IOI . , .

Storage

Labora-

By inserting spores

sweet

tory.

and hyphae into

p 0 t a -

the end of potato.

toes.

108. . .

. . .do .

. do .

1 For more complete data, the reader is referred to the experiments in the preceding pages corresponding
to the numbers of this column.

2 Experiments Nos. 4 to 7, inclusive, are combined in the body of the text, p. 258.

Discussion of Inoculation Experiments

Twenty-two sets of inoculations have been made with Plenodomus
destruens , 17 of which were on sweet-potato plants. Eighty-four sweet-
potato plants in nine different sets were inoculated by wounding the
lower part of the stem and inserting spores and hyphae. Seventy-eight

died of the disease. Seven plants were wounded in a similar manner

and inoculated with hyphae only, and all became infected. Eleven vines
in two sets were inoculated at the node several feet from the hill and io

became diseased. Spores and hyphae were smeared on the lower part

of the stems of io plants, care being taken to cause no wounds, and 7

became diseased. Spores suspended in sterile water were poured about

10 plants, and 7 died from the organism. The foliage of 26 plants in

262

Journal of Agricultural Research

Vol. I, No. $

three different sets was sprayed with the spores suspended in water, but
the disease was not produced thereby. Ten plants each of Ipomoea
hederacea and Ipomoea purpurea , and 5 plants of Ipomoea coccinea were
inoculated by inserting spores and hyphae into the lower part of the
stem. Three plants of Ipomoea coccinea were infected, the other species
not being injured by the fungus.

Two sets of inoculations have been made with potatoes taken from
storage. After inoculation one set was kept in the laboratory room in a
cloth bag and gave negative results. In the other experiment the potatoes
were placed in a damp chamber and kept moist with filter paper satu¬
rated with mercuric chlorid. Under these conditions the potatoes rotted
readily. (PI. XXVIII, figs. Ay B, C, and D.) The organism was
recovered in pure culture from the pycnidia formed thereon and from the
rotten tissue within.

The results of these experiments show that the foot-rot organism is a
vigorous wound parasite of Ipomoea batatas. In the greenhouse and
in the field infection can be readily produced by wounding the plant, but
this method is not imperative. It has been further shown that the tem¬
peratures and other environmental factors best suited for the growth of
the plants are likewise most favorable for the development of the fungus.
During warm, moist weather, when the plants grow most vigorously, the
disease was more severe than when growth was retarded by low tempera¬
ture. Plants at all ages were about equally susceptible to the disease.

It is also interesting to note in this connection that infection was
readily produced by inoculating with hyphae only, the result showing
that the progress of the disease was more rapid and the plants killed
sooner than when inoculations were made with spores.

HOW THE DISEASE IS PERPETUATED

The exact life history of this fungus will be in doubt so long as a perfect
stage is not known. It is evident, however, that an ascogenous stage is
not necessary to carry it from one season to the next. Diseased speci¬
mens on which there were numerous pycnidia were wintered out in a
wire cage covered over with leaves and some dirt with the hope that an
ascospore stage might develop. On the 27th of the following April the
specimens were examined, and normal pycnospores but no asci were
found.

A second lot of diseased specimens were wintered out in a wire cage
set on the ledge of a north window, where they were subjected to alter¬
nately dry and w^et weather and other atmospheric changes. When these
were examined on May 20, 1913, numerous normal conidia were present
and the organism recovered in culture.

There are at least two ways by means of which this disease may be
carried from one year to another: (1) On the dead vines and (2) on the

Dec. iof 1913

Foot-Rot of the Sweet Potato

263

potatoes in storage. In the locality in which this disease occurs, the
hotbeds are started about April 1, or even sooner, so that infection of
young plants might easily take place from pycnospores that had endured
as late as May 20. In old fields the beds are often made from the soil
on which sweet potatoes have been grown the previous year, thereby
providing the best conditions possible for direct infection of the new
crop. Furthermore, it was previously pointed out that the foot-rot
organism spreads from diseased stems to the potatoes and develops
pycnidia thereon. Experiments have also shown that under hotbed
conditions the organism will grow from diseased potatoes on to the slips
produced therefrom. Therefore, owing to the comparative obscurity of
diseases of this type, infected roots might readily be overlooked when
selecting seed, thereby making the sprouts growing from such potatoes
liable to infection.

The brown, spherical, thick-walled, chlamydosporelike bodies were
found in abundance embedded in the cortex of diseased parts of plants
wintered out in the wire cages. What function these forms have is not
yet known, although it is possible that they are able to reproduce the
fungus and serve to carry the organism through unfavorable conditions.
Repeated attempts, however, to germinate them have always given
negative results.

SOME PHYSIOLOGICAL CHARACTERISTICS OF THE FUNGUS
CHARACTER OF GROWTH ON DIFFERENT CULTURE MEDIA

The foot-rot fungus grows well on some kinds of media, but Sparsely
on others. The growth on some media may be regarded as character¬
istic of the organism and is unlike that of any other fungus with which
the writer is familiar.

A comparative study of growth has been made on nine different cul¬
ture media — i. e., corn meal, string-bean agar, string beans, Irish-potato
cylinders, sweet-potato cylinders, sweet-potato stems, rice, beef bouillon,
and beef agar. These different media have not been selected for any
particular reason, except that they are those commonly used and can
easily be duplicated. Five tubes (flasks in case of com meal) were
inoculated on November 25, 1912, with conidia from a 2 5 -day-old cul¬
ture grown on com meal. The tubes and flasks were kept in the light
on a table in the laboratory, the temperature of which varied from 180
to 240 C. They were kept under observation until January 31, 1913,
after which, owing to the dried condition of the cultures, no more notes
were made. The following records, given in number of days from the
beginning of the experiment, show the nature of the growth on the
different media.

264

Journal of Agricultural Research

Vol. I. No. 3

Corn Meal (10611)

2 days. — No visible growth.

4 days. — Yellowish white growth about 1 cm. in diameter. Hyphae growing close to
the medium.

7 days. — Hyphal growth about 4 cm. in diameter, slightly yellowish. Numerous
minute black pycnidia covering an area of about 2 cm. in diameter in the
center of the growth.

9 days. — Hyphal growth covering most of the surface of the medium and pycnidia
formed over about two-thirds. Spores just beginning to exude from pycnidia.
11 days. — Pycnidia covering most of surface of medium; exudate of spores forming
small viscid droplets.

14 to 17 days. — Abundant discharge of spores from the pycnidia.

21 days. — Spore discharge collecting in large globules, forming an almost continuous
covering over the surface of the medium.

25 days. — No change.

40 days. — Surface of medium completely covered with a slimy liquid containing
pycnospores.

67 days. — Hyphae hyaline. Numerous intercellular and terminal chlamydosporelike
bodies.

Com meal is the best of the media used for the development of pycnidia. The
pycnospores are first expelled in about one week, the process continuing for 30 or
40 days thereafter. At the end of that time the medium is covered with a slimy
liquid in which the spores are suspended. This liquid, often amounting to 5 c, c.,
is characteristic of growth on this medium and is apparently not due to the water
added, since that is taken up by the corn meal.

String-Bean Agar (1037)

4 days. — Sparse white growth.

7 days. — Heavy, white flaky growth of erect hyphae covering one-fourth of slant.

9 days.— Light-colored pycnidia collected in spots on surface of medium.

11 days. — No increase in mycelial growth; pycnidia dark; pycnospores exuding from
r**. pycnidia.

14 to 17 days. — Perceptible increase in the exudation of pycnospores.

21 days. — Exudate colorless, forming large droplets and uniting.

25 days. — No apparent change.

40 days. — Hyphal growth covering most of slant. Spores normal.

67 days. — Hyphae hyaline. A few chlamydosporelike bodies.

String-bean agar is only a fair medium for the growth of this fungus. The pycnidia
were sparingly formed as compared with the growth on com meal.

String Beans (1063)

4 days. — White, loose, flaky growth covering one-third of medium.

7 days. — White, loose, flaky growth covering three-fourths of medium.

9 days. — Felty grayish white growth of somewhat erect hyphae. Pycnidia collected
in spots. Pycnospores present.

11 days. — Pycnidia black.

14 days. — Slight exudate of spores from pycnidia.

17 days. — Slight increase in the discharge of spores.

21 to 25 days. — Exudates uniting, colorless.

1 A number is given to and a description made of each medium when it is prepared in the laboratory so
that it can be readily duplicated when desired. Unless otherwise stated, all media were prepared in the
laboratory of the Office of Cotton and Truck Disease and Sugar-Plant Investigations.

Dec. io, 1913

Foot-Rot of the Sweet Potato

265

40 days. — Medium studded with pycnidia. Exudate abundant. Pycnospores not
typical, being immature in appearance and irregular in shape.

67 days. — Hyphae hyaline. Many chlamydosporelike bodies. Long cylindrical
bodies present. (PI. XXV, G .)

Irish- Potato Cylinders (1036)

4 days. — Dense, felty white growth covering all of potato cylinder. Medium slightly
darkened.

7 days. — Scattered dark (not black) pycnidia forming.

9 days. — Pycnidia abundant, irregularly scattered; black, rather large.

11 days. — Pycnidia black and conspicuous; uniformly scattered over the medium.

14 days. — A slight exudate of spores from pycnidia.

17 days.— Pycnidia crowded together. Slight discharge of spores.

21 to 25 days. — Pycnidia numerous. No discharge of spores from the pycnidia.

40 days. — Potato cylinder studded with pycnidia. No discharge of spores. Pycno¬
spores abnormal, being apparently immature and irregular in shape.

67 days. — Hyphae hyaline. A few chlamydosporelike bodies. Long cylindrical
bodies present. (PI. XXV, G.)

Sweet-Potato Cylinders (1064)

4 days. — White procumbent growth of fairly dense hyphae covering one-half of potato
cylinder. Medium changed to a light chocolate-brown color.

7 days. — Feltlike growth covering all of medium. Potato cylinders changed to a
chocolate-brown color. Pycnidia forming; surface of medium grayish.

9 days. — Pycnidia crowded together, forming a felty grayish surface on the medium.

11 days. — Pycnidia formed in a dense grayish mass over surface of medium. Spores
exuding from the pycnidia.

14 to 25 days. — Slight discharge of spores.

40 days. — Medium covered with pycnidia. Spores exuding abundantly from one
tube, a little from another, and none from the remaining tubes.

67 days. — Many long cylindrical bodies. Hyphae hyaline.

Sweet- Potato Stems (1049)

4 days. — A sparse spreading growth of white hyphae covering one-fourth of stem.

7 days. — Sparse, grayish, somewhat irregular, cottony growth of erect hyphae.
Pycnidia black, larger than on com meal, and resembling those on the vines
under natural conditions.

9 days. — Pycnidia black, uniformly distributed over medium. Spores exuding from
the pycnidia.

11 days. — Pycnidia numerous; pycnospores exuding from pycnidia in brownish
globules.

14 to 17 days. — Increase in the discharge of spores from the pycnidia.

21 to 25 days. — Exudates from the pycnidia uniting.

40 days. — Stems studded with pycnidia with long beaks. Discharge of spores from
the pycnidia less than on com meal.

67 days. — Hyphae somewhat brown. A few long cylindrical bodies

Rice (967)

4 days. — No visible growth.

7 days. — Very slight mycelial growth. Many black, somewhat large pycnidia.

9 days. — Mycelial growth sparse. Spores just beginning to ooze from the pycnidia.

11 days. — Surface of medium studded with black pycnidia. Spores discharged from
the pycnidia in small globules.

266

Journal of Agricultural Research

Vol. I, No. 3

14 to 17 days. — Discharge of spores from the pycnidia increasing.

21 to 25 days. — Exudates from the pycnidia increasing and uniting.

40 days. — Pycnidia abundant, forming a black crust over the surface of the medium.

Exudate of spores from the pycnidia a yellowish slimy mass.

67 days. — A few long cylindrical bodies present. Hyphae mostly hyaline. No
chlamydosporelike bodies.

Rice, with the exception of com meal, is the best of all the media tried. Spores
are formed abundantly and exuded in large droplets from the pycnidia. A very scant
mycelial growth is formed on rice.

Bouillon (5725 J)

3 days. — No visible growth.

6 days. — A very sparse flaky white growth arising from individual spores lodged
against the glass below the surface of the medium.

8 days. — Growth below the surface of the medium from individual spores enlarging
and adhering to the glass. No floating hyphae. Slight surface growth against
the glass.

10 days. — Increase in mycelial growth.

14 days. — Pycnidia forming a black ring against the glass at the surface of the medium.
29 days. — Pycnospores very few, poorly developed, and not typical of the spores on
com meal or rice.

56 days. — Hyphse 'hyaline, with many intercellular, spherical swellings singly or in
chains.

Beef Agar (5726 *)

3 days. — Grayish, thick, felty growth, extending % cm. above surface of medium;
irregular in outline.

6 days. — Growth spreading; white pycnidia forming,

8-10 days. — Pycnidia forming a black line on the surface of the medium at the
point of contact with the glass; elsewhere on the surface of medium they are
collected into spots.

14 days. — Pycnidia large and black and increasing in number.

29 days. — No pycnidia at the edge of mycelial growth except in contact with glass.
Spores few and imperfectly formed.

56 days. — Hyphas hyaline, with many intercellularly or terminal spherical bodies
several times the diameter of the hyphse arranged singly or in chains. Very
few typical spores.

These results of tests with the different media1 2 bring out clearly
the fact that the development of the foot-rot organism is decidedly good
on some media and very poor on others. Numerous pycnidia and an

1 This medium was prepared in the Laboratory of Plant Pathology.

2 In addition to the results of growth obtained on the nine media discussed in the preceding pages, the
fungus was cultivated on a number of others, but not for the purpose of comparing the growth at the end
of stated periods of time; hence, these have not been included in the general description. Growth of the
fungus has been studied on mature stems of cotton (Gossypium herbaceum ), sweet clover ( Melilotus clba),
also on immature stems of sweet clover, lupine ( Lupinus sp. ; common varieties from Germany), oak (Quer-
cus sp.), tomato ( Lycopersicon esculenium) , and sweet potato ( Ipomoea batatas). The growth and produo
tion of fruiting bodies varied greatly on the different media. On oak and cotton there was but a sparse
growth, although a few pycnidia were formed. On tomato there was practically no growth. Numerous
fruiting bodies were produced on stems of sweet clover (mature and immature), sweet potato, and a fair
growth of hyphae with production of pycnidia on lupine. Mycelium is so sparingly formed that when
produced abundantly it is a sign that the medium is not suited to the best development of the fungus.
The production of pycnidia, on the other hand, is evidence that the medium is a most favorable one.

Dec. 10, 1913

Foot-Rot of the Sweet Potato

267

abundance of pycnospores are produced on corn meal and rice. On
beef agar and bouillon, on the other hand, the pycnidia are compara¬
tively few and mostly sterile. The development of pycnidia on steamed
sweet-potato stems was very similar to that found in nature, except that
the beaks were longer. A fair growth only was made on string beans
and Irish-potato and sweet-potato cylinders. Com meal and rice, how¬
ever, are the only media tried on which the growth could be regarded as
showing typical development, it being quite apparent in most other cases
that the conditions were quite abnormal.

GERMINATION OF PYCNOSPORES

Germination begins in about 6% to 7 hours in hanging drop cultures.
Growth in sterile or hydrant water is slow at first, the germ tube reaching
only about one-half to one and one-half times the length of the spore in
24 hours at room temperature (21.50 tO22.5°0.). In nutrient media a much
better growth is made. At the end of 48 hours the germ tube reaches a
length several times that of the spore and begins branchings. (PI. XXV,
H.) Growth is fairly rapid thereafter, although if compared with certain
Fusaria it would be regarded as a slow-growing organism both in arti¬
ficial culture and on the host. Preliminary to germination, the pycno¬
spores swell up, lose their original shape, and become more nearly
spherical.

INFLUENCE OF TEMPERATURE ON THE GERMINATION OF PYCNOSPORES

The minimum, optimum, and maximum temperatures for germination
of the pycnospores were determined by inoculating about 2 c. c. of rice
water in test tubes with spores of the fungus from a young culture on
corn meal. One set of cultures was placed in each of six thermostats,
the range of temperature being indicated in Table II. Another set was
similarly inoculated and placed in the laboratory room as a check. The
cultures were examined at the end of 18 hours, and those that had ger¬
minated freely were thrown out. The other cultures were continued
for 24, 42, or 48 hours, as necessity required.

Table II. — Limiting temperature for the germination of pycnospores.

Thermo¬

stat.

Temper¬

ature.

Time in hours.

VI .

°C.

11. 5 to 12.0
15.0 to 17.0
21.0 to 22.0
25.0 to 26.0
35*5 to 36.0
37*4 to 37.5
40.0 to 40.5

No germination .

Slight germination
Fair germination . .

VIII....
Room . .
X .

Slight germination . . .

Fair germination .

Good germination. . . .

XII .

XIII. ...

XIV. ...

Fair germination .

Fair germination. .

No germination.

268

Journal of Agricultural Research

Vol. I, No. 3

While no absolute limits have been established, Table II shows that
only a very slight germination of spores took place in thermostat VI
(i i .5 to 120 C.) at the end of 24 hours. As the temperature was increased,
germination became better until the optimum was reached in thermo¬
stats X (25.00 to 26.0° C.) and XII (35. 50 to 36.0° C.). Germination
was somewhat reduced in thermostat XIII (37.40 to 37.5 0 C.) and com¬
pletely prohibited in thermostat XIV (40.0° to 40.50 C.). Of the tem¬
peratures tried the minimum for germination would be found in ther¬
mostat VI (n.50 to 12.00 C.), the optimum in thermostats X (25.00 to
26.0° C.) and XII (35. 50 to 36.0° C.), and the maximum in thermostat
XIV (40.0° to 40. 50 C.).

VIABILITY OP PYCNOSPORES

Just how long the spores will retain their viability in a dried condition
is not known. The pycnospores on material collected on August 22,
1912, and kept in an envelope in the laboratory would not germinate
in plates of beef agar made on November 27. Hanging-drop cultures
were made with hydrant water in Van Tieghem cells from the same
material on December 1 1 , with similarly negative results. On the other
hand, pure cultures made on August 15, 1912, on corn meal retained
their viability to June 18, 1913. These results are not directly com¬
parable, since there is always a certain amount of moisture present in
the medium when the culture is started. Furthermore, as was previ¬
ously pointed out, this organism produces a considerable amount of
liquid on corn meal, even though there is no surplus water present in
the culture when inoculated.

INFLUENCE OF TEMPERATURE ON GROWTH

The influence of temperature on the growth of Plenodomus destruens
in cultures was determined by the use of 10 thermostats ranging in
average temperatures from 1.090 to 37.3 0 C., and in the laboratory
with an average temperature of 21.90 C. These temperatures varied
somewhat, as will be seen by referring to Table III, where the average
maximum and minimum temperature for each thermostat is recorded.

Cultures were made on January 15, 1913, on sterilized rice (1085) in
test tubes, it having been previously ascertained that this was a favor¬
able medium for the growth of the fungus. Five tubes were placed in
each of the 10 thermostats and one set in the culture room in the labo¬
ratory as a check. The cultures were kept in the incubators and under
observation for 21 days. Table III contains a record of the growth of
the organism in each thermostat and in the laboratory room on the
different dates covered by the experiment.

TablK III. Record of growth in laboratory room and in 10 thermoslats maintained at different temperatures (°C. ).

Dec. io, 1913

Foot-Rot of the Sweet Potato

269

1 Temperature readings were made about 9.15 a. m. and 4.15 p. m. each day. * The cultures were kept in a culture room in the middle of the laboratory.

270

Journal of Agricultural Research

Vol. I, No. 3

It is seen from Table III and also from figure 2 that the temperatures
of thermostats I, II, III, and V (1.090 to 9.00 C.) are prohibitive of growth.
A sparse growth took place in thermostat VI (11.20 to 14,0° C.) and
reached its maximum growth in the laboratory room (17. 20 to 24. 50 C.).
The best growth was obtained at an average temperature of 21.90 C. and
the next best in thermostat XI (29.40 to 30.8° C.). The growth of mycelia
in thermostat XI (29.40 to 30.8° C.) was better at the outset than at any
other temperature, although the production of pycnidia and spores was
not as good at the end of the experiment as in cultures growing in the
laboratory room. The medium in thermostat XI was decidedly dis¬
colored, a change which did not occur at any other temperature. In

AOS 4.S 7i€ 3.0 /as /5<2 SG4 /7S 2/3 302 373

thermostat XII (36.4° to 38.0° C.) a very slight growth of hyphae took
place during the first three or four days. No further development took
place thereafter. While these results do not definitely fix the limiting
temperature for growth, they show that the optimum probably lies some¬
where between 21.90 and 30.20 C., the minimum close to 12. 6° C., and the
maximum at about 37.3 0 C.

At the conclusion of the incubation period all the cultures were taken
from the thermostats and placed on a table in the laboratory room. At
the end of 10 days there was a good growth in all the tubes except in
those that were in thermostat XII (36.4° to 38.0° C.), cultures of which
had been killed in 21 days.

Dec. io, 1913

Foot-Rot of the Sweet Potato

271

A comparison of these results with similar experiments carried on
with the pycnidial stage of Diaporthe batatatis 1 shows that Plenodomus
destruens is limited to a narrower range of temperatures in its growth in
artificial cultures. The optimum temperature for growth of the dry-rot
organism was 30 C. higher than that of the foot-rot fungus. At the
lower temperatures the former made as good a growth at an average
temperature of 7.50 C. as the latter did at an average temperature of
12.6° C. At the higher temperature the foot-rot fungus was killed
when exposed for 21 days at an average temperature of 37.30 C., while
the dry-rot organism made some growth when exposed for 18 days at an
average temperature of 3 7.8° C.

influence; of light on the growth and production of pycnidia

It was found that the conidial stage of the dry-rot fungus (Diaporthe
batatatis) produced pycnidia only sparingly unless exposed to light.1
Contrary to these results, the foot-rot organism on rice cultures grew
equally well in darkness and in the light. Pycnidia were formed in
about 3 days, and the spores began exuding in small droplets in about
10 days.

DISSEMINATION OF THE DISEASE

From what we already know of the foot-rot disease it is evident that
there are several ways in which the organism may be carried from one
place to another. In view of the fact that the pycnospores will live
through the winter on the dead vines until as late as May 20, the plants
in near-by hotbeds and even in the fields are liable to infection from this
source. It has been shown that the organism causes a serious disease
of the stem of the plant and grows from there to the roots, forming
pycnidia on the surface. It is evident, therefore, that the use of such
potatoes for seed might account for a large part of the infections.

There is no way of determining to what extent insects, the wind, and
such agencies are responsible for the distribution of the disease, but they,
do doubt, play an important part. It is believed that this disease, like
many others, is also distributed from one field to another on farm imple¬
ments, the hoofs of animals, or by means of stable manure, etc. It is
a well-known fact that many farmers are careless about the disposition
of diseased and decayed sweet potatoes. Without suspecting the risk
they are taking, they often throw them on the manure pile or feed them
to stock without cooking. In either case the organism, if present on the
potatoes, might eventually be carried to the field. The wider distribu¬
tion of the disease— i. e., from one locality to another — must largely be
accounted for by the exchange of seed potatoes and seed plants.

1 Harter, L- L*» and Field, Ethel C. A dry rot of sweet potatoes caused by Diaporthe batatatis . U. S.
Dept. Agr., Bur. Plant Indus., Bui. 281, 38 p., 4 pi., 1913-

272

Journal of Agricultural Research

Vol. I, No. 3

POSSIBLE METHODS OP CONTROL

The suggestions here given for the control of foot-rot are not based on
experimental evidence, but on what would seem obvious from a knowl¬
edge of the disease and the methods of handling the crop. It has
already been pointed out (1) that the disease occurs both on the stem and
roots of sweet-potato plants; and (2) that the pycnospores of the fungus
can live through the winter and late enough the following spring to
infect the new crop. With these facts in mind it will be clear that pre¬
cautionary and sanitary measures should be employed. One of these
should consist in the careful selection of healthy potatoes for seed. Se¬
lection should be made preferably in the fall at digging time and any
suspicious potatoes should be discarded. They should be carefully ex¬
amined again in the spring when the disease is more easily recognized,
and all those that show any sign of disease should be rejected. While
disinfection of the seed in a solution of mercuric chlorid (1:1,000) will
not destroy the fungus buried beneath the surface of the potato, it will
kill all adhering spores and clean the potatoes so that diseased spots can
be more readily detected. After immersing for five minutes in the solu¬
tion, the potatoes should be rinsed in water and thoroughly dried. It
is advisable that disinfection be done on a clear, warm day, just before
the potatoes are put in the bed.

Soil that is likely to be infected with the disease should not be used in
the preparation of the hotbed. If, however, disease-free soil can not be
obtained, then it should be disinfected by steaming for one hour at a
temperature of ioo° C. If steam sterilization is not feasible, the soil may
be soaked in a formaldehyde (40 per cent) solution (1 : 200). If the latter
method of disinfection is employed, the soil should be treated at least
10 days before it is to be used, and it should be occasionally stirred to
assist in the escape of the gas.

All decayed, diseased, or discarded potatoes should not be fed raw to
stock, or thrown on the manure pile to compost, but should be cooked;
neither should the potatoes be thrown on the ground around the hotbed.
These practices are too common, and are liable to infect otherwise
disease-free beds.

Crop rotation is a good practice, whether for the control of diseases or
not, and should be practiced by every farmer. It is not yet known how
long this disease retains its vitality in the soil without sweet potatoes as
a host, but probably for several years. At least three years should be
allowed between crops whenever diseases of this type are found, although
it is doubtful if this length of time will completely eradicate it from the
soil, but it should reduce it considerably.

Dec. io, 1913

Foot-Rot of the Sweet Potato

273

SUMMARY

(1) The foot-rot has been hitherto unknown on the sweet potato
(. Ipomoea batatas). It is caused by the fungus Plenodomus destruens ,

(2) The organism is a very destructive wound parasite of the sweet
potato in the vicinity of the Dismal Swamp, Va., and occurs at Cape
Charles and Keller, Va.

(3) It kills the plant by the destruction of the cortex of the stem
near the ground.

(4) Pycnidia are abundantly formed on the diseased area of the stem
about the time the plant dies, or soon thereafter.

(5) The disease, while primarily found on the stem, invades the roots
and vines also.

(6) The fungus is cultivable on most artificial media, but gives the
highest development on com meal, rice, and stems of the sweet potato.

(7) The parasitism of the organism has been proved by numerous
inoculations of plants grown on the Potomac Flats and in the green¬
house.

(8) Successful infection experiments were carried out with reisolations
of the fungus from inoculated plants.

(9) The organism is parasitic on Ipomoea coccinea , but not on /. pur¬
purea and /. hederacea .

(10) Sweet potatoes from storage are decayed by the fungus when
inoculated under sterile conditions and kept moist in light.

(n) Light has no apparent effect on the production of fruiting bodies
in pure cultures of rice.

(12) The fungus makes its best growth, as measured by abundance
and rapidity of sporulation, in rice cultures at an average temperature of
about 2 1. 90 C.

(13) The fungus can live through the winter on dead vines of the
sweet potato.

(14) The disease is probably disseminated principally by means of
“seed roots” and the slips produced therefrom.

(15) Seed beds should be sterilized, and potatoes to be used for seed
should be carefully selected.

DESCRIPTION OF PLATES

Pi/ATJS XXIII. Parts of sweet-potato plants, showing the presence of pycnidia: A, On
the stem just above the ground; B, on the root.

XXIV. Portion of sweet-potato vines several feet from the hill, showing the

results of a natural infection of the foot-rot fungus. The organism
was recovered from these vines before being photographed.

XXV. Microscopic characters of the foot-rot fungus: A. Section through

a pycnidium on the root; B, section through a pycnidium on the
stem; C, hyphae, from artificial culture; D and F, chlamydospore-
like bodies found on the host and in some culture media; F,
pycnospores; G, club-shaped bodies often found in pycnidia; II,
germinating pycnospores.

XXVI. Two sweet-potato plants in pots, demonstrating the parasitism of the
foot-rot fungus: A, Inoculated; B, not inoculated.

XXVII. Nine-day-old cultures on synthetic agar: A, The conidial stage of
Diaporthe batatatis ; B, Plenodomus destruens.

XXVIII. Sweet potatoes inoculated with Plenodomus destruens: A, Inoculated
at the end; B, a section of A showing extent of rot; C, inoculated
at the side; D, section of C showing the extent of rot.

(274)

ADDITIONAL COPIES of this publication
-A may be procured from the Superintend¬
ent or Documents, Government Printing
Office, Washington, D. C., at 25 cents per copy

Subscription price per year - - - - 52.50

Plate XXII

Plate XXIV

Plate XXVI

Plate XXVII

JOURNAL OF AGRIOTJORAL RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. I Washington, D. C., January io, 1914 No. 4

ENVIRONMENTAL INFLUENCES ON THE PHYSICAL
AND CHEMICAL CHARACTERISTICS OF WHEAT

By J. A. LE ClERC, Chief , and P. A. Yoder, Assistant Chemist ,
Plant-Chemistry Laboratory , Bureau of Chemistry

INTRODUCTION

A former series of experiments 1 conducted in the Bureau of Chemistry
showed that neither the composition nor the physical characteristics of
wheat are to any great extent hereditary. The protein, gluten, and ash
contents, as well as the size of the berry, the weight of a bushel, and the
flintiness of the kernel, were found to be dependent upon the climatic
conditions prevailing during the growing period of the plant. Seed of
Kansas wheat containing 20 per cent of protein and showing 100 per cent
of flinty kernels and seed of California wheat containing 10 per cent of
protein with 13 per cent of flinty kernels when grown side by side in
South Dakota yielded crops of identical composition and physical appear¬
ance. The same was true of these Kansas and California seeds when
grown in California. The crops grown in California were, however,
entirely unlike those grown in South Dakota, owing to the great differ¬
ence in climatic conditions. It was shown in a most conclusive manner
that environment plays a major part in influencing both the chemical
composition and the physical appearance of a wheat crop. Cropping
through a number of generations under widely different environments
therefore does not alter permanently or make a noticeable impression
upon the transmissible physical and chemical properties of wheat.

Similar experiments, involving the transference of soil, are reported
by Shaw and Walters.2 In the main, their observations, based on crops
grown throughout a period of three years in one locality, harmonize with
the conclusions here presented, which are founded on the wider range of
experimental data now at hand, involving crops grown for four years on
three different types of soil in three different localities having widely

1 L,e Clerc, J. A., and Eeavitt, Sherman, Tri-local experiments on the influence of environment on the
composition of wheat. U. S. Dept. Agr., Bur. Chem. Bui. 128, 18 p., 1910.

2 Shaw, G. W., a^d Walters, E. H. A progress report upon soil and climatic factors influencing the
composition of wheat. Cal. Agr. Exp. Sta. Bui. 216, p. 549-574, 1911.

Journal of Agricultural Research, (27 5) Vol. I, No. 4

Dept, of Agriculture, Washington, D. C. Jan. io, 1914

E-i

17073—14 - 1

276

Journal of Agricultural Research

Vol. I, No. 4

varying climatic conditions. In some particulars, however, the conclu¬
sions which seemed justifiable from their experiments are not borne out
by these more extensive data.

The experiments discussed in this article were designed to study fur¬
ther the environmental influences and to show the r61e exerted by the
soil and the part played by climatic conditions, such as rainfall, sunshine,
humidity of the atmosphere, temperature, winds, and elevation above
sea level. As in the case of the previous experiments,1 they were car¬
ried on in cooperation with the Office of Cereal Investigations of the
Bureau of Plant Industry. The agricultural experiment stations of
Maryland, Kansas, and California cooperated by growing the crops.

CONDUCT OF THE EXPERIMENTS

In order to distinguish between the role played by soil and that by
environment other than soil, samples of soil were interchanged among
three localities, Maryland (College Park), Kansas (Hays), and California
(Davis) , which differ widely in climatic conditions. From each locality
sections of a normally fertile wheat-producing soil 5 feet square and 3 feet
deep were dug up in 3-inch layers, sacked, and replaced in the same
original position. To obviate any differences due to this manipulation
a portion of soil 5 feet square and 3 feet deep from each locality was like¬
wise dug up in 3 -inch layers, sacked, and stored until the soils from the
two other localities had arrived, when all three samples were placed in
their respective positions. A fourth plat of soil of the same size was
allowed to remain undisturbed in each locality to determine whether the
treatment to which the three other soils had been subjected would exert
any influence on the composition of the grain. Thus, there were 12
experimental plats, 4 in each locality, as shown in the following plan :

California:
Plat of
Plat of
Plat of
Plat of
Kansas :
Plat of
Plat of
Plat of
Plat of
Maryland :
Plat of
Plat of
Plat of
Plat of

TWELVE EXPERIMENTAL, PEATS

undisturbed California soil, or check plat.

disturbed California soil.] , , , . . , , , , , .

(Each taken up m 3-inch layers and replaced m

I original order.

Kansas soil.
Maryland soil.

undisturbed Kansas soil, or check plat.

disturbed Kansas soil.] — , A , . . - . , ,

California soil [Each taken up m 3-inch layers and replaced m orig-

Maryland soil.

inal order.

undisturbed Maryland soil, or check plat.

disturbed Maryland soil.] — , , , . . , , ' , , , .

California soil [Each taken up m 3 -inch layers and replaced m

T;r- ... [ original order.

Kansas soil. J b ■*

1 he Clerc and heavitt. Op. dt.

Jan. io, 1914

Environmental Influences on Wheat

277

During the first two years, 1908 and 1909, Crimean wheat obtained
from seed grown in Kansas was used on all 12 plats. As this variety
was not adapted to conditions prevailing in Maryland and California,
Turkey wheat was selected for 1910, 1911, and 1912. The change from
Crimean to Turkey wheat did not interfere, however, with the object of
the experiment, which was to determine the influence exerted by cli¬
matic conditions and soil on the composition of the crop.

The following determinations were made according to the methods
given in Bulletin 107, Revised, of the Bureau of Chemistry, entitled
“Official and Provisional Methods of Analysis/'

Water ; weight of 1 ,000 grains ; weight of a bushel ; flinty grains ; nitro¬
gen; alcohol-soluble nitrogen; fat; fiber; pentosans; sugars; ash; phos¬
phoric acid; and potash. The alcohol-soluble nitrogen was determined
by treating a certain quantity of ground wheat with a 70 per cent solu¬
tion of alcohol at ordinary temperature, with frequent shaking, for sev¬
eral hours, and then allowing the solution to stand overnight. An
aliquot part was taken and the nitrogen therein determined. The
amount of nitrogen thus obtained divided by the total quantity of
nitrogen in the sample gave the gliadin number.

TABULATION OF BATA

The data are collected in a number of tables. In Table I, first col¬
umn, is given the analysis of the original seed grown in Kansas in 1908,
which was used as seed on all the plats for the following year's crop.
The other analyses in tTable I and the data in Tables II to IV were
obtained on crops grown in 1909, 1910, 1911, and 1912, the results be¬
ing grouped by locality. The data from the different soil plats and the
check-soil plat in each locality are arranged in adjacent columns in
Table I. In Table II the same data, exclusive of check-plat data, are
rearranged, the results from the same soils being grouped in adjacent
columns. Averages derived from these data are given in Tables III, TVr
and V. In Table III are shown the averages of all the constituents
from crops grown in California, Kansas, and Maryland, not including the
check-soil plat, throughout the four years of the experiment. Table IV
gives the averages obtained from data on the crops grown on the soils of
California, Kansas, and Maryland for each of the three localities and for
all four years. Finally, in Table V are shown the averages for the undis¬
turbed or check-soil plats and for the corresponding plats in which the
soil had been taken up in 3-inch layers and replaced.

278

Journal of Agricultural Research

Vol. I, No. 4

Table I. — Composition of wheat grown on different plats of soil in California, in Kansas,
and in Maryland in iqoq, iqio, iqii, and IQI2.

CRIMEAN WHEAT.

Original seed and 1909 crop.

Determination.

Physical properties:

Water . per cent. .

Weight per 1,000 grains,

grams .

Weight per bushel, .pounds. .

Flinty grains . per cent. .

On water-free basis:

Nitrogen . do -

Protein (NX 5- 7) . do -

Alcohol-soluble nitrogen,

per cent .

Gliadin inprotein. .percent. .

Fat . do -

Fiber . do -

Pentosans . do -

Sugars . do —

Ash . . . do -

Phosphoric acid . do —

Potash . . . .do -

Phosphoric acid in ash,

percent .

Potash in ash . percent. .

9. 20

26.3
57- 7

2. 58
14- 75

1.03

8. 70
2. 52
2. 05
.96
•55

Wheat grown in
California on —

9.64

36* 2
62. 7
100

2. 59

14. 76

1. 23
46
1.

8.98

34*6

61. 5
100

2. 78
IS. 84

1. 16

41
1.82
2- 33
8.49

3. 21
1. 63

.68

• 45

42
28

9.00

36.4

61.5
75

2. 01
11. 46

.82
4i
1. 82

2.43

8. 16
3- 73
1. 63
. 70
.46

8. 88
25*4
97

2.03
n* 57

•7i
35
1. 84

2- 39
8. S3

3- 2 6
1. 90

.89
• 56

3°

Wheat grown u
Kansas 1 on —

vv titat gwwii J.JJL

Maryland on —

9. 56
21. 2

2. 69
15- 33

1. 10
41

2. 16
2. 69
8-37
2. 89
2- 39
I. 23

9. 48
23. o
80

2- 57
14.65

1.05

2. 05
2. 62
8. 31
2. 64
2. 30
1. 18
* 63

9. 22
22. 2

d S

TURKEY WHEAT.

1910 crop.

Determination.

Wheat grown in
California on —

Wheat grown in
Kansas on —

Wheat grown in
Maryland on —

California check
soil.

California soil.

Kansas soil.

Maryland soil.

California soil.

Kansas soil.

Kansas check
soil.

Maryland soil.

California soil.

Kansas soil.

Maryland soil.

Maryland check
soil.

Physical properties:

Water .

.per cent. .

9. 81

9. 68

9.67

8.99

9-39

9-03

9- 30

9. 12

9,00

10. 66

9- 73

Weight per 1,000 grains..

31- 2

28.3

34-3

21. 5

26. 1

22. 6

23-3

24. 0

28.0

3i-5

25*9

A . . * .

Weight per bushel . .

. .pounds. .

60. 5

61.8

58.3

56.9

57- 2

55-8

57* 7

Flinty grains .

.per cent. .

100

IOO

100

On water-free basis:

Nitrogen .

2. 16

2- 39

I. 86

2. 86

2. 80

3- 28

3-23

3- 12

1. 80

1. 90

2. 05

Protein (NXs-7) .

_ do _

12. 31

I3-63

10. 60

16. 28

15.98

18. 73

18. 41

17. 81

10. 27

10. 85

11. 68

Alcohol-soluble nitrogen.

... .do. . . .

.96

I. os

• 74

1.23

I. 44

I. 32

1. 29

• 75

Gliadin in protein .

.... do ... .

Fat .

2. 01

2. 13

2- 13

2. 11

1.86

2. 04

I. 81

2.02

1. 67

1. 76

1. 78

Fiber .

2. 26

2. 15

2. 28

2-35

2. 72

2. 79

2. 78

2. 80

2. 65

3* 01

2. 63

Pentosans .

8. 27

8. 32

8-57

9- 25

8. 64

8-93

8. 78

8. 64

8. 70

8- 54

8.84

Sugars .

3- 40

3- 53

3- 81

3- 43

3- 13

3-38

3* 11

3*33

2. 90

2. 99

3. 06

Ash .

_ do _

1.87

1. 84

1. 82

2. 05

1. 99

1. 97

2. 29

1.97

2. 09

2. 07

2. 22

Phosphoric add .

_ do _

.84

• 79

.86

1. 02

.85

.81

1. 08

. 80

1. 09

1. 21

Potash .

.... do _

. 60

.61

. 5s

• 65

. 61

. 66

. 69

. 64

. 57

. 61

Phosphoric acid in ash . . .

. do _

Potash in ash .

. do _

3°

1 Owing to a severe drought the crop failed to mature.

Jan. io, 1914

Environmental Influences on Wheat

279

Table I. — Composition of wheat grown on different plats of soil in California , in Kansas,
and in Maryland in 1909, 1910, 1911, and 1912 — Continued.

1911 crop.1

Determination.

Wheat grown in
California on —

ea-J

l'8

Wheat grown in
Kansas on —

Wheat grown in
Maryland on —

California soil.

Kansas soil.

Maryland soil.

Maryland check

soil.

8.83

8.97

8-93

8- 73

27.4

29.4

27. 1

26.4

60. $

62. 2

59*9

2,00

2. 20

2*37

2.31

11. 38

12.52

13*52

13. 18

•85

.88

1. 04

.96

2. 04

2. 05

1.83

1.87

2.41

2- 33

2.49

2.44

8.08

8. 22

8. 25

8-39

3*25

3*45

3*33

3-34

2. 23

2. 20

2. 10

2. 17

1.24

1. 16

1.09

r.17

.67

•65

.67

5<5

Physical properties:

Water . per cent, .

Weight per 1,000 grains _ grams. .

Weight per bushel . pounds. .

Flinty grains . per cent. „

On water-free basis:

Nitrogen . do _

Protein (NX 5. 7) . do _

Alcohol-soluble nitrogen . do _

Gliadin in protein . do _

Fat . do _

Fiber . do _

Pentosans . do _

Sugars . do _

Ash . do _

Phosphoric acid . do _

Potash . do _

Phosphoric add in ash . do _

Potash in ash . do _

38.4

34-6

37* 7
46

9.00
23-5 12-9

9* 30
13*3

8.28

[2.5

100 100

3*70
21. 11

4.09

23*31

4* 07
23. 18

1.94

2*95

8.84

1-95

2.94

9. 12

1.83

3*17

9*57

‘2. 58
1. 14
.69

2. 56
1. 18

2. 78

1912 crop.

Physical properties:

Water .

.per cent. .

8*43

8. 29

8. 67

8. 70

io* 55

10. 18

IO.3O

10* 45

10. 22

9* 65

10. 13

10. 17

Weight per 1,000 grains ,

. . .grams. .

29*3

30.5

31*8

23*9

22. 0

21.4

28. 6

16. 2

25*7

19. 19

22.4

Weight per bushel . .

. . pounds . .

64. 3

65. 1

57* 7

60. 1

60. 3

Flinty grains .

.per cent. .

98'”

On water-free basis:

Nitrogen .

2.05

2. 24

2.40

3*17

2.68

2. 78

3.62

3* 29

1.85

2. 12

2. 28

2.05

Protein (NX 5. 7) .

_ do _

11. 68

12. 77

13.68

18.07

15* 29

15*87

20. 65

18. 78

10. 54

12. 11

11.68

Alcohol-soluble nitrogen .

_ do _

.85

•97

1. 02

1.40

1. 14

1. 20

I.64

i*35

• 70

*S5

.89

•83

Gliadin in protein .

. do _

Fat .

_ do _

1. 89

i* 93

1.88

2- 05

1.88

2. 17

1.82

2. 01

1. 89

1. 97

1.88

Fiber. . . .

_ do _

2. 28

2. 20

2. 28

2. 69

2.68

2- 65

2. 48

2. 24

2. 72

2- 53

2.88

2. 87

Pentosans .

_ do _

8.08

7*95

8. 05

8. 70

8.31

8. 27

8. 52

8. 98

8.44

8. 62

9*36

9*03

■ Sugars .

, . . . . do ....

3. 56

3. 78

3. 80

3. 91

3. 37

3. 41

3. 12

2. 95

3. 08

3. 32

Ash .

2.07

2.14

2. 18

2. 20

2.47

2. 20

2*45

2.83

2. 24

2.48

2. 46

Phosphoric add .

_ do _

1. 02

1. 07

1.09

I.03

1.23

1.02

1. 26

I*3l

1. 14

1. 19

1. 27

1.24

Potash .

_ do _

. 62

.62

. 62

*59

■ 74

.67

.66

* 79

* 71

• 72

.80

• 79

Phosphoric acid in ash. . ,

Potash in ash .

1 The data for the 1911 samples grown in California were furnished by Prof. Shaw, of the Universityof
California, under whose supervision the field work in that State was conducted.

s8o

Journal of Agricultural Research

Vol. I, No. 4

Tabi,E II. — Composition of wheat grown on plats of California, Kansas, and Mary¬
land soils in California, in Kansas, and in Maryland .

1909.

Analysis of wheat grown on —

California soil in —

Kansas soil in —

Maryland soil in —

Determination.

>»

Physical properties:

Water . per cent. .

Weight per 1,000 grains _ grams. .

Weight per bushel . pounds. .

Flinty grains . per cent. .

On water-free basis:

Nitrogen . do _

Protein (N X 5-7) . do _

Alcohol-soluble nitrogen _ do _

Gliadin in protein . do _

Fat . do _

Fiber . do _

Pentosans . do _

Sugars . do _

Ash . do _

Phosphoric acid . do _

Potash . do _

Phosphoric acid in ash . do _

Potash in ash . do. . . .

8.98
34-6
61. s
100

2. 78
15.84
1. 16

1.82

2- 33
8.49

3- 21
1.63

v .68

•45

9- 56
21. 2

9. 00

36.4

61.5
75

9. 48 8. 88

23.0 25.4

9. 22
22. 2

2. 69
15- 33

1. 10
41

2. 16
2. 69
8-37
2. 89
2- 39
1. 23

2. or
11. 46
.82
4i
1. 82

2-43
8. 16
3* 73

1.63

• 70
.46

2. 57
14. 65

1. 05
4i

2. 05
2. 62

8.31

2. 64
2. 30
1. 18
•63

2.03
«• 57
. 71

2. 34
13*34
.92

35
1. 84
2-39
8- 53
3. 26
1. 90

40
2. 15
2. 59
9-03
2. 82
2.09

.89
• 56

1910.

Physical properties:

Water . per cent.

Weight per 1,000 grains _ grams.

Weight per bushel . pounds.

Flinty grains . . per cent .

On water-free basis:

Nitrogen . do. . .

Protein (NX 5.7) . do. . .

Alcohol-soluble nitrogen . do . . .

Gliadin in protein . do. . .

Fat . 4.do. . .

Fiber . do. . .

Pentosans . do. . .

Sugars . do. . .

Ash . do. . .

Phosphoric acid . do. . .

Potash . do. . .

Phosphoric acid in ash . do. . .

Potash in ash . do. . .

9.00

9- 39

9.00

9.67

9-03

10. 66

8. 99

9. 12

9* 73

28.3

26. 1

28.0

34* 3

22. 6

31-5

21.5

24.O

25.9

58.3

61. 8

56. 9

57* 7

55. 8

100

100 '

100

2. 39

2. 80

1. 80

1.86

3-28

1.90

2.86

3- 12

2.05

13-63

15.98

10. 27

10. 60

18. 73

10. 85

16. 28

17- 81

11.63

I. 05

1. 23

• 74

1. 44

• 75

1. 20

2. 13

1.86

I- 67

2.13

2.04

1. 76

2. 11

2.02

1.78

2. 15

2. 72

2. 65

2. 28

2.79

3.01

2* 35

2. 80

2. 65

8.32

8.64

8. 70

8*57

8. 93

8- 54

9*25

8. 64

8.84

3*53

3* 13

2. 90

3*81

3.38

2-99

3- 43

3-33

3.06

1. 84

1.99

2, 09

1.82

1.97

2, 07

2.05

1.97

2. 22

• 79

.85

.86

.81

1. 09

1. 02

.80

1. 21

. 61

•55

.66

• 57

.65

.64

. 61

1911.

Physical properties:

Water . per cent. .

Weight per 1,000 grains _ grams. .

Weight per bushel . pounds. .

34-6

Flinty grains . per cent . .

On water-free basis:

Nitrogen . do _

Protein (NX 5.7) . do _

Alcohol-soluble nitrogen . do _

Gliadin in protein . do _

Fat . do _

10. 56
.76
43

Fiber . do , .

Pentosans . do _

Sugars . do _

Ash . do _

1.88

Phoshoric acid . do _

Potash . do _

Phosphoric acid in ash . do ... .

Potash in ash . do _

9.00

12.9

100

3- 70

1*94

3-95

8.84

2. 58

I. 14

8.83

27.4

60.5
25

2. OO

11.38

.85

43
2.04
2.41
8. 08
3.25

2. 23
1. 24
.67
56
30

37-7

9.61

.70

1. 76

9- 30
13-3

8.97

29.4

62. 2

23- 5

8* 79
13. 8

8.93

27. 1

100

3*97

2-37

12. 52

13. 20

22. 62

13-52

.88

.80

1.04

2.05

2. 12

1.83

2-33

3-21

2-49

8. 22

9. 12

8. 25

3.45

3* 33

2. 20

1.78

2. 09

2. 10

1. 16

.86

I.09

•65

.64

.65

Jan. io, 1914

Environmental Influences on Wheat

281

Table II. — Composition of wheat grown on plats of California, Kansas, and Mary¬
land soils in California, in Kansas, and in Maryland — Continued.

1912.

Analysis of wheat grown on —

Determination.

California soil in —

Kansas soil in —

Maryland soil in —

California.

Kansas.

Maryland.

j California.

Kansas.

Maryland.

California.

Kansas.

Maryland.

Physical properties:

Water .

.per cent. .

8. 29

io- 55

10. 22

8.67

10. 18

9- 65

8. 70

10. 45

10. 13

Weight per 1,000 grains .

. . .grams. .

30-5

22. 0

25- 7

31-8

21. 4

25- 7

23-9

16. 2

19.9

Weight per bushel .

.pounds. ,

65. 1

57. 7

Flinty grains .

.per cent. .

98 “

100

75 ~

On water-free basis:

Nitrogen .

2. 24

2.68

1.85

2.40

2. 78

2. 12

3- 17

‘3-29

2. 28

Protein (N X 5-7) .

_ do _

12.77

15.29

10.54

13-68

15-87

12. 11

18.07

18. 78

Alcohol-soluble nitrogen

_ do _

■97

1. 14

• 70

1. 02

1. 20

.85

1. 40

1-35

Gliadin in protein .

Fat .

....do....

1-93

1.88

1. 89

1.88

2.17

1.97

2. 05

2. 01

1.88

Fiber .

_ do _

2. 20

2.68

2. 72

2. 28

2.65

2- 53

2. 69

3-24

2.88

Pentosans .

_ do _

7* 95

8.31

8.44

8. 05

8. 27

8.62

8- 70

8. 98

9-36

Sugars .

. . . .do. . . .

3* 78

3. 80

3. 08

Ash .

_ do _

2. 14

2.47

2. 24

2. 18

2. 20

2. 24

2. 20

2.83

2.48

Phosphoric add .

_ do _

1.07

1.23

1. 14

1. 09

1. 02

1. 19

1-03

i- 3i

1. 27

Potash .

. . . .do. , . .

• 63

. 74

, 71

. 62

• 67

. 72

* 59

* 7°

.80

Phosphoric acid in ash.

_ do _

Potash in ash .

TablB III. — Averages and extremes in wheat grown on plats of California , Kansas, and
Maryland soils in California, in Kansas, and in Maryland.1

California.

Kansas.

Maryland.

Determination.

Averages.

Extremes.

Averages.

Extremes.

Averages.

Extremes.

Mean.

Divergence
from mean.

1 Minimum.

Maximum.

Mean.

Divergence
from mean.

Minimum.

Maximum.

Mean.

Divergence
from mean.

Minimum.

10. 66

Physical properties:

Water . per cent. .

8. 98

0.31:

8. 29

9.68

9- 53

o- 57

8. 79

io- 55

9- 53

0. 46

8.83

W eight per 1 ,000 grains,
grams .

30- 2

4-5

21. 5

37- 7

19. 1

4-5

12. 9

26. 1

25. 6

2. 7

19.9

3i-5

Weight per bushel,
pounds .

62.8

i-5

61. 5

65. 1

57-2

55-8

58.3

60. 1

57-7

62. 2

Flinty grains . per cent . .

100

On water-free basis:

Nitrogen . per cent. .

2. 42

•35

1. 86

3- 17

3* 30

.41

2. 68

4.09

2. 18

•23

1. 80

2. 69

Protein (N X 5-7).
per cent .

13* n

2. 01

9. 61

18. 07

18.83

2-34

15.29

23-31

12. 43

1. 29

IQ. 27

15*33

Alcohol-soluble nitro¬
gen . per cent. .

.93

.18

• 70

1. 40

1. 27

.09

1. 14

i*44

.90

. 10

.70

1. 10

Gliadin in protein,
per cent .

Fat . per cent. .

1. 97

. 12

1. 82

2. 13

.08

1. 86

2. 17

1. 94

. 14

I. 67

2. 16

Fiber . do _

2- 34

. 11

2. 15

2. 69

2. 89

.18

2. 65

3- 21

2. 63

- 13

2. 33

3. 01

Pentosans . do _

8-45

. 22

7- 95

9- 25

8.76

. 26

8. 27

9. 12

8.56

.29

8. 08

9- 36

Sugars . do _

3- 61

. 22

3- 21

3- 91

3-3 2

. 08

3- 13

3-4i

3- 03

. 18

2. 64

3- 45

Ash . do -

1. 90

. 16

1. 63

2. 20

2. 30

.28

1.97

2. 83

2. 22

.09

2. 07

2. 48

Phosphoric acid . . do -

. 90

• 13

.68

I. 09

1. 02

•17

. 80

1. 31

1. 18

•05

1.09

1. 27

Potash . ,do _

• 57

. 06

•45

• 65

.68

.04

. 61

■ 79

.67

•05

• 57

.80

Phosphoric acid in ash,
per cent .

Potash in ash, per cent . .

1 Not including check plats.

282

Journal of Agricultural Research

Vol. I, No. 4

Table IV— Averages and extremes in wheat grown in California , in Kansas , and in
Maryland on plats of California, Kansas, and Maryland soils.1

Determination.

Physical properties:

Water . per cent. .

Weigh tperi, ooo grains ,

grams .

Weight per bushel,

pounds .

Flinty grains. per cent.

On water-free basis:

N itrogen . per cent , .

Protein (N X 5-7),

per cent .

Alcohol-soluble nitro¬
gen. - .per cent. .

Gliadin in protein,

percent .

Fat . per cent. ,

Fiber . do _

Pentosans . do _

Sugars . do _

Ash . do _

Phosphoric acid . . do _

Potash . do _

Phosphoric acid in ash,

per cent .

Potash in ash. per cent . .

1 California soil.

Kansas soil.

Maryland soil.

Averages.

Extremes.

Averages.

Extremes.

Averages.

Extremes.

0 a

-•

0 d

•

s I

EfS

.fc

9-35

o* 53

8. 29

10. 22

9. 46

0.47

8. 67

10. 66

9. 29

0. 48

8. 70

10.45

26. 5

4-5

12. 9

34- 6

27.9

6. 1

13-3

37-7

22. 1

3-i

13- 8

27. 1

60. 9

i-5

58.3

64-3

60. 4

2. 2

56.9

65. 1

(2>

(2)

100

2. 48

•44

1. 80

3-7®

2. 52

• 53

1. 86

4.09

2. 75

•53

2.03

3-97

13-88

2. 56

10. 27

21. 11

13- 94

3*os

9* 61

23-31

15-44

2.97

«• 57

22. 62

. 16

• 70

I- 23

•94

.18

■70

1.44

1. 05

. 22

• 7i

1. 40

1-93

. 11

1. 67

2. 16

1. 98

. 11

1. 76

2. 17

1.97

. 12

1. 78

2. 15

2- 55

. 22

2. is

2- 95

2- 59

. 22

2. 28

3- 01

2- 73

. 24

2-35

3- 24

8- 41

. 21

7-95

8. 84

8.48

.28

8. 05

9. 12

8. 87

.28

8. 25

9* 36

3- 33

. 28

2. 89

3-90

3- 48

.24

2. 99

3- 81

3- 30

. 22

2. 82

3- 9i

2- 13

.23

i- 63

2. 58

2. 08

. 11

1. 63

2. 56

2. 16

. 20

1. 78

2. 83

I. 04

. 18

. 68

1. 24

1.03

•15

• 7o

I. 19

I- 05

• 15

. 80

i- 31

.64

. 06

■45

• 74

. 61

.07

.46

• 72

.66

. 06

• 56

.80

1 Not including check plats.

Only i sample.

Table V. — Averages and extremes for the years igog, igio, igu , and igi2 in wheat on
disturbed and undisturbed plats 1 for all localities ( California , Kansas, and Maryland)
and years.

Determination.

Physical properties

Water . per cent. .

Weight per i,ooo grains . .grams. .

Weight per bushel . pounds. .

Flinty grains . per cent. ,

On water-free basis:

Nitrogen . per cent. .

Protein (N X 5.7) . do. . .

Alcoholic- soluble nitrogen,

per cent

Gliadin in protein . .

• , .per cent. .

Fat .

Fiber .

Pentosans .

Sugars .

Ash . .

Phosphoric acid . . . do _

Potash . . do _

Phosphoric acid in ash . do _

Potash in ash . do _

Disturbed.

Undisturbed.

Averages.

Extremes.

Averages.

Extremes.

Diver¬

Mean.

gence

Mini¬

Maxi¬

gence

Mini¬

Maxi¬

from

mum.

from

mum.

mean.

9-3i

0.51

S. 29

10. 18

9- 33

0. 65

8.28

10. 30

25.8 ,

.49

13 -3

34 6

27. 6

5* 7

12. 5

38. 4

Only two samples.

100

2. 78

•45

2. 24

409

2. 76

.66

2. 05

4.07

^5* 25

2.84

10. 56

23-31

15-32

3- 62

11. 68

23. 18

1. 06

•IS

• 76

1.44

1.09

•23

.83

1. 64

I. 97

. 11

1.82

! 2. 17

1. 86

.08

1. 67

2. 01

2. 55

. 26

2. 15

2. 94

2. 56

.29

2. 18

3- 17

8- 59

.41

7- 95

9*36

8. 61

•35

8. 08

9- 57

3-44

• 14

3- 21

3- 78

3- 34

• 14

3- 11

3- 56

2. 09

•23

1.63

2. 56

2. 16

•30

1. 60

2. 78

.96

• 17

.68

1. 27

I. 06

*15

• 79

I. 26

.64

. 06

•45

. 80

.64

.07

.48

■ 79

45 |

1 Only data that are strictly comparable are used. Disturbed-plat data are used only if the determina¬
tions for the correspond mg check plats were also made, and vice versa.

Jan. io, 1914

Environmental Influences on Wheat

283

PHYSICAL CHARACTERISTICS
WEIGHT OF 1,000 GRAINS OF WHEAT

In California the grains were almost uniformly plump and heavy, not
varying far from 30 grams for each thousand, except in the case of the
samples grown on the soil obtained from Maryland. In Kansas they
were less plump, 1,000 grains weighing about 23 grams in 1910, 13 grams
in 1 91 1, and 20 grams in 1912. In Maryland the weight of 1,000 grains
was quite uniform throughout the series of four years. As a rule, the size
of the grains in each locality for each year was uniform, irrespective of
the type of soil in which they grew. There were, however, a few notable
exceptions to this rule: The grain grown on Maryland soil in each year
from 1909 to 1912 in California, as well as that grown on the Maryland
soil in 1912 in Kansas, was decidedly lighter in weight than that grown
in the same locality on the other soils. This would seem to indicate that
some soils play an important part in influencing the size of the grain.

Between the localities there was usually a much greater difference in
the weight of 1 ,000 grains than was noted between the soils. (See Table II.)
The weight of 1,000 grains, then, is distinctly dependent, as a rule, on
climatic or seasonal conditions rather than on soil characteristics. The
fact that environment plays the chief r61e in influencing the weight is
again brought out in the tables of averages, which show a great difference,
for example, 30.2, 19. 1, and 25.6 grams for California, Kansas, and
Maryland, respectively, when averaged by localities (see Table III), and
a relative uniformity of 26.5, 27.9, and 22.1 grams, respectively, when
averaged by source of soil (see Table IV).

Table I shows that in about 80 per cent of the samples investigated
the weight of 1,000 grains of seed grown on different soils in any one
locality was sufficiently uniform to permit the conclusion that climate
and not soil is the chief factor affecting the size of the grain. From
Table III it is seen that the California-grown samples averaged the
heaviest and the Kansas-grown samples the lightest.

WEIGHT OF ONE BUSHEL OF WHEAT

The weight of a bushel of wheat runs more or less parallel with the
weight of 1,000 grains. If the samples weighing over 61 pounds to the
bushel are compared with those weighing less than 60 pounds, it will be
found that the weight of 1 ,000 grains of the former was, on an average,
33.4 grams, and that of the latter, 25 grams. In many cases, owing to
the small amount of material, it was impossible to make a weight-by¬
bushel determination.

FLINTY GRAINS

Classifying the grains of each sample into those which were wholly
dark or flinty and those which appeared to be light brown or mealy, a
remarkable uniformity is found in the groups arranged by locality in

284

Journal of Agricultural Research

Vol. I, No. 4

which they grew (see Table I) and a dissimilarity in groups arranged by
the source of soil (see Table II) . The averages by localities (see Table
III) differ greatly, being 86, 99, and 35 per cent for California, Kansas,
and Maryland, respectively. The averages by soils are very uniform,
being 71, 69, and 85 per cent, respectively. (See Table IV.)

These averages do not show the great variations actually found in the
different regions in any one year, for seasonal variations of the individual
localities tend to equalize the averages. In Table I, for example, while
the samples grown in California and Kansas in 1910 in each of the three
soils were for the most part flinty, those grown in the three soils in
Maryland were all more or less starchy or mealy. Similar figures are
noted in 1911, when the Kansas samples grown on all three soils yielded
wheat which was practically 100 per cent flinty, while on the same soils
in Maryland the percentage of flinty kernels was less than half as great.

CHEMICAL CONSTITUENTS

In considering the composition of the wheat it will be. seen that many
of the organic and inorganic constituents undergo as great variations as
have already been noted with respect to the physical characteristics.
On the other hand, there are a number of these constituents which
showed very little variation, or no regularity in such variations as exist.
Among those showing but little variation may be mentioned the gliadin
number and the potash in the ash, and among those showing no pro¬
nounced regularity in the variations are the fat, fiber, pentosans, and
sugars. With those exhibiting variations of a regular character belong
particularly the nitrogen and protein, the ash, the phosphoric acid, and
the phosphoric acid in the ash.

PROTEIN

As the protein of wheat is its most important constituent, it will be
of more than usual interest to note the changes produced by difference
of soil and by change of environment. As a rule, there was a remarkable
uniformity each year among the samples grown in any one locality,
independent of the soil upon which they grew. Thus, in 1910, 1911, and
1912 the protein in wheat grown in California was almost uniformly
low, about 13 per cent; in Maryland it was also low, about 11 per cent;
while in Kansas it was high, nearly 18 per cent. This fact is more
clearly brought out in Table III, which shows the average protein con¬
tent to be 13.11, 18.83, and 12.43 per cent for California, Kansas, and
Maryland, respectively.

In Table IV, where the results are arranged according to source of
soil, it will be seen that the wheats grown on California soil in all three
localities had an average protein content of 13.88 per cent, those grown
on Kansas soil, 13.94 Per cent, and those on Maryland soil, 15.44 Per cent.
This shows a rather striking uniformity and again emphasizes the rela-

Jan. io, 1914

Environmental Influences on Wheat

285

tively small r61e played by the soil in influencing the protein content of
wheat. There was a greater similarity between the protein contents of
the samples grown in Maryland and California, both relatively humid
regions, than between the protein contents of samples from either of
these localities and those from Kansas, which has a comparatively dry
climate.

There are a few exceptions, however, to the rule that soil influences the
composition of wheat to only a slight degree. Among the most striking
of these were the protein results obtained in 1909 in California on Cali¬
fornia soil, in 1910 and 1912 in California on Maryland soil, as well as in
Kansas on Kansas check soil; that is, 4 out of 42 cases did not follow
the general rule. Since about 90 per cent of the results obtained followed
the general rule, and the exceptions noted were in different localities
and on different soils and not always on the same soil in any locality, it
is probably safe to assume that the contrary results given by the other 10
per cent of samples were accidental. These few exceptions among the
prevailing regularities may serve to emphasize the fact, too frequently
overlooked in plat experiments of this kind where many factors may affect
the results, that a regularity needs to be traced through a great number
of individual instances before it is safe to draw conclusions from it.
Thus, in this experiment a consideration of the data from the 1909 crop
alone might show that the soil has a marked determining influence upon
the protein content and that the California soil tends to produce a wheat
of relatively high protein content. That such a conclusion would be
erroneous is evidenced by practically all the data of the three following
years, for in no other case during 1910, 1911, and 1912 was there a larger
amount of protein in wheat grown on the California soil than in that
grown on the two other soils. In fact, those wheats were invariably
lower in protein content.

While these exceptions may be considered as purely accidental, the
following question is suggested by such variations from the rule : Is there
in the physical, chemical, or biological characteristics of the soil a real
difference which at first exerts a determining influence on the composi¬
tion of the crop, but which may be obliterated in the course of a year or
two after putting the soil down in a different locality? Some weight is
lent to such a hypothesis by the fact that the slight differences in protein
content in the crops grown in Maryland the first year after the exchange
of soils were much the same as the exceptionally great differences in the
crops grown in California. Unfortunately, the Kansas crop was a com¬
plete failure, and it is impossible, therefore, to know in what way the
soil there would have influenced the composition of the crop during the
first year. To answer this question, more observations during the first
few years of similar soil exchange experiments would be necessary, using
larger plats to partly eliminate any tendency for soils to equalize after
being together in one locality, if such a tendency does exist.

286

Journal of Agricultural Research

Vol. I, No. 4

It seems justifiable to conclude that climate is the principal factor
influencing the protein content of wheat, and that soils, when used as in
this experiment, have little or no influence.

GUADIN IN PROTEIN

With very few exceptions, the amount of alcohol-soluble nitrogen or
gliadin bore a close relation to that of total nitrogen. The percentage
of gliadin in the wheat grown on the different soils in the three localities
during the years 1909 to 1912 remained practically constant at 41 per
cent, except in the case of wheat grown on Maryland soil and on Cali¬
fornia check soil in California in 1909, and on Maryland soil in California
in 1 91 1. These 3 exceptions out of 42 samples can not be explained and
must be assumed to be accidental. From Table II it would seem that
those samples grown on Maryland soil in California in 1909, 1911, and
1912 and in Maryland in 1912 formed exceptions to the rule. When
general averages are considered, however, practically no differences in
gliadin number due either to difference of soils or to change of seasonal
conditions are noted. Table III gives the average gliadin numbers of
the samples grown on each of the three soils in California as 41 ; in Kan¬
sas, 42; and in Maryland, 40. Table IV shows the gliadin number of
the wheats grown on California soil in each of the three localities to be
42 ; on Kansas soil, 41 ; and on Maryland soil, 40. There seems to be a
slight tendency for the Maryland soil to be low in gliadin. The differ¬
ences are, however, small and probably no weight should be given them.

PAT

The amounts of fat agreed very closely in the case of wheat grown on
the different soils in any one locality, only 3 out of 42 samples showing a
greater variation than 0.2 per cent, which may be assumed to be the
limit of error for fat determinations. When averaged by locality, the
results were 1. 97, «2 .00, and 1.94 per cent for wheat grown in California,
Kansas, and Maryland, respectively. When averaged by source of soils,
the results were 1.93, 1.98, and 1.97 per cent for samples grown on Cali¬
fornia, Kansas, and Maryland soils, respectively. The results taken as a
whole indicate that fat is not affected to any great extent by climatic
or soil conditions.

FIBER

The fiber showed a somewhat greater variation in amount than did
the fat. The results as a whole indicate that a greater influence is
exerted by seasonal or climatic changes than by differences in soils.
This is shown in Table III, with the average fiber content of 2.34, 2.89,
and 2.63 per cent in the wheats grown on the three soils in California,
Kansas, and Maryland, respectively.

The wheat grown in the three localities on California soil gave 2.55
per cent of fiber, on Kansas soil, 2.59, and on Maryland soil, 2.73. (See

Jan. io, 1914

Environmental Influences on Wheat

287

Table IV.) These averages agree with one another more closely than
do those in Table III, proving that soils play a minor role in influencing
the fiber content.

PENTOSANS

The pentosan content followed generally the fiber content, being high
where the fiber content was high and low where the fiber content was low.

SUGARS

The sugar content of the samples grown in California was somewhat
higher than that of those grown in Kansas or in Maryland.

ASH

If soil itself has any influence on the composition of the wheat, it is
reasonable to expect that the mineral constituents especially will be thus
influenced. Even here, however, in the case of ash, the soil factor is a
minor or negligible one. There was a decided regularity in the ash
content, and, like the physical properties and the protein content, this
regularity consisted in an approximately uniform ash content of the
samples grown during any one year in any one locality. Thus, during
each of the four years California produced from all soils crops with a low
ash content of about 1.9 per cent, while Kansas produced crops rela¬
tively higher in ash, averaging 2.30 per cent, and Maryland nearly as
high, Varying somewhat, however, from year to year, with an average of
2.22 per cent. The average ash content of all crops grown on each of
the three soils, irrespective of the locality, showed but slight variation,
being 2.13, 2.08, and 2.16 per cent for California, Kansas, and Maryland
soils, respectively.

PHOSPHORIC-ACID CONTENT OP THE WHOLE WHEAT AND OF THE ASH

In most cases the amount of phosphoric acid rose or fell in the same
proportion as the ash, so that the percentage of phosphoric acid in the
ash remained practically constant, averaging 47 per cent for California,
45 per cent for Kansas, and varying from 41 to 51 per cent in these two
localities. The crops grown in Maryland, however, on all soils had a
strangely high amount of phosphoric acid, averaging 53 per cent of the
ash and varying from 51 to 59 per cent. There is no explanation for
the fact that in Maryland all the soils used in this experiment supplied
to the grain mineral constituents with a percentage of phosphoric acid
much higher than that supplied by the same soils in California and in
Kansas. It was apparently due to some climatic or seasonal conditions
prevailing in Maryland. The kind of soil did not, however, affect the
amount of phosphoric acid in the wheat or in the ash, for Table IV shows
that the average in the wheat grown in the three localities on plats of
California soil was 1.04 per cent, on plats of Kansas soil, 1.03 per cent,
and on plats of Maryland soil, 1.05 per cent, and the phosphoric acid in
the ash was 48 per cent in each case.

288

Journal of Agricultural Research

Vol. I, No. 4

POTASH CONTENT OP THE WHOLE WHEAT AND OP THE ASH

The potash in the wheat, like the total ash, was seemingly influenced
more by climatic and seasonal variations than by the soil, so that the
amount of potash in all samples rose or fell in practically the same pro¬
portion as the amount of total ash, and the percentage of potash in the
ash— about 30 per cent — remained very nearly constant for all localities,
soils, and seasons included in the experiment. This is further shown by
the similarity of the averages, whether by locality (see Table III), with
averages of 29, 30, and 30 per cent for California, Kansas, and Maryland,
respectively, or by soils (see Table IV) with averages of 30, 29, and 29
per cent, respectively.

correlation between physicae properties and chemicae

CONSTITUENTS

Although the relationship or interdependence between the physical
properties and chemical constituents does not show in these results as
markedly as might be expected, such relationships may be distinctly
traced in some of the constituents. Thus, as has often been pointed out
by others, a distinct correlation exists between the protein content and
the physical appearance or between the protein content and the weight
of 1,000 grains, high protein being more or less parallel with flintiness
and with lightness of grains. The table of averages (see Table III) shows
that the Kansas samples, containing 18.83 Per cent of protein, averaged
99 per cent of flinty grains and weighed at the rate of 19.1 grams for
1,000 grains, while the Maryland samples, containing 12.43 per cent of
protein, averaged but 35 per cent of flinty kernels and weighed 25.6 grams
for 1,000 grains, and the California samples, containing 13.11 per cent of
protein, averaged 86 per cent of flinty grains and weighed as high as 30.2
grams for 1 ,000 grains. The results in Table IV show a similar tendency
in these respects, the samples grown on Maryland soils in the three locali¬
ties being somewhat richer in protein and having at the same time more
flinty kernels and weighing less for each i;ooo grains than the samples
grown on California or Kansas soils. The differences in this case, how¬
ever, were very much less notable than those due to climatic variations.
(See Table III.) There was a less noticeable parallelism between the
fiber and pentosans, a high fiber content, as a rule, being accompanied by
a high pentosan content, and vice versa. The California-grown samples,
which were the heaviest, contained the smallest amount of fiber and pen¬
tosans, while the Kansas samples, which were the lightest, contained the
greatest amount.

The fact that the ash and protein contents were low in the California-
grown samples and high in the Kansas-grown samples might lead one to
expect that the ash was a function of the protein content. This is not
borne out by an examination of Table III, where it is seen that the ash

Jan. io, 1914

Environmental Influences on Wheat

289

of the samples grown in Maryland was appreciably higher than that of
the samples grown in California, while the protein of the former was less
than that of the latter. On the other hand, the ash content of the
Kansas samples was only slightly higher than that of the Maryland-grown
samples, although the protein content of the former was 50 per cent
higher than that of the latter.

COMPARISON BETWEEN RESULTS FROM DISTURBED AND UNDISTURBED

PLATS OF THE SAME SOIL

Attention has thus far been directed primarily to the composition of
the wheat samples grown for several years in each locality on each of the
three soil plats which had been taken up in 3-inch layers and interchanged
among the three localities. As previously mentioned, a check plat
of equal size, in which the soil had not been disturbed, was planted each
year in each locality, and samples from it were analyzed for comparison.
A fear that manipulation of the soil would produce abnormal conditions,
influencing the character of the crop, was not justified by these results
(Table V), at least not as evidenced by the physical appearance and the
chemical composition. The slight differences between the crops from the
disturbed and undisturbed plats of the same soil are apparently either
accidental or due to errors in sampling or in analysis. This is further borne
out by the results from both the seed-exchange experiments 1 and from
the soil-exchange experiments (pp. 278-28 1 ) . It is simply a verification of
the conclusion already drawn, that the soil factor plays but a very sub¬
ordinate part or is entirely devoid of influence in determining these
characteristics in the crop.

Such great differences exist in respect to one constituent, however,
that they must be classed as exceptions to the rule. The percentage of
phosphoric acid averaged 0.96 per cent in the samples from disturbed
plats and 1.06 per cent in those from undisturbed plats, or, if expressed
as the percentage of phosphoric acid in the ash, it is 46 and 49 per cent,
respectively. It might seem that the undisturbed soil could give a little
more phosphoric acid to the grain than the disturbed soil. These dif¬
ferences, being only slightly greater than the limit of error in analytical
work, probably have no significance.

CONCLUSIONS

As is to be expected in plat work in the field, especially with such small
plats as were used for these experiments, there are many variations in the re¬
sults which seem accidental, in that they can not be interpreted according
to any definite law. There are, however, certain variations which appear
with such regularity that important conclusions may be drawn from them.

An inspection of the tables should show whether climatic conditions
or soil characteristics have a strong determining influence upon the

1 Le Clerc and Leavitt. Op. cit.

29°

Journal of Agricultural Research

Vol. I. No. 4

properties or composition of the crop. If the adjacent data in Table I
under each locality are similar and distinctly unlike the corresponding
group in another region, it is evident that the locality — that is, the cli¬
mate — has exerted a strong influence. Likewise, if a similarity exists in the
data in the adjacent columns in Table II as regards crops from the same
soil and there is a distinct difference between them and the corresponding
data from other soils, it is clear that the soils in themselves have a deter¬
mining influence, regardless of the locality in which the soils happen to be.

To avoid erroneous conclusions concerning any property or constituent,
due to accidental differences occurring in individual groups of data, it is
necessary to make a survey of all the data on hand regarding that prop¬
erty or constituent. In a measure the averages drawn from the several
groups of data furnish quantitative values which may indicate the per¬
sistence or the nonpersistence of such differences. The average diver¬
gences from these means, together with the minima and the maxima,
supply further quantitative evidence along this line. Such averages
and the corresponding minima and maxima are brought together in
Tables III, IV, and V.

This experiment, covering a period of four years, in which three fairly
good wheat soils, one each from California, Kansas, and Maryland, were
put down side by side in each of these three localities and cropped with
the same variety of wheat, shows that the soil does not exert the chief
or preponderating influence in determining the physical properties or the
chemical constituents of the grain crop. No attempt has been made to
trace out from these experiments the manner in which the climatic factors
thus exert the chief determining influence on the composition of the
wheat crop. The following possibilities may, however, be considered :

(1) Differences in humidity may cause a difference in the transpiration
of the plants, which in turn may react on the composition of the crop.

(2) Variations in the amount and distribution of sunlight may influence
diversely the photosynthesis of the plants.

(3) Differences in temperature and in the succession of hot and cold
periods may cause varying vegetative activities in the plants.

(4) The climatic differences, such as the humidity, rainfall, temper¬
ature, and sunlight, may bring about changes in the physical, chem¬
ical, or biological characteristics of the soil which in turn may react on
the crop.

From this it should not be assumed that it is impossible for soil which
has been transferred from one locality to another to become so changed
by climatic environment that the character of the wheat grown thereon
would be approximately the same as that grown in soil belonging to
the second locality. This has been suggested to explain the facts ob¬
served during this experiment — namely, that wheats grown on the three
soils in Kansas are very different from the same variety of wheat grown

Jan. io, 1914

Environmental Influences on Wheat

291

on the same soils transported to Maryland. In view of the further fact,
generally accepted by agriculturists, that the same variety of wheat
grown over certain large areas having similar climatic conditions pos¬
sesses approximately the same physical and chemical characteristics,
notwithstanding the inherent differences in soil on which they were
grown or the differences of fertilizers applied to these soils, it would
seem that climate plays a greater role than soils as such in influencing
the composition of wheat.

Of the biological factors, those bearing on nitrification might be the
most influential in affecting the protein content of the crop. Yet it is
a noteworthy fact that the application of nitrate as a fertilizer increases
the protein content of the crop to only a slight degree. Considering the
great difference existing between the protein of the Maryland and Kan¬
sas crops, it may therefore be concluded that even if nitrification were
greater in Maryland soil transferred to Kansas than in Maryland soil in
Maryland, that fact would not be sufficient to explain the wide variation
between the composition of the wheat grown on the four plats in Mary¬
land and on the four plats in Kansas.

It is also shown that the crops from the plats which had been taken
up in 3 -inch layers and replaced had approximately the same physical
and chemical characteristics throughout as the crops from the corre¬
sponding plats which had not been thus disturbed. On the other hand,
it is shown that the climatic factors collectively have a strong determin¬
ing influence, especially upon the crude-protein content, the ash content,
and the percentage of phosphoric acid in the ash. The results from
this experiment thus harmonize with the findings previously published 1 —
namely, that environment rather than what has been usually termed
heredity is the major factor in determining the physical and chemical
characteristics of the wheat crop. They indicate, further, that it is the
climatic environment which exercises the primary influence of the envi¬
ronmental factors.

i7°73 — ; r4 - 2

1 I^e Clerc and I^eavitt. Op. tit.

A DROUGHT-RESISTING ADAPTATION IN SEEDLINGS

OF HOPI MAIZE

By G. N. Coixins,

Botanist , Office of Acclimatization and Adaptation of Crop Plants,

Bureau of Plant Industry

INTRODUCTION

A study of the maize grown by the Hopi, Zuni, and Navajo Indians of
New Mexico and Arizona has brought to light an adaptive character that
promises to be of economic importance in dry regions where germination
is uncertain.

These southwestern Indians have preserved from pre-Columbian times
a type of maize able to produce fair crops in regions where the better
known varieties of the East fail for lack of sufficient water. An impor¬
tant factor in the drought resistance of this type of com is its ability to
force the growing shoot of the seedling to the surface of the soil when
planted at a depth of a foot or more. At such depths less specialized
varieties die before reaching the surface.

The literature of corn contains reports of many experiments conducted
to determine the proper depth of planting, but the results are confusing
and contradictory. It has generally been realized that the optimum depth
is influenced by differences in soil and climate, but that the proper depth
might vary with different varieties seems not to have been appreciated.
The experiments referred to later, as well as many unpublished data
showing the varying behavior of types when planted at different depths,
indicate that it is unsafe and unscientific to generalize with respect to
cultural factors without taking type, varietal, and even individual differ¬
ences into account.

MORPHOLOGY OF THE MAIZE SEEDLING

To explain this drought-resistant character, it will be necessary to
discuss briefly the different parts of a maize seedling. (See fig. i.) The
primary root, or radicle, which is the first organ to emerge from the ger¬
minating seed, is soon followed by the shoot or plumule. Inclosing the
shoot is the cotyledonary sheath, or coleoptyle, a tubular organ which
is closed and pointed at the upper end. Between the base of the coleop¬
tyle and the seed the axis is somewhat elongated. With seeds germinated
in the laboratory this elongation is so slight that it might easily be over¬
looked. Nevertheless, this small organ has not escaped the notice of mor¬
phologists, and its nature has been the subject of much discussion. It has

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(293)

Vol. I, No. 4
Jan. 10, 1914
0-8

294

Journal of Agricultural Research

Vol. I, No. 4

been variously called “hypocotyl, ” “mesocotyl, ” and “epicotyl.” By
some it is held to be an intemode, by others merely an elongated node.

The choice of a name for the organ depends on the interpretation of
the homologies of the other parts of the embryo, particularly as to what
is considered as constituting the cotyledon. If the sheath, or coleoptyle,
be thought of as the cotyledon, the most appropriate name would be
hypocotyl. Although this interpretation was accepted by Richard
(1811),1 Hofmeister (1858), and Sachs (1875), there seems to be little
evidence in its favor and it is summarily dismissed by other mor¬
phologists.

The two remaining views are as follows :

(1) The scutellum alone is the cotyledon, the epiblast (absent in
maize) representing a second leaf and the coleoptyle a third. The elon¬
gated axis between the coleoptyle
and scutellum is thus considered an
internode and is then given the name
“ epicotyl.” Among the supporters
of this hypothesis are the following:
Warming (1879-80), Hackel (1887),
Bruns (1892), Van Tieghem (1897),
and Holm (1908-9).

(2) All these organs, scutellum,
epiblast, and coleoptyle, are viewed
as parts of a more highly specialized
cotyledon, in which case the term
“mesocotyl” is applied to the portion
between the coleoptyle and scutellum.
With various modifications this last
interpretation is adopted by Van
Tieghem (1872), Hagelmaier (1874),
Klebs (1881), Schlickum (1896), Celakovsky (1897), and Goebel (1905).

Van Tieghem originally subscribed to the view that the coleoptyle was
a part of the cotyledon, but as a result of further investigations aban¬
doned that position and adopted a modification of the views of Warming
to the effect that the mesocotyl and coleoptyle represent a metamer dis¬
tinct from the scutellum. The epiblast he held to be a rudimentary
second cotyledon. Van Tieghem carried this interpretation to its logical
conclusion and adopted the view that the apparent similarity between
the grasses and other monocotyledons did not represent homologies, but
that the two groups were phylogenetically distinct. He further held, on
the strength of anatomical differences, that the portion of the axis
between the scutellum and the coleoptyle is in some grasses an internode
and in others an enlongated node. The evidence regarding the mor¬
phology of the mesocotyl appears so conflicting that a definite interpreta-

Fig. i. — Diagram of seedling maize plant,
giving terminology of parts.

1 For “literature cited” see p. 301.

Jan. io, 1914

Drought-Resisting Adaptation in Maize

295

tion satisfactory to all morphologists seems very remote. With organs
that pertain to the very beginnings of the plant, even the primary dif¬
ferentiation into root, stem, and leaves may not be complete, and to
insist on a definite classification of these primitive organs may be idle.

Studies of seedlings of Hopi maize show that the mesocotyl may fre¬
quently develop up to lengths of 36 cm. / and it has been possible to note
a fact which appears thus far to have escaped notice — namely, that the
mesocotyl may give rise to roots at any point on its surface — but these
roots are threadlike and do not resemble the roots that arise from the
nodes of the culm. They do, however, closely resemble the roots that
arise from the radicle immediately below the seed. (See PI. XXIX,
fig. 1.) In grasses roots usually arise from nodes, not from internodes,
and the presence of roots on this organ in maize distinguishes it sharply
from subsequent internodes and is an argument in support of the inter¬
pretation that this intercalary growth, long though it is, is really a part
of the cotyledon and may properly be termed a mesocotyl. A further
reason for retaining the term “mesocotyl” is because the interpretation
implied by its use permits more direct comparisons with other groups
of monocotyledonous plants, where the organ sheathing the plumule
seems undoubtedly to be a part of the cotyledon.

From observations upon many varieties of maize it has become appar¬
ent that when a grain of corn germinates in the ground this usually insig¬
nificant organ is of vital importance to the life of the plant, for it is
through the elongation of the mesocotyl that the shoot is enabled to
reach the surface. So long as the seedling remains below ground, away
from light, the mesocotyl will continue to elongate until it reaches a
maximum length, which we have found to differ in different varieties,
but which seems reasonably constant within the variety. As the meso¬
cotyl elongates, the coleoptyle, with its firm, sharp point, is pushed
upward through the soil. As soon as the coleoptyle emerges from the
soil, the elongation of the mescotyl ceases, and elongation of the inter¬
node bearing the first true leaf begins, forcing open the coleoptyle.

If the seed is planted so deep that the maximum elongation of the
mesocotyl, which in anatomical structure shows a striking relation to the
radicle, fails to bring the coleoptyle to the surface, the task of penetrating
the soil and reaching light devolves upon the first true leaves. In com¬
parison with the sharp coleoptyle, these leaves are but poorly adapted
for forcing their way through the soil, and if the tip of the coleoptyle
stops more than a few centimeters below the surface these leaves usually
crumple and never reach the light.

In the varieties of maize commonly grown we have been unable to force
the mesocotyl to a length greater than 10 cm., while in the Hopi and
Navajo varieties this usually minute organ has in our experiments fre¬
quently reached the enormous length of 25 or even 30 cm.

1 In Euchlaena also the mesocotyl may reach a length of 28 cm. Van Tieghem gives 3 cm. as the maxi¬
mum length of this organ in grasses.

296 Journal of Agricultural Research voi. 1, no. 4

GERMINATION OF NAVAJO MAIZE

It has been frequently stated that the Navajos, like their neighbors,
the Hopi and Zunis, plant maize at unusual depths, 15, 30, and even 45
cm. having been reported. Since planting at such depths is known to
be impracticable with other varieties, experiments were planned to test
the ability of the Navajo maize 1 to pierce the soil. A representative
experiment is here reported. A box 70 cm. long, 33 cm. wide, and 34
cm. deep was sunk in the ground. A quantity of sandy-loam soil
sufficient to fill the box was slightly moistened and carefully sifted. At
one end the box was filled to within 1 cm. of the top, the soil sloping in
a straight line to within 1 cm. of the bottom at the other end.

Fig. 2. — Diagram showing the average size of seedlings of Chinese, Boone County White, and Navajo

maize planted at different depths.

Five seeds each of Navajo, Boone County White, and Chinese maize
were placed in a row transverse to the inclined surface of the soil, 2 cm.
from the top of the box. A similar row was planted at a depth of 4
cm. from the top, and so on at the following depths: 6, 8, 10, 12, 16,
20, 24, 28, and 32 cm. The box was then filled with the soil and struck
off level with the top. The seeds germinated promptly, and when the
most advanced seedlings had reached a total height of about 60 cm.
the plants which appeared above the surface were dug up, and the
mesocotyl and coleoptyle were measured. (See Table I and fig. 2.)

1 In the fall of 1912 Messrs. Walter T. Swingle and Karl F. Kellerman visited the region about Shiprock,
N. Mex. , in the Navajo Reservation and secured specimen ears of the maize grown by the Navajos. This
collection was kindly placed at the disposal of the writer. Additional seed was later secured through the
-courtesy of Mr. William T. Shelton, Indian agent at Shiprock.

Jan. io, 1914

Drought-Resisting Adaptation in Maize

297

Table I. — Average measurements of seedlings of Chinese , Boone County White , and
Navajo maize planted at different depths.

Depth.

Chinese.

Boone County White.

Navajo.

Cole¬

op¬

tyle.

Meso¬

cotyl.

Coleop¬

tyle

and

meso¬

cotyl.

Cole¬

op¬

tyle.

Meso¬

cotyl.

Coleop¬

tyle

and

meso¬

cotyl.

Cole¬

op¬

tyle.

Meso¬

cotyl.

Coleop¬

tyle

and

meso¬

cotyl.

Cm.

2* 3

2.3

4.6

3-7

3- 2

6.9

5-5

5*0

10. 5

2- 5

3- 5

6.0

3- 1

4. 9

8.0

4-3

6* 5

10. 8

2. 8

5-o

7.8

3-4

6. 1

9- 5

5- 2

10. 2

15*4

2* 5

S-*

8.3

2. 8

7-4

10. 2

4-9

11. 0

i5-9

3- 2

5-8

8.9

3- 1

8.6

11. 7

5-6

12. 2

17.8

4. 0

5- 2

9. 2

3-4

10. 4

13-8

5-o

*5- 1

20. 1

4. 6

12. 4

17. O

4. 3

I7* 5

21. 8

4- 5

IO. O

15-4

4. 7

19. 7

24. 4

9. A

5. 2

23. 0

28. 2

5. 6

26. 5

32. 1

6- 5

29. 0

35* 5

Twelve cm. was the greatest depth from which seedlings of the Chinese
variety appeared at the surface. Seedlings of Boone County White ap¬
peared from all depths up to 20 cm., while plants of Navajo maize
appeared from all plantings, including the very deepest, 32 cm.

There were numerous instances in which the combined length of the
mesocotyl and coleoptyle was less than the depth at which the seed was
planted. This, of course, means that the upper layers of the soil were
penetrated by the true leaves. The maximum depth of soil thus pene¬
trated by the true leaves of the plants of the Chinese variety was 5
cm. One plant of Boone County White maize forced its leaves through
8 cm. of soil. In all of the Navajo plants the coleoptyle reached the
surface.

The extent to which the seedlings of the Chinese and Boone County
White varieties were able to penetrate the soil by means of the true
leaves was doubtless much greater in the carefully prepared soil of the
experiment than it would be under field conditions, where any slightly
compacted lump of soil would deflect the tender leaves and cause them
to crumple. On the other hand, many seedlings failed to come up where
there was less than 2 cm. between the top of the coleoptyle and the
surface of the ground. The results clearly show that the coleoptyle
is the proper organ for penetrating the soil, and where this office devolves
upon the leaves there will be many plants that fail to reach the surface.

It has been observed in many field plantings that the spatulate first
leaf, formerly called the cotyledon, is the first evidence of the germinating
plant. When this occurs in any considerable proportion of the plants,
it is safe to assume that the seed has been planted too deep for the best
results.

298

Journal of Agricultural Research

Vol. I, No. 4

The three types of maize used in the box experiment were also planted
in the field. Four seeds of each of the varieties were planted as follows ::
At the surface and at 5, 10, 20, 30, and 40 cm, below the surface.
The greatest depth from which plants of the Chinese variety reached the
surface was 10 cm., that of the Boone County White was 20 cm.,
while that of the Navajo was 30 cm.

The seeds planted at the surface were naturally the first to appear,
but on June 17, one month after planting, the largest of the Chinese
variety were those from a depth of 5 cm., while the largest plants
of both the Boone County White and the Navajo maize were from the
10-cm. depth. On July 11 the plants that came up from a depth
of 10 cm. were the tallest in all the varieties, including the Chinese,
and to the end of the season this appeared the most favorable depth for
the Chinese and Boone County White varieties. With the Navajo,
however, the plants from a depth of 20 cm. had equaled those from the
10-cm. depth before the end of July, and from that time the plants from
the 20-cm. planting continued to make the most rapid growth, as
though this depth represented the most favorable condition for the
Navajo variety.

DESCRIPTION OF ROOT SYSTEM

We have observed further that the root systems of the Navajo, Hopi,,
and Zuni varieties differ from those of the other varieties; the roots
of their seedlings extend to a greater depth, and there is only a single
root arising from each seed, while in the seedlings of the Chinese and
Boone County White varieties the roots are shorter and more numerous.

The roots of maize are of two kinds : Those that arise from the embryo
or seed, called “seminal roots,” and those produced from the nodes of the
plant. Of the latter class those that arise from the nodes above the
ground are often called “brace roots” or “aerial roots.” In the varieties
commonly grown in the United States there are, in addition to the pri¬
mary root, or radicle, from two to six additional roots that arise from the
base of the cotyledon. These secondary seminal roots, though appearing
somewhat later, usually equal or exceed the radicle in size. In the
Pueblo varieties of maize these secondary seminal roots have been absent
in all seedlings thus far examined, the radicle being the only root arising
from the seed. (See Pis. XXIX and XXX, fig. 2.)

FIELD STUDIES OF PUEBLO VARIETIES OF MAIZE

In September, 1913, opportunity was afforded for a visit to the Zuni,.
Navajo, and Hopi Indian Reservations of Arizona and New Mexico. It
was thus possible to form some idea of the agricultural significance of the
peculiar habits of germination of this type of maize.

The value of deep planting made possible by the greatly elongated
mesocotyl was obvious. In the localities selected by the Indians for

Jan. io, 1914

Drought-Resisting Adaptation in Maize

299

planting maize the soil is sandy, and in the absence of spring rains the
surface layers are, of course, very dry. (See PI. XXXI, figs. 1 and 2.)
The seed, to germinate at all, must be planted deep enough to be in con¬
tact with the moist soil. In Navajo fields near Tohatchi, N. Mex., plants
were dug up, and the remains of seeds were found at depths ranging from
13 to 18 cm. below the surface. Similar depths were found in a Zuni field
near Black Rock, Ariz. (See PI. XXXI, fig. 1.) In a Hopi field at
Polacca, Ariz., near the First Mesa, where the conditions are extreme,
the seed had been planted at a depth of 25 cm. (See PI. XXX, fig. 1.)
It thus appears that there is no fixed depth for planting, the custom being
to plant deep enough to place the seed in moist soil. If the seed were
planted at ordinary depths, germination might be delayed until the
latter i part of June or the first of July, at which time the rains usually
occur; or if the seeds germinated as a result of one of the occasional
showers occurring in May, the plants would die from subsequent desic¬
cation.

Like the long mesocotyl, the simple radicle of the Pueblo varieties of
maize may be looked upon as an adaptation to the extreme conditions that
exist where these types are grown. For six or eight weeks after planting,
no rain can reasonably be expected, and during this time the moisture
is constantly receding from the surface. By concentrating the energy
of the seedling into a single root the latter is forced to greater depths and
consequently kept in moister soil than would be the case were a number
of seminal roots developed.

Under ordinary conditions, where moisture is distributed through the
entire seed bed, the seminal roots become of little importance as soon as
the seedling is established and nodal roots have developed. If a half-
grown or nearly mature com plant is carefully dug up, the seminal roots
and traces of the seed can still be found, but they are usually dry and
shrunken and are obviously of little use to the plant. This was also the
condition found in Navajo and Zuni maize fields, though the seminal
root was more strongly developed than in the eastern varieties. (See
PI. XXIX, fig. 2.) But in the more extreme conditions existing in the
fields near the Hopi villages, where the seeds were planted deeper, it was
found that the seminal roots were relatively much larger and were still
alive and fresh, making it apparent that they retain their function of
supplying moisture and are able to play an important part during the
entire life of the plant.

In one Hopi field at the base of the First Mesa the hills of maize were
planted about 20 feet apart, with from 10 to 20 plants in a hill. The
soil was apparently pure sand washed down by the winter rains and
entirely destitute of vegetation other than the planted maize. An
average hill dug up in the field was found to contain 15 plants ranging
from 60 to 90 cm. in height. (See PI. XXX, fig. 1.) The remains of the
seeds were found at 25 cm. from the surface, and from each seed there

3°°

Journal of Agricultural Research

Vol. I, No. 4

descended a single large seminal root. (See PL XXX, iig. 2.) These
seminal roots were traced to a depth of 35 cm. and extended even farther
down. They were still fresh and densely covered with fine branches.
This mass of 15 seminal roots, while less in volume than the nodal roots
arising near the surface, was apparently playing an important part in the
support of the plants. The mesocotyls connecting the seminal roots
with the plants above, while dry on the outside, were filled with live
tissue quite unlike the dry and shrunken mesocotyls found in plants of
similar age grown under more favorable conditions.

When planted by the Indian methods, the Hopi and Navajo varieties
of maize have been found superior to the more improved eastern varieties
for these very dry regions. At the time of our visit there was a small
field near Kearns Canyon that had been planted by eastern methods.
The plants were in rows and thinned to one stalk to the hill. There
had evidently been a fair germination, but the plants had died without
reaching maturity and had produced no seed. At the same time, in
the nearest Indian fields at Polacca the plants were dark green and
maturing a fair crop, though the season was said to have been unusually
dry. (See PI. XXXI, fig. 3.)

Even under irrigation the somewhat larger strains grown by the
Navajos have been found to compare very favorably with eastern types.
Several acres of Navajo maize were seen at Shiprock, N. Mex., under
irrigation. The fields were very uneven, apparently the result of alkali,
but in the better portions the yield was good. The plants were standing
about 2 feet apart in the row, the rows 4 feet apart, and nearly every
plant was bearing from two to four fair-sized ears. (See PL XXXII.)

The ears from 36 plants, representing a number of distinct types, were
collected. The 36 plants bore in all 94 ears, weighing 37.6 pounds, an
average of 15.2 ounces per plant. The plants producing these ears
averaged only a little over 5 feet in length.

CONCLUSIONS

Throughout the western part of the Great Plains area the difficulty
of securing uniform germination is a serious obstacle to the growing of
maize. With the varieties commonly grown, if the seed is planted at
the customary depth, many seeds fail to germinate from insufficient
moisture; if planted deep enough to come in contact with moist soil,
the plants may fail to reach the surface.

The agricultural Indians of the Southwest have continued from pre¬
historic times to grow maize successfully in regions where drought, and
especially the absence of spring rains, makes it much more difficult to
start the crop than in the Great Plains. A study of the varieties grown
by the Hopis and other agricultural Indians shows that these varieties
possess two special adaptations: (1) A greatly elongated mesocotyl that
permits deep planting and (2) the development of a single large radicle

Jan. io, 1914

Drought-Resisting Adaptation in Maize

301

that rapidly descends to the moist subsoil and supplies water during the
critical seedling stage.

This indigenous type of maize seems to have attracted little attention,
perhaps because it has been included in the popular mind with a series
of inferior varieties commonly known as “ squaw corn.” But the Pueblo
Indians of Arizona and New Mexico have strains sufficiently productive
to compare favorably with improved varieties even when grown under
irrigation. The peculiar adaptations of this type definitely indicate its
value for the semiarid regions and warrant experiments to determine
the possibility of its utilization.

LITERATURE CITED

Bruns, Erich.

1892. Der Grasembryo. Flora, Jahrg. 76, p. 1-33.

Celakovsky, L.*

1897. Uber die Homologien des Grasembryos. Bot. Ztg. Jahrg. 55, p. 141-174.
Goebel, K. E.

1905. Organography of Plants. Pt. 2, Oxford, p. 416.

Hackel, Eduard.

1897. Gramineae. Engler, Adolf, and Prantl, K. A. E-, Die Natiirlichen
Pflanzenfamilien. T. 2, Abt. 2, p. 10.

Hegelmaier, Friedrich.

1874. Zur Entwicklungsgeschichte monokotyledoner Keime nebst Bemerkun-

gen uber die Bildung der Samendeckel. III. Bot. Ztg. Jahrg. 32,
col. 661. •

Hoemeister, Wilhelm.

1858. Neuere Beobachtungen iiber Embryobildung der Phanerogamen, Jahrb.
Wiss. Bot. [Pringsheim] Bd. 1, p. 154.

Holm, Theodor.

1908-9. Observations on seedlings of North American Phaenogamous plants.
Ottawa nat. v. 22, p. 165-174, 1908; p, 235-244, 1909.

Klebs, Georg.

1885. Beitrage zur Morphologie und Biologie der Keimung. Untersuch. Bot.
Inst. Tubingen, Bd. 1, p. 536.

Richard, L. Cl.

1811. Analyse botanique des embryons Endorhizes ou monocotyledon^, et
particuli&rement de celui des Gramin6es. Ann. Mus. Hist. Nat. [Paris],
t. 17, p. 223-251; 442-487.

Sachs, Julius.

1875. Text-book of Botany. Oxford, p. 541.

Schlickum, August.

1896. Morphologischer und anatomischer Vergleich der Kotyledonen und ersten

Laublatter der Keimpflanzen der Monokotylen. Stuttgart, p. 56.
(Bibliotheca Bot. Heft 35.)

Van Tieghem, Philippe.

1872. Observations anatomiques sur le cotyledon des gramin6es. Ann. Sci.
Nat. Bot. s. 5, t. 15, p. 236-276.

1897. Morphologie de l’embryon et de la plantule chez les gramindes et les

cyp6racees. Ann. Sci. Nat. Bot. s. 8, t. 3, p. 259-309.

Warming, Eug.

1879-80. Forgreningen og Bladstillingen hos Slaegten Nelumbo.. [Footnote.]
Vidensk. Meddel. Naturhist. Forening. Kjjzfbenhavn, p. 446-448.

DESCRIPTION OF PLATES

Plat^ XXIX. Fig. i, — A seedling of Hopi maize with mesocotyl 18 cm. long. The
seed was planted in sand 20 cm. below the surface. There is a
single seminal root with threadlike branches similar to those arising
from the mesocotyl. The first nodal roots have begun to form at
the base of the coleoptyle. One-half natural size.

Fig. 2. — The root system of a plant of Zuni maize dug from a field
near Zuni, N. Mex., showing the well-developed, single seminal root
and the comparatively feeble nodal roots. Natural size. The field
from which this plant was dug is shown in Plate XXXI, figure 1.

XXX. Fig. i.-r—A hill of Hopi maize containing 15 plants grown under con¬
ditions of extreme drought at the base of the First Mesa near Polacca,
Ariz. The ears can be seen borne at the surface of the ground.

Fig. 2. — A plant of Hopi maize. One of the smaller plants from the
hill shown in figure 1 . The remains of the seed are scarcely visible
at the sharp bend of the mesocotyl, 25 cm. below the surface of
the ground.

XXXI. Fig. 1. — A field of Zuni maize near Zuni, N. Mex. One of the hills
near the center containing but a single plant shows a relatively
large ear borne at the surface of the ground.

Fig. 2. — A hill of Zuni maize in the field shown in figure 1. Note
the large ears borne near the surface of the ground.

Fig- 3- — A hill of Hopi maize making luxuriant growth under condi¬
tions of extreme drought. Note the manner in which the low-
spreading plants shade the ground. Polacca, Ariz.

XXXII. Fig. 1. — A single plant of Navajo maize grown under irrigation at
Shiprock, N. Mex.

Fig. 2. — The basal portion of the plant of Navajo maize shown in
figure 1, with leaves and husks removed. The ears from this plant
after drying weighed 2 pounds.

(302)

Plate XXX

SOME DISEASES OF PECANS

By Frederick V. Rand,

Scientific Assistant, Fruit-Disease Investigations, Bureau of Plant Industry
INTRODUCTION

The pecan, Carya illinoensis (Wang.) K. Koch,1 is an indigenous tree
of the hickory group, which has long been famous for the excellent quality
of its fruit. From the time when the earliest settlers first gathered the
nuts from native forest trees the pecan has been growing steadily in
favor.

Until recently the entire supply has come from the wild forest trees
and from a comparatively few, more or less isolated seedling orchards.
During the last 15 years, however, artificial propagation by budding
and grafting has gradually assumed a commercial importance until at
the present time a large number of excellent horticultural varieties are
available. These are being planted on a large commercial scale and
through an ever-widening range.

The pecan is found native on low, rich ground in the neighborhood of
streams from the valley of the Mississippi River in Iowa through south¬
ern Illinois and Indiana, western Tennessee to central Alabama and
Mississippi, western Louisiana through Arkansas and Missouri to south¬
eastern and western Kansas, eastern Oklahoma, and the valley of the
Concho River, Tex. It is also found in some of the mountain regions
of Mexico. As a native tree the pecan is most abundant and attains
its largest size in southern Arkansas, eastern Oklahoma, and middle
to eastern Texas.2 As a cultivated tree, however, it is by no means
confined to the sections above enumerated. Plantings of greater or
less extent have been made in Virginia, North Carolina, South Carolina,
Georgia, Florida, New Mexico, California, Oregon, and Washington, with
small experimental plantings in several other States.

Notwithstanding the highly colored statements of some of the early
promoters of pecan culture, this tree, like all of our cultivated fruit trees,
has its insect and fungous enemies. Possibly they would form a shorter
list than would those of some of our common fruits, but they are none
the less real and important, for, whenever a plant is brought under culti¬
vation or taken out of its native range, new diseases and new problems
with old diseases are sure to follow.

Other things being equal, the larger the number of individuals of a
host species growing in a given area the greater the chances any particu-

1 Synonyms: Carya oVvnaeformis Nutt.; Hicona pecan Brit.; Juglans pecan Marsh.

2 Sargent, C. S. Manual of the Trees of North America. Boston, 1905, p. 134.

Journal of Agricultural Research,

Dept, oi Agriculture, Washington, D. C.

(303;

Vol. I, No. 4

Jan. 10. 1914

304

Journal of Agricultural Research

Vol. I, No. 4

lar parasite has of successfully reproducing itself from season to season,
and consequently the more general and severe will be its injury over
that area. Thus, a disease occurring occasionally or with but slight
injury upon more or less isolated host individuals may under conditions
of close orchard planting assume an entirely different aspect, becoming
more nearly seasonal in its occurrence and causing a much greater per¬
centage of injury. A large part of the assumed difference in injury by a
disease under native and under orchard conditions is, however, often
merely psychological. In orchard culture the ideal sought is a thrifty
growth and abundance of high-grade fruit for every tree planted. Any
deviation from this ideal is quickly noted by the grower; whereas little
consideration is given to the facts that under native conditions large
numbers of individuals succumb to disease for every one that persists
and reaches maturity and that careful observations and comparisons are
seldom made with those which do reach maturity.

Nevertheless, the general fact remains that well-known diseases are
often more destructive under orchard conditions. Further than this,
diseases of hitherto unknown occurrence upon a particular host may
suddenly make their appearance. Some of these may have been present
but previously unnoticed, while others may be actually new to the host.
They are often brought to a locality with the introduction of new plants,
and with the widening of the range of a host the diseases of related plants
will be encountered sooner or later. Furthermore, a parasite is often
more destructive when brought to a new locality, either because of the
absence of its former enemies or because of other conditions more favor¬
able to its growth and reproduction in the new environment.

It has long been known that where a considerable number of plants
or animals are exposed in a similar way to the attacks of a parasitic
disease more or less difference will be noted in their behavior toward
the disease. In many cases some individuals will be found which seem
to be entirely immune, others which are very susceptible to attack, and
still others with varying grades of immunity or susceptibility between
two extremes. In localities favorable to the growth and spread of a
disease this condition works for the general benefit of the species attacked.
Those individuals least susceptible to injury will be most successful in
reproducing themselves, and thus a more or less immune race will be
developed. On the contrary, if a race has arisen amid conditions un¬
favorable to the development of a particular disease, or in its entire
absence, growth and reproduction will have taken place with little or
no relation to the disease. If such a race is exposed to the disease, it
is probable that a large percentage of its individuals will be found to be
susceptible.

These relations between host and parasite, though only a few among
many, may at least serve to indicate the extreme complexity of all prob-

Jan. io, 1914

Some Diseases of Pecans

305

lems having to do with living things. Partly because of this complexity
most problems of disease control are problems of “ better and worse”
rather than of 4 ‘good and bad,” for very few varieties prove to be abso¬
lutely immune, and very few artificial methods of control are entirely
effective.

The present paper deals only with certain distinct and more or less
troublesome fungous and bacterial diseases of pecans.1 For the most
part these studies were carried on during the years 1911 and 1912.

NURSERY-BLIGHT

[Caused by Phyllosticta caryae Peck]

HISTORY AND DISTRIBUTION

Nursery-blight is one of the worst known diseases of the pecan to
affect nursery seedling trees. However, in spite of the fact that young
trees are often defoliated from this cause by midsummer, no definite
investigation has hitherto been carried out and published, so far as
could be ascertained. This may be due partly to the fact that the pecan
nursery business is of comparatively recent origin and partly to the
obscurity of the causal fungus.

The distribution of this disease has been found to correspond very
closely with that of the pecan scab and the brown leaf -spot. Affected
specimens have been received from most of the pecan-growing States,
and personal observations have further demonstrated its presence at
Petersburg, Va. ; Orangeburg, Summerton, and Charleston, S. C. ;
Albany, De Witt, Hardaway, Baconton, Thomasville, and Cairo, Ga. ;
Tallahassee, Newport, Monticello, Glen St. Mary, Jacksonville, St. Augus¬
tine, Palatka, and Belleview, Fla.; New Orleans, La.; and at San Anto¬
nio, Boeme, Kerrville, Waco, and San Saba, Tex. Strains of the fungus
obtained from as widely separated points as Florida and Texas have
been similar in cultural characters and have caused the same symptoms
upon artificial inoculation, thus demonstrating the disease in both cases
to be of the same origin. Wherever observations have been made the
disease has for the most part been found to affect young trees, and by
far the greatest injury has been to the 1 and 2 year old nursery stock.2
Mature trees are seldom seriously injured.

1 No discussion of the scab, a serious disease of pecans, is included in the present paper.

2 A very effective control of the nursery-blight with Bordeaux mixture was obtained in two different
localities during the season of 1911, and there appears to be little reason to doubt that it will prove effica¬
cious in other localities and seasons. The quantity of spray material used and the cost of application under
nursery conditions are small, and it is thought that the increase in size and vigor, together with better
conditions for budding, will amply repay the small cost in material and labor necessary for the treatments.
It is obvious that the first application should be made before the disease has gained much headway in the
spring. Three to five subsequent applications may then be given at intervals of three to four weeks, accord¬
ing to the season.

3°6

Journal of Agricultural Research

Vol. I, No. 4

SYMPTOMS OF THE DISEASE

So far as has been observed nursery blight affects only the leaf blade,
but infections occur from early spring well on through the season, so
that under conditions favorable to the development of the disease the
young trees have little opportunity for growth. Generally the first
indications of infection appear in the form of minute roundish spots,
which are dark reddish brown on the upper leaf surface and blackish
on the lower. (PI. XXXVII, Fig. C.) These slowly increase in size
until a diameter of 2 to 5 mm. is often reached in the individual
spots. With increase in size the center of the spot on the upper surface
assumes an ashen-gray color, which is usually bordered with reddish
brown, while the lower surface remains black throughout or with an
occasional tiny ashen-gray spot in the center of this dark-colored area.
(PI. XXXVII, Fig. /.) The gray color in both cases is caused by a rais¬
ing of the epidermis, thus leaving an air space between it and the tissues

Fig. i. — Cross section of pecan leaf recently infected with the nursery-blight fungus ( Phyllosticta caryae

Peck) from pure culture. X260.

below. The leaves are often considerably peppered with these spots,
and by their coalescence larger areas are often involved. Very fre¬
quently the spots elongate and coalesce along the midrib and larger
veins, thus giving a very characteristic appearance. The parenchyma
cells and vascular bundles are often killed and discolored over large
areas. Whenever the vascular tissue becomes involved to any great
extent the supply of water is cut off from below and the leaf soon dries
up and falls. Figure i shows the microscopical appearance of the
diseased cells in a recently infected leaf.

MYCOEOGICAE AND PATHOEOGICAE STUDIES
Isolation of the Fungus

Rough microscopical examination of a considerable range of diseased
material disclosed no fungous or bacterial form which was at all con¬
stantly associated with the symptoms. Occasionally a tiny thin-walled
pycnidium was encountered, but no spores were found and usually no
fungous mycelium or fruiting body of any kind. Cultures made during
1910 and 1 91 1 from material several days old gave only saprophytic
fungi as shown by the negative results of all the inoculation tests.

Jan. io, 1914

Some Diseases of Pecans

307

Experimentation had already shown that the disease was readily con¬
trollable by Bordeaux mixture, and hence it was thought highly probable
that it was of parasitic origin. Consequently, in the summer of 1912,
materials for making cultures were taken directly into the field with ttie
idea of locating the cause, if possible, by any of the ordinary methods of
isolation. Leaves showing very recent infection were taken from the
highest parts of the trees where there was little or no spattering from the
soil. These leaves were placed in sterile Petri dishes and taken to the
temporary laboratory, where the tiny spots were cut out at once with
sterile scissors and transferred by the ordinary poured-plate method to
Petri dishes of com meal and synthetic agar.1 After 24 to 48 hours
colonies became visible which had evidently originated from the diseased
areas, and their appearance was quite uniform in all the cultures, except
in a few cases where contaminations had entered. Transfers were then
made to tube cultures. In this way strains of the fungus were obtained
from Monticello, Fla., from Albany, Ga., and from San Antonio, Waco,
and San Saba, Tex.

Inoculations

Circumstances connected with field travel prevented the making of
any inoculation tests during the summer of 1912, but the following
summer and winter trials were carried out upon pottted seedling trees in
the greenhouse. The trees were sprinkled, inoculated from pure cul¬
tures, and covered after inoculation for several days with bell jars. Three
strains of the fungus were used in this work: One from Texas, one
from northern Florida, and a strain reisolated from an artificially infected
leaf.

Experiment No. i (Oct. 8, 1912). — The young leaves on four trees were inoculated
from 1 ^-months-old, nonsporiferous synthetic-agar cultures (Florida strain 122), the
slimy mycelial mass being smeared over portions of both leaf surfaces. These four
trees and two moistened but uninoculated check trees were left under bell jars for
five days. After a week small dark-brown specks were noted over the inoculated
areas. In three weeks these spots were 1 to 2 mm. in diameter and in every way
similar to natural infections. The check trees remained uninjured.

Experiment No. 2 (Nov. 9, 1912). — The young leaves on three seedlings and the
matured leaves on two others were inoculated as above only from 3- weeks-old, sporifer-
ous com-meal-agar cultures (Florida strain 122). The five inoculated and five check
trees were left under the bell jars for three days. Observation after two weeks showed
the production of small, roundish, dark-brown specks, which at three weeks had
become 1 to 3 mm. in diameter with small ashen-gray areas in the center. The lower

1 Synthetic agar. — (i) 1,500 c. c. of distilled water and 36 grams of agar. Cook in double boiler for one
hour at 15 pounds pressure.

(2) 500 c. c. of distilled water, 200 grams of dextrose, 40 grams of peptone, 20 grams of ammonuim nitrate,
5 grams of magnesium sulphate (crystals), 10 grams of potassium nitrate, 5 grams of potassium acid phos¬
phate (K2 HPCb), and 0.2 gram of sodium chlorid.

Boil in double boiler for 30 minutes, add agar and cook for five minutes. Restore to volume, titrate,
cool to 60 0 C., and add whites of two eggs. Cook to coagulate eggs, filter, tube, and sterilize.

This formula is modified from that given by Francis Darwin and E. Hamilton Acton in their Practical
Physiology of Plants, ed. 3, 1901, p. 68.

I7°73°— !4 - 3

3°8

Journal of Agricultural Research

VoJ. I, No. 4

surface of the infected areas was almost black. Infection had taken place upon all
the leaves inoculated, while none of the check trees showed any signs of the disease.

Experiment No. 3 (Dec. 7, 1912). — The mature leaves of four seedlings were inocu¬
lated from 3 -weeks-old, sporiferous corn-meal-agar cultures (Florida strain 122), and
those of three other seedlings from nonsporiferous synthetic-agar cultures of the same
age and strain. The seven inoculated plants and four checks were kept under bell
jars for three days. Observations at two weeks showed the leaves of the first set with
tiny dark-brown specks scattered over the inoculated areas and with some of the
spots beginning to show the grayish centers. The leaves inoculated from the syn¬
thetic-agar cultures were similar, but not quite so far advanced. The check trees
all remained uninjured.

Experiment No. 4 (Dec. 7, 1912). — The mature leaves of three seedlings were in
like manner inoculated from 3-weeks-old, sporiferous corn-meal-agar cultures (Texas
strain 127). At the end of one week the spots were just becoming visible, and after
two weeks the centers were turning gray on the upper surface, while the borders
remained the typical dark brown. There were no evidences of the disease on the
two check trees. All five trees had been covered with bell jars for the first three
days.

Experiment No. 5 (Dec. 18, 1912). — Two trees were inoculated from com-meal-
agar cultures isolated from one of the trees of experiment No. 3 (strain 163). Typical
infections appeared at five to seven days, and these gradually increased in size for
three weeks, finally taking on the grayish center and dark reddish brown border
above, with the color almost black below. The check trees remained healthy.

Experiment No. 6 (Dec. 18, 1912). — The mature leaves of three seedlings were
inoculated from sporiferous com-meal-agar cultures of two weeks’ incubation (Florida
strain 122). In this case the pycnidia were broken up in sterile distilled water and
sprayed upon the leaves. The three check trees were sprayed with sterile distilled
water, and all six trees were left under bell jars for three days. On removing the bell
jars it was noted that tiny dark-colored specks were forming over much of the areas
inoculated. These later proved to be the typical spots of the nursery-blight. No
evidence of disease appeared on the check trees.

Experiment No. 7 (Dec. 23, 1912). — The sporiferous pycnidia from young com
meal-agar cultures (Florida strain 122) were broken up in sterile distilled water and
sprayed upon the upper and lower surfaces of the leaves of three seedling pecan trees,
the leaves having previously been washed. Three days after inoculation sample
inoculated and check leaves were collected. These were killed and bleached in
alcohol, stained with eosin, and examined superficially under the microscope. The
conidia themselves, being almost bacillar in size, could not be seen with the low
power necessary in any such examination. However, here and there could be dis¬
tinguished a very fine mycelial growth stained pale pink by the eosin, and in a number
of cases hyplise were clearly seen entering the leaf through stomatal pores or openings
left by the breaking off of leaf hairs and resin glands. In one case the branching
hypha could be followed some distance beneath the epidermis from the stoma through
which it had entered. The check leaves showed no such fungous growth entering the
leaf.

After a week an examination of the leaves left on the trees showed tiny dark-colored
spots scattered over the inoculated areas, while at two weeks the typical grayish
centers had developed. The check leaves were still without injury.

In the above detailed experiments the leaves of 24 pecan seedlings
were inoculated at different stages of maturity and with three strains of
the fungus. Every inoculation was successful, and in no case did any of
the check trees show signs of the disease. These data seem to establish
the parasitism of the fungus beyond any doubt.

Jan. 10, 19x4

Some Diseases of Pecans

309

From the facts that most of the infections occur within 2 or 3 feet of the
soil surface, that such infections may take place through stomata and
other openings in the epidermis, and that pycnidia are of rare occurrence
upon the leaves while still attached to the tree, it seems very likely that
the general development of pycnidia takes place upon the dead and
decaying leaves after they have fallen to the ground and that most of the
infection occurs through the spattering of spore-bearing material from
the soil.

Cultural Studies

THERMAE TESTS

Four series of thermal tests were carried out, corn-meal-agar cultures
being incubated for two to three weeks at temperatures ranging from
i° to 40° C. No change occurred at i° or at 40°, while at 50 and 36°
growth, where it occurred at all, was so small as to be scarcely discernible.
The growth of the colony was extremely slow at 8°, but increased con¬
siderably in rate at 120. At 140, 160, and 20° the rate was nearly the
same, though with a very gradual increase toward the higher tempera¬
ture. The optimum for the temperatures tested occurred at 30°, while
at 32 0 growth was very similar to that at 120 to 140. Incubation of two
or three weeks at 370 to 40° invariably killed the fungus, no subsequent
growth taking place when again held at optimum temperatures.

Thus, incubated in com-meal-agar slant tubes, the fungus made at least
some growth at temperatures ranging from 50 to 36° C. (41 0 to 970 F.)r
with a very gradual decrease in rate from the optimum (30° C. or 86° F.)
downward, and a rather rapid decrease upward. The comparatively
high optimum temperature, together with the wide range of effective
growth at lower temperatures, will assist in explaining the extended and
continuous period of infection observed under held conditions.

cueturae characters

The more obvious characters of the fungus as grown upon a number
of culture media are as follows :

Beef- Agar Slant Tubes. — The colonies are at first somewhat convex, pale ocherous
in color, with slightly roughened but glistening surface, and without aerial mycelium.
Later, the surface becomes much wrinkled, often presents a corallike growth in the
older parts, and approaches a light Venetian red in color. A moderate production
of pycnidia usually takes place in cultures 1 or 2 months old. Colonies often attain
a diameter of 10 to 12 mm.

Com-Meal-Agar Slant Tubes (PI. XXXVII, fig. H). — Where little aerial mycelium
is present, the colonies are at first about the same color and general appearance as in
the young beef-agar cultures. The cottony aerial mycelium becomes a faint pinkish
white and is often present in considerable luxuriance. The submerged parts some¬
times give a pale-violet tinge to the agar, but little or no direct diffusion of color into
the medium has taken place. Pycnidia are produced in abundance and range from
75 to 150 ju in diameter. At first they are a pale-ocherous color, but later change to
dark brown or almost to black. Many cross connections between the hyphae have

3i°

Journal of Agricultural Research

Vol. I, No. 4

been observed, and swollen cells are commonly scattered here and there through the
mycelium. Colonies often cover the slant, but unlike those on beef agar they are
seldom much wrinkled.

Corn-Meal Flasks. — On this medium the colonies with i or 2 months’ growth attain
a diameter of 5 or 6 cm., and become deeply convoluted or wrinkled. The cottony
aerial mycelium where present is similar in color to that on corn-meal agar, while
the underlying pseudoparenchyma en masse takes on a yellowish bumt-sienna tinge.
Pycnidia were not observed.

Filter Paper.— Growth on filter paper moistened with sterile distilled water gave
small colonies of a pale-violet color and with or without a scant pinkish white aerial
mycelium.

Oxalic- Acid- Agar Slant Tubes. — The colonies are raised to convex, pale ocherous
around the margin and approaching a sepia brown throughout most of the central
portion. With age the vegetable dye of the medium becomes bleached, so that the
color of ordinary beef agar is finally assumed. No pycnidia were observed.

Synthetic- Agar Slant Tubes 1 (PI. XXXVII, fig. G). — The colonies are very convex,
with moist and glistening surface. The mycelial mass is extremely viscous, much
convoluted, burnt sienna to brown in color, and with the drying out of the cultures
assumes various shades of olive green, violet, brown, and reddish brown. Numerous
cross connections between the hyphae were noted, but no pycnidia have yet been
developed on this medium.

Morphology and Taxonomy

Several of the diseased spots from fresh material were killed in Carnoy’s
fluid, embedded in paraffin, and sectioned both vertically and horizon¬
tally in order to locate the course of the fungous growth within the
tissues. The mycelium was found to be septate, very fine, and nearly
or quite hyaline; and even in the stained vertical sections it was often
distinguishable with difficulty. This readily explains the fact that
examination of rough mounts from field material rarely gives any evi¬
dence of fungous growth within the leaf tissues. The mycelium was
best located in the stained horizontal sections, where it could be dis¬
tinctly seen ramifying through the intercellular spaces just above the
lower epidermis and throughout the mesophyll tissue. (Tig. 2.) Where
the spots involved the vascular tissue, the hyphae were often seen extend¬
ing immediately parallel to the vessels, the latter in such cases being
dead and discolored. In many cases this intercellular mycelium had
developed scattered swollen cells with large vacuoles, but thus far no
definite formation of pycnidia has been observed upon artificially infected
leaves. Upon field material, however, the tiny dark-colored fruiting
bodies are occasionally encountered upon the upper leaf surface.

In 1887 a Phyllosticta occurring on Carya alba was described by Peck,2
which from his description and an examination of type specimens,
appears to be identical with the nursery-blight fungus. Peck’s descrip¬
tion is as follows:

1 For the formula for preparing synthetic agar, see p. 307.

2 Peck, C. H. Plants not before reported. 40th Ann. Rpt.. N. Y. State Mus. Nat. Hist., 1886, p. 57.
1887.

Jan. io, 1914

Some Diseases of Pecans

3ii

Phyllosticta caryae, n. sp. — Spots large, irregular, often confluent, at first yellowish,
then brown, sometimes becoming grayish in the center; perithecia minute, .004 inch
broad, punctate, epiphyllous; spores irregularly elliptical, .0002 inch long, .00008
broad.

Living leaves of hickory, Carya alba , Piffard, August.

Several months afterwards Ellis and Everhart1 described under the
same name a Phyllosticta occurring on species of Carya at Newfield,
N. J. On account of Peck’s priority, the specific name of Ellis and
Everhart’s fungus was later changed by Saccardo to caryogena.2

Fig. 2.— Horizontal section of leaf recently infected with the nursery-blight fungus in pure culture. Xis°-

After examination of Peck’s material the two species were finally
considered by Ellis and Everhart as identical, and the following descrip¬
tion and statement was published:

Phyllosticta Caryae Pk, 40th Rep. 57. 1887.

P. Caryae K. & E. Journ. Mycol. 101. 1888.

P. caryogena . Sacc. Syll. 10:119. 1892.

Exsicc. Ell. & Evrht. X. A. F. 2155, 2677.

On various species of Carya from Maine to Kansas.

Spots large, irregular, often confluent, often acute at each end, with a nerve of the
leaf running through the center, .5-1 cm. diam., yellowish at first, becoming brown,

1 Ellis, J. B., and Everhart, B. M. New species of fungi from various localities. Jour. Mycol., v. 4,
no. 10, p. 101, 1888.

2 Saccardo, P. A. Sylloge Fungorum, v. 10, Patavium, 1892. p. 119.

312

Journal of Agricultural Research

Vol. I, No. 4

with the margin darker. Perithecia epiphyllous, minute, lenticular, black-brown,
100 fi broad. Sporules oblong or ellipsoid-oblong, 5 — 8X2 — 2.5/i. The fungus is also
found on old insect-galls on the same leaves. The 40th Rep. was given to the public
in May, 1888. P. Caryae E. & E. was not published till October, 1888. P. Caryae Pk.
and P. Caryae E. & E. are evidently the same 1

The leaf spots upon the pecan assume the reddish brown color at a
very early stage of development, though this is often preceded by a
slight yellowing of the tissue at the point of infection. Furthermore,
the grayish center is almost invariable in its appearance during the
later stages. Individual spots have rarely been found by the writer to
exceed 4 or 5 mm. in diameter, but by the coalescing of several initial
infections diseased areas at least up to 8 or 10 mm. have frequently
been observed.

The majority of the pycnidia have been found to vary but little from
100// in diameter, but extremes of 50 to 150 n have been noted for mature
pycnidia in culture. In the latter case they are usually much lighter in
color than on the host, assuming macroscopically a tawny appearance.
On the pecan leaf and occasionally in culture the fruiting bodies are
dark brown to black.

Conidia as observed on this host have corresponded closely with
Peck’s fungus, ranging within the limits of 3.8 to 6 by 1.5 to 2 /*. In other
points also the pecan fungus corresponds closely with the two descrip¬
tions quoted above.

Thus, on account of the close relationship between the hosts and the
many points of resemblance between the fungi and the disease symp¬
toms, it seems best to consider the nursery-blight fungus as identical
with Phyllosticta caryae Peck rather than to burden mycological litera¬
ture with another name. At least this course should be followed until
cultural and cross-inoculation work can demonstrate a specific difference.

BROWN LEAF-SPOT

[Caused by Cercosporafusca, emend, sp.]

history and distribution

With the growth of the pecan industry the brown leaf spot has grad¬
ually been receiving more notice among orchardists. Since it is by no
means as serious a trouble as the pecan scab, it has not merited the
attention given the latter. No published record has been found, except
a brief description of the fungus, and no work establishing the cause or
demonstrating a method of control.2 However, next to the pecan scab
it is perhaps the worst and most generally distributed leaf-spot disease

1 Ellis, J. B., and Everhart, B. M. The North American Phyllostictas. Vineland, N. J., 1900. p. 35.

2 The brown leaf spot has occurred to a limited extent at points where spraying tests were being carried
out on other pecan diseases and has been effectively controlled with three treatments of Bordeaux
mixture.

Jan. io, 1914

Some Diseases of Pecans

3i3

affecting the mature trees and consequently has been considered worthy
of investigation as well from a practical as from a mycological standpoint.

For several years specimens of leaves showing this disease have been
received from widely different parts of the pecan-growing territory,
while within the last two years the writer has made personal observa¬
tions in the field over much of this region. From these observations
and studies in field and laboratory it may definitely be said that the
brown leaf-spot occurs in South Carolina, Georgia, Florida, Alabama,
Louisiana, and Texas and that an exceedingly similar if not identical
disease has in numerous instances been seen on other species of hickory.
Furthermore, there is little doubt that its range is much greater than
that above indicated, since it has been found in nearly every pecan sec¬
tion visited by the writer during the last three years.

Observations in several States during the past two years have shown
very little difference in resistance to the disease among the different
varieties. For example, in one orchard examined, containing 45 varie¬
ties of the pecan, the brown leaf -spot was so uniformly distributed that
no appreciable difference in the amount of injury could be detected
among the different varieties. From a number of such observations
over a wide territory it may be safely assumed that little difference in
resistance exists among the varieties now commonly planted.

SYMPTOMS OF THE) DISEASE)

The leaf blade is the only part of the tree known to be affected. (PI.
XXXVII, fig. A.) In ordinary seasons or when only a few spots occur,
there is little or no appreciable injury, but occasionally under conditions
very favorable to the progress of the disease partial defoliation may
result. Infections occur from the early part of summer on until fall,
and under proper conditions of moisture and temperature even well-
matured leaves may develop the disease. Several days after infection
(ordinarily 3 to 10) the condition becomes evident through the formation
of a tiny dark reddish brown spot, which is usually somewhat angular
in outline and bounded by the veins of the leaf. The spots from the
earliest visible stages extend through the leaf tissue and appear about
the same in form and color on both surfaces. The size increases grad¬
ually until the diseased areas often attain a diameter of 10 or even in
some cases 15 mm. With increase in area the spot often loses its angular
outline and the margin becomes more indefinite, while at the same time
the center of the spot may in some cases assume a somewhat lighter
reddish brown color with the darker brown as a border. Very often,
however, the spots remain angular and with definite margin, though
in such cases they seldom attain a diameter of more than 2 or 3 mm.
Microscopical examination showed the cells within the affected areas to
be de£Jid, more or less opaque, and brownish in color. (Fig. 3.)

314

Journal of Agricultural Research

Vol. I, No. 4

MYCOlvOGICAL, AND PATHOEOGICAE STUDIES
Isolation of the Fungus

Examination of a wide range of material during the last three years
has invariably shown the same type of fungous growth and spore for¬
mation, while no other fungi have been found, except in the later stages
of the disease. It was considered very probable that the fungus above
mentioned was the cause of the diseased condition which it accompanied,
and so on October 4, 191 1 , single spore cultures were started, using conidia
from material collected August 29, i9ii,at Baconton, Ga. Synthetic
agar was used, and the germination was followed under the microscope
from day to day. Growth was rather slow at first, but continued until
at the end of a month colonies 5 to 15 mm. in diameter had been formed.

Prom one of the strains obtained in
this way the first inoculation tests
were made.

Inoculations

The inoculation work was carried
out during the winter and spring
of 1912 upon young seedling pecan
trees in the greenhouse. The leaves
to be used in the tests were mois¬
tened with water immediately pre¬
ceding inoculation, and since no
definite spore formation has taken
place in culture, bits of the mycelial growth were placed directly on the
upper or lower sides of the leaves thus moistened. The small potted trees
were then generally left for several days under bell jars, with slight ven¬
tilation at the base, to insure proper humidity for growth of the fungus.
Check trees in each experiment were treated similarly, with the excep¬
tion of the inoculation.

Fig. 3. — Cross section of a leaf infected with the
brown leaf-spot fungus from pure culture. X 250.

Experiment No. i (Feb. 29, 1912). — The young leaves of two potted seedlings
were inoculated from 3-weeks-old oxalic-acid and synthetic-agar cultures (strain
33), the first tree being covered with a bell jar and the second left open. Two
check trees were placed under a bell jar. The inoculated and check trees were
all sprinkled with tap water on the second and fifth days and the bell jars were removed
on the latter date. At the end of two weeks most of the inoculated leaves on the first
tree had developed small, reddish brown areas from mere angular flecks up to irregu¬
larly circular spots 1 mm. in diameter. Very little infection had occurred on the tree
left uncovered after inoculation, but several distinct spots were noted. Eater, many
of the spots had increased in size up to 7 or 8 mm., with the development of tawny
clusters of conidia visible to the naked eye upon the upper leaf surface. In no case
did the check trees show signs of the disease.

Experiment No. 2 (Apr. 16, 1912). — In a similar manner the tender leaves of a
seedling were inoculated on both surfaces from a month-old com-meal-agar culture
(strain 33). This tree and the check were left under a bell jar for five days. Observa¬
tion after a month showed a large number of the somewhat angular young spots up

Jan. io, 1914

Some Diseases of Pecans

315

to i mm. in diameter, but no spore formation had as yet occurred. After two months
the spots were well scattered over the inoculated areas, and some of them had at¬
tained a diameter of 10 mm. The pale tawny coni dial tufts were at this time very
abundant on the upper surface. No infection had taken place on the check tree.
After three months many of the smaller spots had coalesced to form reddish brown
areas up to 20 mm. in diameter.

Experiment No. 3 (Apr. 29, 1912). — The tender leaves of two seedlings were
inoculated on both surfaces from 5 -wee ks-old com-meal-agar cultures (strain 33).
These and the two check trees were left under bell jars for six days, the leaves being
sprinkled on the second and fourth days. At 10 days infection was just becoming
evident, while at the end of one month all but one of the inoculated leaves were
peppered with the more or less angular reddish brown spots. After six weeks the
development of conidial tufts began to take place on the upper leaf surfaces.

Experiment No. 4 (May 28, 1912). — The tender leaves of two seedlings were inocu¬
lated on the lower surface from a month-old synthetic-agar culture (strain 33) and
these, with the single check tree, were covered with bell jars for six days. Observa¬
tions after three weeks showed the typical spots of this disease up to 3 and in one
case 6 mm. in diameter. The conidial tufts were just beginning to form. No in¬
fection occurred on the check tree.

Experiment No. 5 (May 29, 1912). — The young leaves of two seedlings were inocu¬
lated from a month-old culture (strain 33) on sterile pecan wood, and the tree was
left under a bell jar for several days. At the end of one month numerous somewhat
angular reddish brown spots were evident on all the leaves, and these varied in size
from mere specks to areas 10 or 15 mm. in diameter. After six weeks the development
of conidial tufts had commenced. No infection occurred on the single check tree.

Experiment No. 6 (June 7, 1912). — The rather mature leaves of two seedlings were
inoculated on both surfaces from a 4- weeks-old synthetic-agar culture (strain 113, an
isolation from the artificially infected leaves described in experiment No. 1). The
air was hot and dry at this time, and hence the bell jars were left over these trees and
the three checks for eight days. Observations at the end of two weeks showed the
development of typical spots on all the inoculated leaves, and at one month the
formation of conidial tufts had begun.

Experiment No. 7 (3 p. m., May 23, 1912). — The tender leaves of a young tree of
the Schley variety at Arlington Farm, Virginia, were inoculated on both surfaces
from a 4- weeks-old prune-agar culture (strain 33). The day was cloudy, but the hot,
dry weather of the following week prevented infection.

Experiment No. 8 (2.30 p. m., June 7, 1912). — A second young Schley pecan tree
at Arlington Farm was inoculated from a 4-weeks-old synthetic-agar culture (strain
1 13). The day was cloudy, and the leaves were covered with moistened cotton to
further insure the growth of the fungus. The weather was rather hot and dry for
several days afterwards, but this period was followed by a day or so of rain. Eater
observations showed a moderate number of the typical spots on the inoculated leaves,
while the check leaves showed no signs of the disease.

Experiment No. 9 (3 p. m., June 15, 1912). — In a similar manner the four to six
young shoots of three Schley pecan trees at Arlington Farm were inoculated from
6- weeks-old corn-meal flask cultures (strain 33). In this case the shoots on one inocu¬
lated and one check tree were covered by heavy paraffined paper bags containing
moist blotting paper to insure a high humidity around the inoculated leaves, while
those on one check and two inoculated trees were left uncovered. Showers occurred
on the two following days. Examination in the fall showed many of the typical spots
developed on the inoculated leaves covered by the bags and on those of one tree left
uncovered. There was no infection on any of the check trees.

Experiment No. io (June 15, 1912). — The young leaves of one potted seedling and
the mature leaves of another were inoculated from a 6-weeks-old corn-meal flask

316

Journal of Agricultural Research

Vol. I, No. 4

culture (strain 33) and covered with bell jars. An uninoculated check tree was cov¬
ered in the same way. The trees were sprinkled twice between June 15 and 20, and
the bell jars were removed on the latter date, at which time definite infection was noted
on the first inoculated tree, but none was on the second or on the check tree. An
accident to these trees prevented further observations.

Experiment No. ii (Dec. 18, 1912). — The partly mature leaves of a seedling
pecan were inoculated from an 8-weeks-old com-meal-agar culture (strain 113). This
and one check tree were left under a bell jar for three days, when tiny reddish brown
specks could be recognized over the inoculated areas. After bleaching and staining,
these leaves were examined for the mode of entrance of the fungus into the leaf.
Many cases were found in which the mycelial threads had passed through the open¬
ings in the stomata. In all probability this mode of infection occurs in the field from
the germination of the spores, but this point has not been proved by artificial infection,
on account of the lack of distinct coni dial formation in culture.

Cultural Studies

THERMAL TESTS

Several series of com-meal-agar slant cultures were grown for two to
three weeks in constant-temperature incubators ranging from 1 0 to 40° C.
No growth took place below 50 or above 35 °. After two to three weeks'
incubation growth at 8° had barely started, while the rate gradually
increased up to 30° (86° F.), this giving the highest rate for the tem¬
peratures tested. Growth at 32 0 was about the same as at 140. Cul¬
tures incubated two to three weeks at 36° and 40° gave no signs of life
when subsequently held at room temperature, while those incubated at
20 and 40 made a perfectly normal growth when placed under favorable
conditions.

cultural characters

The cultural characters of the fungus as grown upon several of the
more common media are briefly described below. No distinct develop¬
ment of conidia has been observed in cultures of the fungus.

Beef-Agar Slant Tubes. — The colonies are convex, approximately raw umber
in color, glistening and smooth at first, but later becoming wrinkled and finally attain¬
ing a diameter of 10 to 12 mm. Aerial mycelium where present has been very sparse.
The submerged mycelium consists of a pale-olive, tangled mass of hyphae with many
swollen and contorted cells.

Beef Broth. — The entirely submerged and dirty- whitish colonies consist of a
rounded filmy mass of threadlike mycelium with but few swollen cells. Some of the
hyphae are beaded in appearance.

Corn-Meal- Agar Slant Tubes (PI. XXXVII, fig. K). — The submerged growth which
is usually the most prominent part is seal brown to black, while the somewhat cottony
aerial mycelium is pinkish. After an incubation of one to two weeks a distinct violet
tinge is assumed by the whole agar plug, and the combination of pigment and gela¬
tinous medium gives an opalescent appearance to the whole. Colonies often reach a
diameter of 1 5 mm . Except for the rather scant aerial mycelium , the growth is entirely
below the surface of the medium where the hyphae consist of more or less distorted,
dark olive-brown cells.

Corn-Meal Flask Cultures. — The colonies are cottony to plushlike in surface
appearance, with a wide variation of color comprising white to pale pink in the cottony
parts and shades of raw sienna, burnt umber, and Venetian red elsewhere. A diameter

Jan. io, 1914

Some Diseases of Pecans

3i7

of 50 or more mm. is often attained by individual colonies after a growth of several
weeks.

Filter Paper. — Growth on filter paper moistened with sterile distilled water
caused the formation of dark reddish brown circular spots very similar in appearance
to those formed on the leaf, while for a radius of 10 to 12 mm. around the spot the paper
took on a pinkish cast. An extremely scant white to pinkish aerial mycelium was
often developed.

Oxalic-Acid-Agar Slant Tubes. — Colonies are more or less convex, becoming
wrinkled with age. The rather scant aerial growth is white to pale pink, while the
submerged mycelium is seal brown to black and made up of densely anastomosing and
variously contorted hyphae. The colonies are rarely over 10 mm. in diameter. After
several weeks' growth the medium loses its pink color and assumes the shade of ordi¬
nary beef agar.

Potato Cylinders. — The colonies are very convex, with white to pinkish aerial
mycelium and olive-gray surface growth which becomes much wrinkled with age.
A diameter beyond 8 to 10 mm. is rarely attained. The potato cylinder assumes a
dark-gray cast for several millimeters beyond the outermost fungous growth, due
evidently to enzym action.

Prune-Agar Slant Tubes. — The colonies are little or not at all raised above the
surface of the agar, with a fine, velvety, Indian red aerial growth. In the older and
drier parts of the culture a scant white to pinkish aerial mycelium develops. A
diameter of about 15 mm. is usually attained.

Synthetic- Agar Slant Tubes (PL XXXVII, Fig. J). — The colonies are extremely
convex with a light to dark olive-green velvety surface growth. Numerous guttate
drops of liquid are scattered over the surface during the earlier stages. A dark-brown
to black, leathery pseudoparenchyma is developed beneath the surface, and with age
the whole colony becomes considerably wrinkled. Growth continues until the agar
has almost completely dried down, so that the whole slant surface of the medium is
often eventually covered by the fungus.

Morphology and Taxonomy

A comparison of the characters of this fungus with the description of a
Clasterosporium published by Heald and Wolf and an examination of
their type material deposited in the pathological herbarium of the Bureau
of Plant Industry have shown that the two are undoubtedly the same
species. Their description is as follows:

Clasterosporium diffusum. — Maculis indefinite marginatis, amphigenis; irregularibus,
aequaliter brunneis, 5-10 mm. diam.; hyphis effusis prostratis, saepe laxe gregariis
atque erectis; conidiis curvulis, clavatis, pluriseptatis, brunneis, 45-135 X 4-5 /*.

On Hicoria pecan (Marsh) Britton. Victoria, 2536; Gonzales, 2695 (type); Yoakum
2770, Halletsville, 2783.

This fungus produces circular or irregular, indefinite margined, brown spots, which
are uniformly brown on both surfaces of the leaflets. Dark-brown hyphae run
throughout the dead tissue or creep over either surface of the affected area, or are
sometimes aggregated to produce clusters of erect conidiophores.1

After a careful study of this fungus from both the humid and semi-
arid parts of the pecan belt it has seemed to the writer to conform more
nearly with the Cercospora than with Clasterosporium characters.

Typically, the latter is saprophytic and possesses a prostrate or creeping
mycelium with sporophores either short or differing but little from the

1 Heald, F. D., and Wolf, F. A. New species of Texas fungi. Mycologia, v. 3, no. 1, p. 21, 1911.

3i8

Journal of Agricultural Research

Vol. I, No. 4

conidia. The latter are borne singly, rarely in clusters, and are largely
straight, with rounded ends.

The Cercosporas, on the other hand, are mostly parasitic, and form
leaf spots. The sporophores are developed in thick bundles, either
through the stomata and from mycelium within the leaf tissues which
often takes the form of a stroma beneath the stomatal opening or by
sporophores breaking through the epidermis irregularly. The conidia
are longish-cylindrical or spindle-shaped, occasionally somewhat club-
shaped, straight or bent, and often with a long drawn-out point.

As observed in the humid sections, the typical forms of this fungus
had the densely clustered sporophores which, occurring mostly on the
upper leaf surface, have arisen from a stroma breaking through the
epidermis rather than through the stomatal openings. (Fig. 3.) The
mycelium is largely within the leaf tissue and is intercellular, but is also
found creeping over the leaf surface and giving rise here and there to
single conidiophores. In the semiarid sections the latter type of spore

formation appears to be the more frequent. The conidia are long, usu¬
ally somewhat club-shaped, and with the apical end the more pointed.

It will be seen that the fungus possesses some characters of both
genera. However, since under conditions favorable to fungus growth the
Cercospora characters greatly predominate, it has seemed best to place
it in this genus. Of course parasitism or nonparasitism should scarcely
be given a generic value, but this point at least adds further weight to
the present decision. Furthermore, since a Cercospora diffusa has been
previously described by Ellis and Everhart* 1 as occurring upon leaves of
Physalis lanceolata , it becomes necessary to change also the specific name
of this pecan fungus. The emended description of the fungus is given
below.

Cercospora fusca, emend, sp.

Syn. Clasierosporium diffusum Heald and Wolf.

Leaf spots up to 10 or 15 mm. in diameter, at first somewhat angular and bounded
by the veins, becoming roundish to irregular and with more indefinite margin, dark
reddish brown on both leaf surfaces.

1 Ellis, J. B., and Everhart, B. M. Additions to Ramularia and Cercospora. Jour, of My col v 4 no

1, P. 3. 1888. *

Jan. xo, 1914

Some Diseases of Pecans

3i9

Mycelium dark brown and septate, intercellular, sometimes also creeping over the
leaf surfaces.

Conidiophores mostly epiphyllous, short and erect, typically in dense, tawny
clusters from stromata developed beneath the epidermis and later bursting through,
also arising singly from the prostrate surface mycelium.

Conidia pale olive brown, highly variable in size, 30 to loop. or more by 3 to 6jx (see
figs. 4 and 5), usually curved, typically subclavate, multicellular, septa less frequent
toward the more pointed apical end.

Habitat. — hiving leaves of Carya illinoensis , Southern States. Also possibly occur¬
ring on other species of Carya. Diseased nuts or leaves seen by the writer at Orange¬
burg, Sumter, and Charleston, S. C.; Americus, Albany, DeWitt, Hardaway, Bacon-
ton, Thomasville, Cairo, Bainbridge, and Valdosta, Ga.; Tallahassee, Newport, Monti-
cello, Glen St. Mary, St. Augustine, Palatka, Gainesville, Ocala, and Belleview, Fla.;
New Orleans, ha.; and at Waring and San An¬
tonio, Tex. Reported also by Heald and Wolf1
from Victoria, Gonzales, Yoakum, and Hallets-
ville, Tex.

/oa

PECAN ANTHRACNOSE

[Caused by Glomerella cingulata (Stonem.) S. and v. S.]

HISTORY AND DISTRIBUTION

Pecan anthracnose, variously known
among pecan growers as “leaf -blotch”
and “rust,” was first noted by the writer
during the summer and fall of 1910, at
which time single-spore strains of the
causal fungus were obtained from per-
ithecia matured on the leaves in a damp
chamber. Studies of these cultures were
carried out during the following winter,
and a preliminary description of the
fungus later appeared under the name
►of Mycosphaerella convexula.2

Further cultural studies of the fungus brought out changes in its
morphology sufficient to throw it out of the genus Mycosphaerella, and
these, together with cross-inoculation experiments upon the apple, indi¬
cated its close affinity to the apple bitter-rot caused by Glomerella
cingulata .3 No other published information concerning this disease has
come to the notice of the writer.

Pecan anthracnose seems to be well distributed throughout the eastern
part of the pecan-growing territory, but it has thus far usually occurred
only to a limited extent at any one place. Diseased leaves or nuts have
been collected by the writer at Orangeburg, Sumter, Summerton, Charles¬
ton, and Aiken, S. C. ; at numerous places in southern Georgia and

3 * & e

W/DTH OP" SFORJETS /A/ M/CftOA/S.

Fig. 5.— Diagram showing measurements
in width of 200 conidia.

1 Heald, F. D., and Wolf, F. A. Loc. cit.

2 Rand, F. V. A pecan leaf-blotch. Phytopathology, v. 1, no. 4, p. 133-138, 3 fig., 1911.

3 Rand, F. V. Further studies on the pecan '‘rust/’ Science, n. s., v. 35, no. 913, p. 1004, 1912.

320

Journal of Agricultural Research

Vol. I, No. 4

northern Florida, including Albany, Dewitt, Baconton, Thomasville,
Cairo, and Bainbridge, Ga., and Tallahassee, Newport, Monticello,
Jacksonville, St. Augustine, and Belleview, Fla.; and at San Saba, Tex.

SYMPTOMS OP THE DISEASE

The disease has been found to occur on both leaves and nuts of the
pecan. On the leaves it appears in the form of irregular, reddish to
grayish brown blotches varying greatly in size and eventually often cov¬
ering the whole leaf. (PI. XXXVII, fig. B.) The color of the affected
areas is the same on both surfaces. Under conditions of moderate
humidity, spores of the Gloeosporium type are developed singly, but
with favorable temperature and moisture the acervuli with exuding pink
spore masses and the black perithecia appear rather thickly scattered
over the diseased blotches. When the greater part of the leaf blade
becomes involved, it usually falls to the ground, and it is here, under
natural conditions, that the acervuli and perithecia are developed.

The blotches on the nuts are also irregular in outline, but are nearly
or quite black and often slightly sunken below the surrounding healthy
tissue. (PI. XXXVII, fig. F.) The perithecia and the densely gregarious
acervuli are formed under the same conditions as on the leaves, but the
perithecia have been found on the nuts with much less frequency. A
serious dropping of the partially grown nuts sometimes occurs from this
cause.

A watery condition of the kernel is frequently found in connection
with the anthracnose blotches. It is doubtful, however, whether it has
anything to do with this disease, for the same condition prevails both
with and without external signs of injury, while both cultural methods
and microscopical study have thus far failed to locate any microorganisms
in the watery kernels.

MYCOEOGICAE AND PATHOEOGICAE STUDIES *

Isolation of the Fungus

No mature perithecia have as yet been found on fresh leaves or nuts,
but at various times during the last three seasons ripe asci have been
readily developed on affected leaves after an incubation of one or two
weeks in a damp chamber. The original strains of the fungus were
obtained in this way from nursery -tree leaves collected in the fall of 1910
at Baconton, Ga., and were each started from a single, apparently 2-celled
ascospore the germination of which was closely followed under the micro¬
scope to preclude the possibility of contamination. On several different
culture media the colonies at once developed perithecia suggesting the
genus Mycosphaerella, and the great majority of the slightly curved
ascospores were apparently 2-celled, though a few showed no signs of a
cross-septum.

Jan. io, 1914

Some Diseases of Pecans

321

After carrying in culture for about two months, a few spores were noted
which were 1 -celled, oblong, and slightly smaller than the typical asco-
spores. These conidia as first noted were borne hyphomycetously, but
later were found developing from dense groups of modified fungous cells
like an acervulus and in size and shape resembling a typical Gloeosporium.
For some time it was thought that this was a contamination, though no
possibility of such an occurrence could be found. However, in order to
determine this point with certainty, single-spore cultures were started
from the 2 -celled ascospores and also from the conidia, each individual
spore being isolated and its germination carefully followed microscopi¬
cally. The resulting two series of cultures were similar macroscopically,
and both soon developed typical perithecia and ascospores, and also the
Gloeosporium conidia. This procedure was repeated 30 or 40 times with
a like result in all cases. In several instances the germinating ascospores
had within 24 to 48 hours developed mycelial threads which were cutting
off conidia in considerable numbers, and in these cases the hyphal con¬
nection could frequently be traced from the parent ascospore to the
conidium.

However, it was noted after eight or nine months' growth in culture
that the 2 -celled ascospores were becoming fewer in proportion to the
1 -celled, and this tendency has continued until now, after more than two
years in culture, the majority of the ascospores are 1 -celled, though still
of the original form and size.

The possibility suggested itself that perhaps many of the apparent
septa were in reality merely a denser layer of cytoplasm across the center
of the single cell and that after many generations of growth in culture this
cytoplasm had for some reason become more homogeneous. Whatever
the explanation, the fact remains that in these original strains a change
has taken place from a majority production of apparently 2 -celled to
that of 1 -celled ascospores and that the production of acervuli has become
quite as abundant as that of the perithecia. It should be added, how¬
ever, that many of the ascospores possessed an undoubted septum, as
clearly brought out by staining.

During the last two years 10 or 12 other strains of the fungus have been
obtained from both conidia and ascospores developed on naturally
infected leaves and nuts. In these cases most of the ascospores have
been unicellular, though a few have been found with a cross partition
clearly brought out by staining.

Inoculations

Several series of inoculation tests have been carried out on the leaves,
but on account of the exigencies of field travel and the difficulty of obtain¬
ing suitable material and conditions only two sets of infection experiments
have been tried on the nuts.

322

Journal of Agricultural Research

Vol. I, No. 4

From the similarity of this fungus to the Glomerella rot of apples and
from the omnivorous character of the latter species, as brought out in a
paper read by Shear1 at the 1911 meeting of the American Association
for the Advancement of Science, it was decided to make several cross¬
inoculation tests on the apple. The results of inoculation tests on leaves
and nuts and of the cross-inoculation work on the apple are given in the
following pages.

LEAVES

Experiment No. i (Feb. 16, 1911). — A distilled water suspension of ascospores
from a month-old corn-meal flask culture (strain 17, Baconton, Ga., 1910) was sprayed
upon the lower surfaces of six potted pecan seedlings. Three of the seedlings were
under bell jars for four days, while the remaining three were left uncovered through¬
out the experiment. Observations at the end of a week showed no signs of infection,
but at four weeks numerous small discolored areas had developed on the foliage of
the first three trees and on that of all but one of the others. The three check trees
which had been sprayed with distilled water and left under bell jars for four days
were sound. No further development of the disease was apparent for some time, but
during the latter part of June large, dull reddish brown areas were noted on the leaves
of the first three inoculated trees and on one of those which had not been covered
with a bell jar. Specimens of these diseased leaves were at once collected, and a
microscopical examination showed the development of an occasional Gloeosporium
conidium. The leaves were then placed in a damp chamber, and after several days
numerous acervuli had developed and were exuding the typical pink spore masses
in abundance. Reisolations of the fungus were made from these acervuli.

Experiment No. 2 (Mar. 15, 1912). — Conidia (strain 17) from 2-weeks-old com-
meal-agar cultures were mixed with sterile distilled water and sprayed upon the
leaves of four potted seedling trees, which were then left under bell jars for five
days. Two check trees were sprayed wTith sterile distilled water alone, one being
left under a bell jar for five days and the other uncovered. After four weeks it was
noted that discolored areas similar to those noted in inoculation experiment No. 1
had suddenly developed, but observation at ten weeks showed no further progress of
the disease. The last of May, however, when the leaves were getting well matured,
the large, dead, brownish areas were fairly numerous on the leaves of three out of the
four inoculated trees. The check trees which had been kept on the same bench but
somewhat removed from the infected trees were entirely normal. Specimens of the
infected leaves were placed in a damp chamber, where in a few days the Gloeosporium
acervuli were formed.

Experiment No. 3 (Apr. 15, 1912). — Conidial masses from a young com-meal-agar
culture (strain 17) were smeared upon both surfaces of the moistened leaflets of two
potted seedlings, one of which was left under a bell jar for several days. Two check
trees were similarly treated, but not inoculated. No discoloration of the leaves fol¬
lowed for several weeks, but on May 20 several dead, brownish areas were noted on the
leaves of the inoculated tree which had been under a bell jar. These leaves were
placed in a damp chamber and in a week had formed numerous acervuli with the
typical pink spore masses.

Experiment No. 4 (May 1, 1912). — Two of the younger leaves from a Sovereign
pecan were placed in a damp chamber and sprayed with a sterile distilled-water sus¬
pension of conidia from a 2 -months-old com-meal-agar culture (strain 17). The sur¬
face of some of the leaflets was slightly abraded with a needle before inoculation. At
the end of two weeks large dull-brown areas had developed on most of the leaflets, both

1 Shear, C. L. Variation in Glomerella. (Abstract.) Science, n. s., v. 35, no. 891, p. 152, 1912.

Jan. io, 1914

Some Diseases of Pecans

323

where abraded and where the surface was left intact. The largest of these irregular
spots covered as much as half the surface of the leaflets, and numerous perithecia were
forming, though only a few were mature at this time. When 3 weeks old the spots
had increased in area so as to involve most of the tissue, and most of the perithecia
were mature, bearing the typical curved ascospores in abundance. No acervuli or
scattered conidia were noted.

Experiment No. 5 (May 29, 1912). — Twenty young seedling pecan leaves were
placed in damp chambers and lightly sprayed with a distilled water suspension of
conidia and ascospores from a 5-weeks-old cora-meal-agar culture (strain 17). On
the third day small brownish areas had developed here and there over the leaf sur¬
faces. On the eighth day these had nearly covered the leaves, and numerous peri¬
thecia, together with an occasional acervulus, had developed in the dead tissue.
(PL XXXVII, fig. B.) These fruiting bodies occurred in greater abundance on the
lower side of the leaves, but frequently on both upper and lower surfaces. The incip¬
ient perithecia and acervuli developed beneath the epidermis, but later burst through
and became partly superficial.

Experiment No. 6 (Oct. 22, 1912). — Six vigorous but mature leafy pecan shoots
were sprayed with a sterile distilled- water suspension of conidia from strain 123
obtained from diseased nuts, and a similar number with strain 150 obtained from a
naturally infected apple. The shoots were cut under water and the lower ends placed
in bottles of water under slightly ventilated bell jars. Nine check shoots were treated
in the same way, except that they were not inoculated.

At three days many of the inoculated leaves in both sets had begun to show the dead,
brownish areas characteristic of the disease. After seven days these areas had in some
cases involved nearly the whole leaf, with the development of acervuli in moderate
numbers. The check leaves were still green and healthy.

Experiment No. 7 (Mar. 25, 1913). — Distilled- water suspensions of conidia from one
apple strain and three pecan strains of the fungus were sprayed over young seedling
pecan leaves in damp chambers. After three days sample leaves from each set were
collected and prepared for microscopical examination. A small percentage of the
conidia, varying somewhat with the different strains, had sent out germ tubes. Some
of the short hyphse were terminated by appressoria. In one or two cases the germ
tube was traced into the opening of a stoma. This method of infection agrees with that
described by Shear 1 for Gloeosporiums on a wide variety of hosts.

After five days several of the leaflets in each set exhibited typical infection areas
up to 30 mm. in diameter. However, on account of a field trip, no further obser¬
vations were made on this experiment.

NUTS

Experiment No. i (Oct. 22, 1912). — These inoculations were from strain 123,
obtained in October, 1912, from blackened nut shucks sent in from Thomasville, Ga.,
by Mr. C. A. Reed. Terminal shoots bearing healthy green pecans were kindly
furnished by Mr. J. B. Johnson, of Manassas, Va. These shoots were cut under water
to prevent clogging of the vascular system, placed with the cut ends in bottles of
water, and sprayed with a distilled- water suspension of the conidia from this strain.
All were then covered with bell jars ventilated at the base to prevent a too great
stagnation of the air, but at the same time to furnish sufficient humidity to insure
germination of the spores. The check shoots were treated in the same way, except
that they were sprayed with distilled water alone.

Group A consisted of 7 shoots bearing 9 nuts, the hulls of which were punctured
with a sterile needle and sprayed with sterile distilled water. Group B consisted of 2

1 Shear, C, L., and Wood, Anna K. Studies of fungous parasites belonging to the genus Glomerella.
U. S. Dept. Agr., Bur. Plant Indus. Bui. 252, no p.t 18 pi., 4 fig., 1913.

*7°73 — !4 - 4

324

Journal of Agricultural Research

Vol. I, No. 4

shoots treated in the same way but unpunctured. Groups A and B were held as checks.
Group C consisted of 8 shoots bearing io nuts the hulls of which were punctured with
a sterile needle and sprayed with a sterile distilled-water suspension of the conidia.
Group D consisted of 6 shoots bearing 8 pecans which were inoculated like group C,
except that the hulls were not punctured. Group E consisted of io nuts removed
from the shoots, their hulls punctured, and inoculated with a suspension of conidia
as in groups C and D, and then placed in damp chambers.

At the end of three days distinct infection had occurred on all the nuts with punc¬
ture inoculations, the tissue of the hulls being blackened for a radius of 2 to 5 mm.
around the needle punctures. The first checks had the tissue blackened for a radius
of about 0.5 mm. around the needle punctures, while the nonpunctured check and
inoculated nuts at this time showed no evidence of infection. Many of the leaves
on the inoculated shoots were developing small, irregular brownish areas, while the
uninoculated leaves were all green and healthy.

After nine days groups A and B appeared as on the third day. The very narrow
margin of blackened tissue in the punctured checks was due merely to the mechanical
injury to immediately surrounding cells, and no further injury occurred throughout
the experiment. (PI. XXXIII, fig. 1, A.) All the pecans in group C (PI. XXXIII,
fig. 1, C ) were blackened over half to the whole of their surface, and acervuli were
beginning to develop over the dead parts, with an occasional exudation of the pink
spore masses. In group D half of the eight nuts had blackened, and acervuli had
begun to develop, but the others gave no evidence of infection. (PI. XXXIII, fig.
1, B.) In group E all the nuts were blackened, and very numerous acervuli with
exuding spore masses had developed. Reisolations of this fungus were made as strain
144. Plate XXXIII, figure 2 , shows three of the inoculated nuts after further develop¬
ment of acervuli.

apples

Experiment No. i (Dec. 30, 1911). — Five sound Jonathan apples direct from cold
storage were placed in damp chambers and inoculated by needle punctures, two of
them with conidia and three with ascospores inserted directly into the punctures.
Three sound apples were punctured with sterile needles and also placed in damp
chambers. All were kept overnight at 35 0 C., and subsequently throughout the
experiment at laboratory temperatures. Examination after seven days showed a
decay very similar in appearance to bitter-rot around all the inoculation punctures.
The check apples were perfectly sound. These cultures were kept for 10 days, with
a gradual increase in the size of the decayed areas and formation of incipient fruiting
bodies but no spore production.

Experiment No. 2 (Mar. 5, 1912). — Twelve Yellow Newtown apples were similarly
inoculated and placed in damp chambers, one half being inoculated with conidia
and the other half with ascospores from an 8-weeks-old com-meal-agar culture (strain
17). Three apples from each set were held at 28° to 30° C. and three from each set
at laboratory temperature. Six apples punctured with sterile needles and placed in
damp chambers were held as checks, one half at 28° to 3 o° and the other half at labo¬
ratory temperature. At the end of a week the cultures were examined, and all the
inoculated apples had developed a decay apparently identical with bitter-rot, but
the brownish and somewhat sunken spots had increased in size much more rapidly
at the higher temperature. Two weeks later perithecia were forming and the conidia
were developing in considerable numbers, but not in such amount as to give the pink
spore masses characteristic of bitter-rot. The check apples at both temperatures
remained sound to the end of the experiment.

Experiment No. 3 (Nov. 30, 1912). — Sound Jonathan and Yellow Newtown apples
direct from cold storage were inoculated with three strains of Glomerella obtained
from the pecan and with one strain obtained from the apple. The cultures used.

Jan. io, 1914

Some Diseases of Pecans

325

were all young com-meal-agar-slant tubes of the same age and bearing the Gloeospo-
rium stage in abundance. Inoculations were by needle puncture in damp chambers,
and, with the exception of group A, all were held at 28° to 30° C. for 48 hours; after
this they were kept at laboratory temperature. Group A was held at laboratory tem¬
perature throughout.

Group A consisted of three Jonathan apples which were inoculated with strain 17,
isolated from diseased pecan leaves collected at Baconton, Ga., in the fall of 1910.

Group B consisted of three Jonathan and four Yellow Newtown apples inoculated
with strain 123, isolated in October, 1912, from diseased nuts from Thomasville, Ga.

Group C consisted of four Yellow Newtown apples inoculated with strain 125 simi¬
larly obtained from diseased nuts collected at Sumter, S. C., in October, 1912.

Group D consisted of three Yellow Newtown apples inoculated with strain 150,
obtained in October, 1912, from an apple naturally affected with bitter-rot.

Group E consisted of six Jonathan and four Yellow Newtown apples treated simi¬
larly but not inoculated.

On the fouth day typical bitter-rot areas had developed in all the inoculated cul¬
tures. In group A the spots were 1 to 3 mm. , while in B to D they were 3 to 20 mm. in
diameter. It should be stated that the progress of the tissue decay was somewhat
more rapid with strains 125 and 123 than with 150, obtained from the apple itself.
In all cases the pale pinkish white mycelium could be seen protruding in tufts from
the needle punctures, and the dark-colored fruiting bodies were developing. There
were conidia evident at this time. The check apples remained sound. (PI. XXXIV,
fig. A.)

On the eleventh day the spots had increased considerably in size, many of them
being 15 to 20 mm. in diameter and becoming confluent. (PI. XXXIV.) Acervuli
extruding the pink spore masses occurred in dense aggregations over the infected
tissues, being considerably more abundant, however, in strains 123 and 150 than in
the other two. No perithecia had developed as yet, and even after six weeks none
had appeared except on the apples inoculated with strain 123.

Experiment No. 4 (Mar. 21, 1913.) — Sound Yellow Newtown apples direct from
cold storage were inoculated as in experiment 3 with two strains of the fungus obtained
from diseased nuts, one each from the pecan leaf and the apple, and one originally
from the nut but reisolated from an artificially inoculated apple.

On the fourth day bitter-rot areas had developed about the needle punctures in the
case of every strain tested, while the check apples remained perfectly sound. (PI.
XXXV.) The decaying spots rapidly increased in size, and after eight days the
formation of acervuli had begun.

Erom these inoculation tests it would appear that this fungus is para¬
sitic on the leaves of the pecan, though usually not actively injurious
until a certain stage of maturity of the leaves is reached, together with
favorable conditions of temperature and humidity. Field observations
also bear out this point.

The limited inoculation work with the nuts, taken alone, would hardly
justify very definite conclusions, but as far as they go the experience
with leaf, inoculations is duplicated. No artificial infection tests have
been made upon very young nut hulls, but from the field observations
of the last two seasons no evidence of injury during the early part of
the season has been obtained. The disease has come to notice only from
mid-season on until fall. These observations are in line with the sea¬
sonal distribution of bitter-rot as it occurs on the apple.

326

Journal of Agricultural Research

Vol. I. No. 4

The cross-inoculations upon the apple, carried simultaneously with
infections by the apple bitter-rot fungus, show that the pecan fungus
from both leaves and nuts is at least physiologically similar to the Glom-
erella of the apple. The morphological characters will be discussed
later.

Cultural Studies

THERMAL TESTS

Several series of corn-meal-agar cultures were grown for two to three
weeks at temperatures ranging from i° to 35 ° C. As a result of these
studies it was found that no growth would take place below 6°, either
with freshly inoculated cultures or with those in which growth had already
started before incubation. At 70 to 8° the growth was extremely slow,
but gradually increased with rise of temperature until the maximum
for the strains tested was reached at about 30°.

cultural characters

Since the fall of 1910 various strains of the fungus have been grown
on the common culture media, and their appearance under different
conditions is briefly given as follows:

Beef-Agar Slant Tubes. — The colony first appears as a colorless, roundish, submerged
mycelial mass which at ordinary temperatures generally covers the slant in five to
seven days, while one to several groups of acervuli or black perithecia have in the
meantime usually begun to form. The growth is at first entirely submerged and the
surface of the slant presents a smooth glistening appearance. However, after some¬
thing like two weeks a small amount of whitish aerial mycelium makes its appearance
toward the upper edge of the slant. In old cultures this subicle may sparsely cover
the whole surface, while the submerged parts become very dark colored.

Corn-Meal Flasks. — Growth becomes visible in two to three days as a roundish
colony several millimeters in diameter, with sparse, white to pinkish, cottony aerial
mycelium in which are usually scattered a considerable number of dark olive-brown
dots. These dots are found to consist of numerous interwoven hyphae with swollen
and contorted cells in process of uniting to form a pseudoparenchyma. These dark
masses later develop either into acervuli or perithecia.

Corn-Meal- Agar Slant Tubes. — The white to colorless growth is at first submerged
or at the surface. After several days acervuli or perithecia usually begin to form
and a scant whitish aerial mycelium may appear at the edges of the slant. The pink
spore masses are often developed without the formation of a dark-colored stroma,
while in other cases this stroma may be the most conspicuous part of the acervulus.
The perithecia are developed within black carbonaceous masses of mycelium which
may or may not be submerged in the medium. In old cultures parts of the submerged
growth often become olive green to almost black.

Cooked-Potato Cylinders. — Growth first becomes evident through a light-brown
discoloration of the tissue immediately around the point of inoculation, and usually
a whitish aerial tuft of mycelium appears within 24 hours at the center of the dis¬
colored area. This breaking down of the tissue progresses rapidly so that after several
days the whole cylinder becomes involved. The white to pinkish cottony subicle
develops somewhat more slowly, but eventually covers the cylinder and bears the
embedded acervuli or perithecia.*

Jan. io, 1914

Some Diseases of Pecans

327

PEDIGREED CULTURES

Starting with a single ascospore and a single conidium from the same
strain of the fungus, two series of com-meal-agar cultures were carried
for five generations. Each generation was grown for three weeks before
transfers were made for the next succeeding generation, and conditions
of temperature and medium were made as uniform as possible throughout
the 1 5 weeks of the test. Observations in every case were taken at three
weeks. In the first strain ascospores were always used in making the
transfers, while conidia alone were transferred in the second strain.

Ascospore Strain. — Generation 1 had numerous black, carbonaceous, perithecial
groups and no acervuli, though a moderate number of conidia were developed hypho-
mycetously.

In generation 2 the perithecia and acervuli occurred in about equal numbers. In
many cases the black perithecial clusters were surrounded with acervuli which were
exuding pink masses of spores.

Generation 3 exhibited dense black masses of perithecia near the base of the slants
and a considerable number of acervuli which were mostly toward the upper part.

Generations 4 and 5 were similar to the last, except that the two forms were more
uniformly scattered over the surface of the cultures.

Conidial Strain. — Generation 1 had numerous acervuli with exuding pink spore
masses, but no perithecia.

Generations 2 and 3 had numerous perithecial groups and acervuli well scattered
over the cultures, with neither form greatly in predominance.

Generation 4 had numerous pink spore masses along the streak, and perithecial
clusters in moderate numbers near the base of the slant.

Generation 5 had both forms in about equal numbers and well scattered over the
surface of the cultures.

Further cultural studies carried out in the same way as the one described
above have given essentially the same results — namely, that a strain
producing both spore forms will continue to produce both ascospores
and conidia even though one form alone is used in reproduction. Varia¬
tions have occurred from time to time, but these have occurred irregu¬
larly and without continuance. Strains of the fungus from single
ascospores have sooner or later always given rise to both ascigerous and
conidial forms. However, some conidial strains have been obtained from
the host which, after two years in culture, still produce only the conidial
form. It would thus appear that there are conidial strains of the fungus
which have lost the power of developing the perfect stage or which at
least have not met with the proper inciting conditions.

Morphology and Taxonomy

The perfect stage has been noted less frequently than the conidial stage,
but nevertheless the perithecia have been occasionally found on both
leaves and nuts. The first evidence of perithecial formation is seen in a
plexus of pseudoparenchyma tissue made up of more or less isodiametric
fungous cells developed in the decaying tissues beneath the epidermis.
This finally develops into the mature perithecium which ruptures the

328

Journal of Agricultural Research

Vol. I, No. 4

Fig. 6. — The anthracnose fungus upon corn-meal agar.

X84; B , conidia, X400; C, ascus, X400.

A , Acervulus,

epidermis and becomes partially superficial. The mature fruiting body
is nearly spherical, but is papillate and occasionally short beaked. From
several hundred measurements it has been found to vary from 80 to
250 p in the longest diameter, with the majority lying between 150 and

200 fx. The perithecia

are black and carbo¬
naceous, and in cul¬
ture several are usu¬
ally developed within
a single carbonaceous
stroma.

The 8-spored asci
vary considerably in
size and shape, but are
usually cylindrical-
clavate. (Fig. 6.) The
extreme measure¬
ments found were 45 to 80 by 9 to 12.5/1, the majority lying within the
limits of 55 to 80 by 10 to 1 1 fi.

The ascospores are unicellular (rarely with a cross partition), oblong,
slightly tapering toward both ends, and usually curved. (Fig. 6.) The
extreme measurements found were 12.5 to
29 by 3.5 to 6{i, the majority lying about
midway between the two extremes as shown
in the accompanying graph (fig. 7) drawn
from measurements of 1 50 spores of a single
strain taken at random and all developed
in corn-meal-agar culture. Measurements of
other strains both from culture and from the
host have come within these limits.

The acervuli have been of much more com¬
mon occurrence on the host. (Fig. 6.) In
their early stages they are scarcely to be dis¬
tinguished from the perithecia, but the pro¬
duction of the characteristic pink spore
masses soon differentiates them even macro-
scopically from the perfect stage. The
production of setae has been found of fre¬
quent though by no means of general occur¬
rence, and to vary even within a single strain.

4 j-tf
AT/C&OM&

Fig. 7. — Diagram showing ascospore
measurements of the anthracnose
fungus. Ay Length of 150 ascos¬
pores; By width of 150 ascospores.

The conidia are ovate to oblong, with blunt, rounded ends (fig. 6)
(occasionally somewhat dumbbell-shaped). Both on the host and in
culture they have been found to develop hyphomycetously, as well as
in acervuli. The measurements taken from several strains on the host

Jan. io, 1914

Some Diseases of Pecans

329

and in culture ranged within the limits of 11 to 22 by 3.8 to 7.6/*. The
accompanying graph (fig. 8) shows the measurements of 1 50 conidia taken
from the same strain and under the same conditions as those noted above
for the ascospores. The conidia have frequently been found to develop
appressoria as described by various authors for the apple bitter-rot
fungus.

From the general pathology and temperature relations, the cross¬
inoculation and cultural studies, and finally from the morphology of the
pecan fungus there can be no doubt of its specific connection with Glo-
merella cingulata (Stonem.) S. and v. S.,the fungus causing bitter-rot of
apple, ripe-rot of grape, and anthracnoses of a
wide range of hosts.

In several instances Glomerella perithecia
have developed upon pecan leaves scattered
among the densely gregarious pycnidia of a fun¬
gus which has since proved to be Phyllosticta
convexula Bubdk.1 The spores of the latter are
almost bacillar in size, measuring 1.5 to 2 by
1 jc, while in many cases only a few pycnidia upon
a leaf matured in damp chamber, so that mor¬
phologically most of these fruiting bodies were
similar to the immature perithecia of Glomerella.

Furthermore, an examination of the fruiting
bodies from type specimens of Sphaerella con¬
vexula (Schwein.) von Thum.2 (, Sphaeria con¬
vexula Schwein.3) shows them to be morpho¬
logically similar to those of Phyllosticta convex¬
ula. The original brief diagnosis of Sphaerella
convexula was from immature material and
without description of asci or ascospores. Sim¬
ilar material has been collected by the writer
at various points in South Carolina, Georgia,
and Florida, including one of the type localities
of the species.

Glomerella perithecia have been developed in a damp chamber, not
only upon disinfected pecan leaves exhibiting the typical anthracnose
blotches and among the pycnidia of Phyllosticta convexula , but also fre¬
quently upon leaves apparently healthy in every respect, showing the
wide distribution of the former fungus and its ability to hibernate on
the living host until the occurrence of conditions favorable to its further
growth.

M/CffOA/S

Fig. 8. — Diagram showing conidial
measurements of the anthracnose
fungus. A , Dcngth of 150 con¬
idia. B , width of 1 50 conidia.

1 Bubdk. Franz, Binige neue Pilze aus Nord America. Jour. Mycol. , v. 12, no. 82, p. 52, 53 , 1906.

2 Saccardo, P. A. Sylloge Fungorum. v. 1, Patavium, 1882, p. 494.

3 Schweinitz, L. D. von. Synopsis fungorum in America boreali media degentium. Trans. Amer. Phil.
Soc. n. s., v. 4, p. 224, 1834.

Berkeley, M. J. Notices of North American fungi. Grevillea, v. 4, no. 32, p. 154, 1876.

33°

Journal of Agricultural Research

Vol. I. No. 4

From these facts it seems entirely possible, if not indeed probable, that
the type fungus of Schweinitz and Von Thiimen was in reality identi¬
cal with Phyllosticta convexula Bub&k and that the immature perithecia
were those of the fungus at present known as Glomerella cingulata .

KERNEL-SPOT

[Caused by Coniothyrium caryogenum , n. sp.]

HISTORY AND DISTRIBUTION

Fortunately this disease has hitherto been of only occasional occurrence.
In the fall of 1907 infected kernels were received by Mr. W. A. Orton,
Pathologist in Charge, Cotton and Truck Disease and Sugar-Plant Inves¬
tigations, Bureau of Plant Industry, from Minden, Ea., accompanied by
the f ollowing statement :

The disease of the pecans is not confined to any one tree or variety. * * * For
six years they have contained the blight, growing worse each year, until I think that
next year there will not be a single good one (nut) found among them. * * * I
have never heard of this disease from anyone else. All our trees are infected.

From these specimens Orton isolated a fungus with brown, septate,
branched mycelium. No further studies were carried out to determine
its parasitism or further cultural characters,1 but examination of these
specimens has shown them to have the symptoms associated with the
kernel-spot.

No other definite reports of the kernel-spot prior to 1910 have come
to notice, but during the last three years occasional specimens from a
number of Southern States have been received by the Office of Fruit-
Disease Investigations. Among these communications the only report
of serious injury was from a point in southern Georgia, where in the fall
of 1 91 1 most of the nuts on a large seedling tree were rendered unfit for
consumption. From this source were obtained the fungous cultures
used in the present inoculation work. Since there were no nuts on this
tree the following year, field studies as to time and manner of infection
could not be carried further.2

SYMPTOMS OP THE DISEASE

Externally there is no evidence of infection and it is only upon freeing
the kernel from the shell that the diseased condition becomes apparent.
(PL XXXVII, fig. E.) On the surface of the kernel the spots are dark
brown to almost black and often slightly sunken. They are in general
irregularly roundish with a fairly definite margin and several millimeters
in diameter. Internally the diseased tissue extends in an approximate

1 Orton. W. A. From unpublished notes.

3 If a tree is badly infected, the nuts should be gathered and burned, in order to lessen the chances of
further spread of the disease.

Jan. 10, 1914

Some Diseases of Pecans

33i

hemisphere beneath the dark-colored spot. The central part of this
hemisphere is dry and pithy, slightly discolored, and surrounded by a
definite dark-brown layer separating the diseased from the healthy parts.
The tissues are slightly disorganized, but are not softened or entirely
broken down. A bitter taste is imparted to the kernel. Microscopically,
the fungus is found to enter the cells of the kernel, where the hyphae
become partially broken up into their constituent cells. Outside the
dark-colored boundary layer the tissues of the kernel are seen to be
slightly discolored, although no signs of the fungus itself are seen here.
It seems probable that enzyms or toxins (or both) excreted by the fungus
may diffuse out into the healthy cells of the host and by partial digestion
prepare the way for the entrance and progress of the parasite.

MY COLOG IC AL AND PATHOLOGICAL STUDIES
Isolation op the Fungus

The affected pecan kernels received in the fall of 1911 from Thomas-
ville, Ga., were washed for five minutes in a solution of bichlorid of mer¬
cury (1 : 500), and in distilled water. Small pieces of the diseased internal
tissue were then cut out under sterile conditions and transferred to
Petri dishes of melted beef agar. Yellowish bacterial colonies resulted
from two of the transfers, but a constant fungous type developed from
all the others. The bacteria and the fungus were isolated and carried
in pure culture for the following cultural and inoculation studies.

Inoculations

In all the inoculations the kernels were freed from the shells under
semisterile conditions and placed upon sterile, moist filter paper in
Petri dishes. Under these conditions the pycnospores or mycelium from
a pure culture were placed upon the kernels either with or without
slight abrasion of the surface. The checks were treated in a similar
manner but without inoculation.

Experiment No. i (Jan. 15, 1912). — The kernels from several stratified nuts were
placed in Petri dishes and inoculated by slight abrasion (1) with spores of the fungus
(strain 99), and (2) with the yellow bacteria (strain 101), while the kernels in the
third dish (3) were merely abraded with a sterile scalpel. After eight days typical
symptoms of the kernel-spot had developed in the first culture. (PI. XXXVII,
fig. D.) The bacteria in the second culture had made a slight growth, causing an
irregular softening of the superficial tissues, but without discoloration or other resem¬
blance to the kernel-spot. The check cultures were entirely sound.

Experiment No. 2 (Jan. 25, 1912). — Kernels of well-cured Stuart pecans were inoc¬
ulated with the fungous spores, six kernels upon the uncut surface, and eight with
a slight abrasion. Four kernels were held as checks. After 12 days typical spots
had formed upon half of the first set and on all of the second set of kernels. Of the
checks, two kernels were perfectly sound, the third exhibited a slight bacterial soften¬
ing at one end, and the last was softened throughout by a growth of Penicillium

33 2 Journal of Agricultural Research voi.i, No. 4

glaucum . In the last two cases the injury was similar in no particular to the
kernel-spot.

Experiment No. 3 (May 13, 1912). — Three Petri dishes containing four to eight
kernels from cured pecans were inoculated by placing macerated pycnidia upon the
uninjured surfaces. A fourth Petri dish was held as a check. After seven days it
was noted that infection had taken place at every point of inoculation in the first
two cultures. In the third, two kernels had become infected with the kernel-spot,
but the remaining two were entirely softened by bacterial contamination. In the
check Petri dish two kernels were sound and two were contaminated and softened
throughout by Botrytis cinerea. In no case was the injury by contamination similar
to the disease under investigation.

Experiment No. 4 (Nov. 20, 1912). — Ten kernels of newly harvested pecans were
inoculated with macerated pycnidia and without abrasion of the surface skin. A
similar number of kernels were held as checks. After nine days, 8 out of the 10 inocu¬
lated kernels had developed the disease. The checks were sound, except for two or
three kernels which had softened and yellowed throughout from bacterial contami¬
nation.

Experiment No. 5 (Dec. 25, 1912). — Eight to ten partially cured kernels of each of
the following varieties were inoculated with macerated pycnidia by a slight abrasion
of the surface: Schley, Curtis, Nelson, Teche, Alley, Pabst, and Van Deman. Check
kernels of each variety were carried throughout the experiment. Similarly, Teche
and Van Deman kernels were inoculated with the two strains of yellow bacteria
(strains 100 and 104) . After five days the bacterial inoculations had caused a softening
of the tissues throughout, but there were no evidences of the kernel spot. The fun¬
gous inoculations had in nearly every case taken, and spots typical of the disease both
externally and internally had developed, regardless of variety. The checks were
sound, except for an occasional contamination with Botrytis cinerea, which had caused
a general softening of the tissue. Reisolations of the fungus were made from each of
the varieties inoculated, and one of these strains was used in the next experiment.

Experiment No. 6 (Jan. 6, 1913). — Three Petri dishes of partially cured Van Deman
kernels were inoculated upon the slightly abraded surface with macerated pycnidia
of the fungus reisolated from artificial inoculation in experiment No. 6. Three dishes
of kernels were similarly inoculated with a Sphaeropsis obtained from old decaying
pecan hulls, while two were held as checks. Observations after five days showed
infection with typical symptoms in every case of inoculation with the kernel-spot
fungus. The Sphaeropsis had caused a general breakdown and softening of the tissues,
with slight discoloration, but with no symptoms like the disease in question. The
checks all remained sound and free from infection of any kind.

No opportunity for field inoculations has presented itself without the
accompanying danger of introducing or spreading the disease, and hence
the infection tests have been entirely confined to the laboratory. How¬
ever, the characters of the disease are so definite and the results of the
inoculation work on kernels in the laboratory have been so largely positive
that the fungus tested (strain 99 and its reisolation) may now be legiti¬
mately regarded as the cause of the kernel-spot. The general disorgani¬
zation and moist softening of the tissues brought about by the bacteria
and by the Sphaeropsis, Botrytis, and Penicillium fungi was entirely
different in appearance and result from the disease under investigation.
Individual infections of the latter occur within limited and well-defined
boundaries and, though giving a pithy consistency to the diseased parts,
never cause a moist softening of the injured tissue.

Jan, io, 1914

Some Diseases of Pecans

333

Cultural Studies

As grown upon corn-meal agar the optimum temperature for the fungus
was found to lie around 20° C. (68° F.). No growth took place below 20
or above 370. The rate was slow at 40, but gradually increased up to
the optimum, and decreased somewhat more rapidly in rate above that
point. At 350 a slight but abnormal growth occurred for a few days,
but at the end of the 3-weeks' test, incubation at the optimum tem¬
perature failed to show any further signs of life in these cultures.

Upon corn-meal agar the submerged growth varies but little from a
sepia brown, while the aerial mycelium shows gradations from that to
whitish. Usually a large number of dark-sepia to almost black pycnidia
are formed upon this medium. The mycelium is straight and but little
branched, with gradations from brown to almost hyaline.

On corn-meal flasks the colonies appear very much as upon the corn-
meal agar, though the aerial mycelium is usually much more luxuriant
and cottony, becoming, however, somewhat felted with age. Pycnidia
are developed in large numbers.

On cooked-potato cylinders the colonies are brown ocher, varying also
to a slightly darker shade. The surface is smooth and glistening, be¬
coming somewhat wrinkled with age. No aerial mycelium or pycnidia
have been observed on this medium. The cells of the hyphae differ from
those grown upon corn-meal agar in being more nearly isodiametric, with
thicker and somewhat bulging walls. The mycelium possesses but few
side branches, and the color varies from pale brown to almost hyaline.
In cultures several weeks old the whole potato cylinder becomes some¬
what softened and turns brown, but no fungous mycelium is found except
near the surface. The starchy contents of the potato cells become largely
digested, though the walls of the deeper lying cells remain intact except
for the breaking down of the middle lamellae.

Upon synthetic agar the growth is brown ocher to sepia in the older
and drier parts. The surface growth often becomes more or less wrinkled
and moist-mealy in appearance in older cultures, while a pale brown
to whitish aerial mycelium may or may not develop. Microscopically
the hyphae very much resemble those developed upon the potato cylin¬
ders, but the thickening and bulging of the walls is often much more
apparent. Indeed, the hyphae frequently break up into their constituent
cells, and it is this behavior that gives the moist-mealy appearance to
some cultures.

Morphology and Taxonomy

The study of this fungus in culture and upon the host has shown it to
conform in characters with the genus Coniothyrium. However, no mem¬
ber of this genus has been found hitherto reported on the pecan or any
nearly related host. It thus becomes necessary to give the fungus a
new specific value until cultural and cross-inoculation work can establish

334

Journal of Agricultural Research

Vol. I, No. 4

its connection with a previously described Coniothyrium occurring upon
some widely differing host. An ^numeration of the characters thus far
observed is given below.

It should be stated that the pycnidia have been observed mostly in
culture, their formation on the host having been confined to the extracted
kernels in a damp chamber. In the latter case their development has
has taken place at or near the surface of the kernel and often accompanied
by a thin subicle of brown to whitish hyphae.

Coniothyrium caryogenum, n. sp.

Upon pecan kernels Coniothyrium caryogenum causes dark-brown, irregularly round¬
ish surface spots with a hemisphere of pithy tissue beneath, which is surrounded by
a brownish layer of host cells.

Mycelium brown, sometimes almost hyaline where not submerged, septate, slightly
branched, straight or within the host cells often separating into the constituent hyphal
cells which are then more or less swollen and thick walled.

Pycnidia roundish, osteolate, thin walled, dark brown, about 200 to 2 50 g. in diameter.

Sporophores short and indistinct. Spores pale brownish, elliptical, i-celled, 2.5 to
3.6 by 1.8 to 2 fi.

Habitat. — Kernels of Carya illinoensis (Wang.) K. Koch. Type specimens from
large seedling tree belonging to Mr. James R. Vann, Thomasville, Ga. Specimens also
received from Raleigh, N. C.; Baconton, Ga.; Monticello, Fla.; Minden, La., and
other points in the pecan belt, including Texas.

CROWN-GALL

[Caused by Bacterium tumefaciens Sm. and Town.]

So far as known, the crown-gall has not hitherto been published as
occurring on the pecan from natural infection. However, in the fall
of 1909 specimens of young trees affected with both the hard and soft
types of galls (PI. XXXVI) were received from a nursery in Mississippi
with the statement that about 0.1 per cent of the stock in the nursery
was infected. The disease has also been observed by the writer at one
point in northern Florida. But, since these two localities have fur¬
nished the only cases reported, it may be considered as of very rare occur¬
rence upon this host.1

On the pecan the tumors occur not only at the collar of the tree but
several inches higher up on the stem and also on the roots. The greater
prevalence of the disease near the surface of the ground is explained by
the fact that the parasite first enters the host tissues through wounds.
Thus, the process of grafting and the subsequent treatment of the stock
readily furnish conditions requisite for infection and further develop¬
ment. The typical appearance of the disease may be inferred from the
name; the galls at first consist of a succulent growth of the young host
cells thrust out from the cambium layer in the form of a tumor which
may attain a considerable size. With age the surface becomes much

1 The only practical method of control hitherto employed consists in rigid nursery inspection. Obviously,
no trees showing the disease should be planted, even though the pecan does not appear to be as seriously
affected as many other plants.

Jan. io, 1914

Some Diseases of Pecans

335

roughened and darker in color and the interior tissues are then more or
less distorted and hardened. Often the interior assumes a distinctly
woody texture, and a roughened bark develops over the surface to form
the “hard-gall” type. With the development of roots from the tumor
tissue the “ hairy-root ” type appears, but this form has not been ob¬
served on the pecan.

EXPERIMENTS WITH THE CROWN-GAEE ORGANISM

Soft galls from the Mississippi nursery (December, 1909) were left for
five minutes in a solution of corrosive sublimate (1 *.500) and washed in
sterile distilled water. Small pieces of the abnormal tissue were then
removed under aseptic conditions from points just under the surface and
near the edge of the galls, and beef-agar cultures started by the ordinary
poured-plate method. In from three to eight days the circular and some¬
what opalescent colonies of the organism appeared, but were much more
abundant in cultures started from the extreme base of the young soft galls
near the juncture between the diseased and healthy tissues. Transfers
were made to beef-agar slant tubes, and with one of the strains thus
obtained the following inoculation tests were made.

Experiment No. i (December, 1910). — Six table beets were inoculated by needle
punctures from young beef-agar cultures of the bacteria, while a like number of beets
were punctured with sterile needles and held as checks.

After five weeks, examination of the inoculated beets showed the development of
typical galls, 3 to 10 mm. in diameter, at most of the needle punctures, while the
checks showed no signs of infection.

Experiment No. 2 (Jan. 12, 1911).— Four potted pecan seedlings were inoculated
by scalpel punctures at the crown from 4-day-old beef-agar cultures, and the soil was
replaced around the base of the tree to preserve the moist condition. Four other
seedlings were treated in the same manner, except that no bacteria were introduced.
The trees were all dormant at this time and remained in this condition until the
latter part of March, when, with the exception of one of the inoculated trees which
died from other causes, all pushed out their foliage in the normal manner.

Examination in June showed a tumor several millimeters in diameter at the crown
of one of the inoculated trees and an apparently incipient infection on a second. All
the other trees had completely healed over, so that the location of the punctures
could scarcely be made out. On September 12, eight months after inoculation,
well-developed galls were found at the crown of two out of the three remaining
inoculated trees. The check trees, together with 59 other pecan seedlings in the
same greenhouse, showed no indications of the disease.

Since these brief studies with the parasitic organism were carried out
merely to indicate the connection of this disease of the pecan with the
well-known crown-gall, no further inoculation and cultural tests were
made. However, cultures of the bacterium were submitted to Dr.
Erwin F. Smith, of the Bureau of Plant Industry, who obtained similar
results in inoculation experiments and further verified the identity of
the organism with Bacterium tumefaciens Sm. and Town., the cause of
crown-gall of plants.

336

Journal of Agricultural Research

Vol. I, No. 4

SUMMARY

The nursery-blight is a serious disease of young trees, but is rarely
found to be injurious in orchards. Its distribution corresponds closely
with that of the host. The casual fungus, Phyllosiicia caryae Peck,
attacks only the leaves of the pecan. Infection first becomes evident
through the formation of tiny circular, dark-brown spots, which increase
gradually in size and finally become grayish white in the center of the
upper surface and usually blackish throughout on the lower. Entire
defoliation of young trees sometimes takes place. Spraying with Bor¬
deaux mixture has proved a very effective method of control. Since the
disease is primarily a nursery trouble, the question of disease resistance
would not be applicable in this connection. All attempts at pure-culture
inoculation have been successful. A combination of high humidity and
temperature seem best to favor the spread of the disease. The fungous
mycelium ramifies through the intercellular spaces above the lower
epidermis and throughout the mesophyll tissue. Pycnidia are few on
the living leaves, but are produced in abundance on some culture media.

The brown leaf-spot usually causes very little injury, but is widely
distributed and occasionally during wet seasons some defoliation may
result. The fungus Cercospora fusca, emend, sp., causes dark reddish
brown spots of uniform color on both leaf surfaces. These are at first
somewhat angular in outline as bounded by the veins of the leaf, but
may later become roundish and more indefinite in their margins. There
appears to be little difference in resistance to this disease among the
varieties now commonly planted. The rather limited observations upon
the effect of Bordeaux mixture were favorable to the control of the
disease. Pure-culture inoculations were highly successful, giving the
typical disease symptoms. The temperature relations were very sim¬
ilar to those of the nursery-blight. The mycelium is largely inter¬
cellular in its growth, but aggregations of fungous cells break through
the upper epidermis to bear the pale tawny conidial clusters, and a
creeping surface mycelium sometimes occurs. True spore formation
has not taken place in culture.

The pecan anthracnose is well distributed, but hitherto has not
usually been very serious at any one point. It has been shown by cul¬
tural and cross-inoculation work to be due to Glomerella cingudata
(Stonem.) S. and v. S., the fungus causing bitter-rot in apples. On the
leaves infection causes the formation of irregular reddish to grayish
brown blotches varying greatly in size and eventually often covering the
whole leaf. On the huts the blotches are also irregular in outline, but
nearly or quite black and often slightly sunken below the surrounding
healthy tissue. The production of acervuli and perithecia occurs under
suitable conditions of temperature and humidity. The problem of control
is largely in the tentative stage, though from the work of Scott and others

Jan. io, 1914

Some Diseases of Pecans

337

on the apple bitter-rot it is thought that Bordeaux mixture will prove
effective. Some indications of difference in varietal resistance have been
observed. High temperature and humidity furnish the optimum condi¬
tions for growth and spread of the disease, as is the case with the bitter-
rot of apple.

The kernel-spot is fortunately rare, but on this account the present
study has been largely confined to laboratory and greenhouse work.
The fungus Coniothyrium caryogenum , n. sp., causes the development of
dark brown to almost black surface spots upon the kernel. Internally the
diseased tissue extends in an approximate hemisphere beneath the dark-
colored spot and is pithy in texture and bitter to the taste. Pure-culture
inoculations have been largely successful. The optimum temperature for
growth was found to be about 70° F. The mycelium enters the cells of
the kernel, where it is often more or less swollen and broken up into its
constituent cells. Pycnidia have been produced abundantly in culture,
but on the host only on the extracted kernels in a damp chamber.

Crown-gall has been found on the pecan in northern Florida and
southern Mississippi. It is similar in appearance to the well-known
crown-gall of plants and has been shown by pure-culture and inoculation
work to be due to the same organism, Bacterium iumefaciens Sm. and
Town.

DESCRIPTION OF PLATES

Plate XXXIII. Fig. i. — Pecan nuts infected with the anthracnose fungus by spray¬
ing with a distilled water suspension of conidia, showing the ap¬
pearance nine days after inoculation. Natural size. Fig. A. —
Four check nuts, two punctured with sterile needle and two
unpunctured. Fig. B. — Four nuts inoculated upon the unpunc¬
tured surface of the hull. Fig. C. — Four nuts inoculated after
puncturing the surface of the hull with a sterile needle.

Fig. 2. — Three of the infected nuts shown in figure i after further
development of the acervuli. X i

XXXIV. Yellow Newtown apples infected by needle puncture with conidia
of the anthracnose fungus from pecan and apple, showing appear¬
ance four days after inoculation. One-half natural size. Fig.
A . — Check apples punctured by sterile needle. Fig. B. — Apples
infected by needle punctures with strain x 50 from the apple. Fig.
C. — Apples infected with strain 123 from a diseased pecan hull.
Fig. D. — Apples infected with strain 125 from a diseased pecan
hull.

XXXV. Yellow Newtown apples infected by needle puncture with conidia of
the anthracnose fungus from pecan and apple, showing appearance
four days after inoculation. Two-thirds natural size. Fig. A. —
Check apple punctured by sterile needle. Fig. B. — Apple
infected with strain 125 from the pecan nut. Fig. C. — Apple
infected with strain 123 from the pecan nut. Fig. D. — Apple
infected with strain 150 from the apple. Fig. E . — Apple infected
with strain 146 from the pecan leaf. Fig. F. — Apple infected
with strain 158, a reisolation of strain 125 after passage through
the apple.

XXXVI. Crown-gall (caused by Bacterium tumefaciens Sm. and Town.) on
pecan nursery trees from southern Mississippi. Natural infection.
Two-thirds natural size. Fig. 1. — The soft type of gall. Fig.
2. — The hard type of gall.

XXXVII (colored). Fig. A. — A pecan leaflet infected with the brown leaf-
spot fungus ( Cercospora fusca , emend, sp.) from pure culture.
Fig. B . — A pecan leaflet infected with the anthracnose fungus
(Glomerella cingulata (Stonem.) S. and v. S.) from pure cul¬
ture. Fig. C. — View of upper surface of a pecan leaflet recently
infected with the nursery-blight fungus ( Phyllosiicia caryae
Peck) from pure culture. Fig. D. — A pecan kernel infected
with the kernel-spot fungus ( Coniothyrium caryogenum, n. sp.)
from a pure culture, showing the appearance eight days after
inoculation. Fig. E. — A pecan kernel with the kernel-spot from
natural infection. Fig. F. — A pecan nut infected with the an¬
thracnose fungus from pure culture. Fig. G . — The nursery-
blight fungus upon synthetic agar after two weeks. Fig. H. —
The nursery-blight fungus on corn-meal agar after two weeks.
Fig. I. — Views of the upper and lower surfaces of pecan leaflets,
showing an advanced stage of the nursery-blight. Natural infec¬
tion. Fig. /. — The brown leaf-spot fungus on synthetic agar
after four weeks. Fig. K. — The brown leaf-spot fungus on corn-
meal agar after four weeks. (All figures are natural size.)

(338)

Plate XXXIII

Plate XXXIV

Some Diseases of Pecans

PlateXXXVII

Journal of Agricultural Research

Vo I I, No 4

A TWIG BLIGHT OF QUERCUS PRINUS AND RELATED

SPECIES

By Deela E. Ingram,

Scientific Assistant, Investigations in Forest Pathology , Bureau of Plant Industry

INTRODUCTION

A twig blight of the chestnut oak (j Quercus prinus L.) was first reported
to the Office of Investigations in Forest Pathology on May 31, 1911, by
Drs. Metcalf and Spaulding, of that office. Specimens were collected
and sent in from York, Pa. Since that time the disease has been reported
and diseased specimens have been received from various points through¬
out Virginia, West Virginia, Maryland, Pennsylvania, New York, and
Connecticut. It is not possible at this time to determine definitely
the exact range of the blight, as sufficient data have not been obtained.
Nothing is known regarding the origin, age, or directions of distribution
of the causal fungus, but apparently it will seriously lower the silvi¬
cultural status of the chestnut oak.1

EFFECT ON HOST

This blight is primarily a disease of the chestnut oak, but occasionally
the American chestnut ( Castanea dentata (Marsh) Borkh.) and the white
oak (1 Quercus alba E.) are attacked. Inoculations in the greenhouse
have proved that a number of other species of oak are also susceptible.

Trees of all ages and sizes may be attacked, but usually only the
small branches of the larger trees are affected. In some cases where
young saplings are attacked the whole tree is killed outright. On the
affected twigs the leaves wither suddenly without yellowing, gradually
shrivel, and turn a chocolate brown. This browning of the leaves and
twigs gives the tree the appearance of the well-known fire-blight of the
pear and the apple. (PI. XXXVIII.) The fungus often stops at the
point where the secondary shoots join the main stem, and, as a result,
the affected twig may rot at the base and fall off. On the diseased twigs
are numbers of small black pycnidia erumpent through the bark. These
are sometimes arranged singly and sometimes grouped. Careful sections
were made of leaves from diseased twigs brought in from the field, but
no mycelium could be found in the tissues. Cultures were also made,
but nothing developed. A microscopic examination of a transverse
section of the wood reveals the presence of abundant mycelium in the

1 Apparently the only practical method of control for individual trees is cutting back the young twigs
several inches below the darkened portion. However, under forest conditions no practicable means of
control is known.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(339)

Vol. I, No. 4
Jan. 10, 1914

340

Journal of Agricultural Research

Vol. I, No. 4

tracheary tubes and throughout the cells of the inner and outer bark.
A study of the distribution of the mycelium in the twigs of different ages
and the relative amount present in the wood and cambium of the dis¬
eased twigs was not undertaken.

MORPHOLOGY OF FUNGUS

Soon after the leaves wither on the affected twigs, small papillae begin
to form under the bark, which in the course of a few weeks break through

Fig. i. — Diplodia longispora: A section of a pycnidium.

in the form of the small, black pycnidia mentioned above. These are
globose to subglobose in shape, very distinctly ostiolate, and dark brown
to black in color. In size they vary from 95 to 145^ in diameter. In
cross section (fig. 1 ) the wall of the pycnidium is made up of practically
two parts: The outer, dark carbonlike layer
and an inner membranous layer of typical
fungous cells. These cells have a decidedly
purplish tinge, merging into hyaline as the
s orogenous layer is reached.

The spores both on the host and in culture
are oval or ovoid (fig. 2, A)} often tapering
somewhat at one end, densely granular, often Fig. 2— Diplodia longispora: stages

Very thick-walled, averaging about 29 x Iljtt phoma stage; Bt Diplodia stage;
in size. At first the spores are hyaline and c* Diplodia spore with two
continuous, but after some time (fig. 2, B) they septa*
take on a yellowish tinge and finally become dark brown in color and
1 -septate. Rarely the septum forms in the hyaline spores before the
color begins to change, but this is not usually the case.

The spores are borne singly on rather short, broad conidiophores,
interspersed with numerous filiform paraphyses, and are ab join ted from
the tip at maturity by a constriction near the end of the conidio-
phore. The conidiophores may become long and filiform in artificial
media. The liberation of the spores from the pycnidium is effected in
damp weather by means of distinct cirrhi, or threads, forced out through
the ostiole of the pycnidium.

Jan. io, 1914

Twig Blight of Quercus Prinus

34i

INOCULATIONS

Inoculations were carried on in the greenhouse on Castanea dentata
(Marsh) Borkh. and on a number of related species of oak — Quercus
prinus L., Q . minor (Marsh) Sarg., Q. gambelii Nutt., Q. lobata Nee.,
Q. texana Buckl., Q. virginiana Mill., Q. alba L., and Q. rubra L.

At the time of the first inoculations small potted trees were used, and
these were mostly in their dormant winter condition.

The inoculations were made by sterilizing the bark with a mercuric-
chlorid solution, making an incision through the bark with a sterile
scalpel, and carefully inserting a portion of the mycelium. The wound
was then carefully protected by a small portion of sterile cotton. Check
plants were kept of all inoculations made.

The first inoculations were made on chestnut on October 24, 1911, as
no chestnut oak was then available. In seven days the inoculated twigs
showed a darkened area in both directions from point of infection.
After one month the twigs were entirely dead from the point of inocula¬
tion outward, and the small papillae of the fungus were visible just
beneath the epidermis. The checks healed normally.

A pure culture of the fungus was obtained from a portion of a diseased
twig that was brought into the laboratory. From this culture inocula¬
tions were made on November 11, 1911, as follows:

Four inoculations on Quercus lobata , two by means of an incision in
the bark and two by simply binding on portions of mycelium in agar
with sterile cotton; three inoculations on twigs of Castanea dentata; and
three inoculations on leaves of Q. prinus . One leaf of 0. prinus was
inoculated on the upper surface through the wounded epidermis and one
on the lower; on the other, the mycelium was simply spread over the
unwounded surface.

An examination after one week showed inoculated twigs of Quercus
lobata blackened for about half an inch each way from the point of inocu¬
lation; the chestnut was slightly darkened. The wounded leaves of
Q. prinus , both inoculations and checks, were somewhat yellowed, but
these subsequently recovered; the unwounded inoculated leaf was
normal; and all were uninjured by the fungus. After some weeks these
leaves were brought into the laboratory and careful sections made, but
no trace of the mycelium could be found in the tissues.

In all, a total of over 50 inoculations were made in the greenhouse to
test the susceptibility of different species of oak and to find the time
when infection most readily takes place. Of these inoculations 50 per
cent were effective. The twigs darkened and the leaves withered, show¬
ing the presence of the fungus. In some the infection did not extend
more than a few inches from the tip, but in others the whole twig died.
In but few cases, however, did the fungus make its way into and up the
main body of the plant.

342

Journal of Agricultural Research

Vol. I, No. 4

Quercus gambelii proved to be the most susceptible when inoculated,
and Q . lobata the second; Q . alba and Q. rubra were slower in showing
the effects of the fungus; while Q. virginiana and Q . texana were not
affected.

In a number of cases the plant was in a dormant condition when
inoculated and seemed not to be affected by the fungus, but at the
leafing-out season no leaves were formed from the point of inoculation
outward to the end of the branch (PI. XXXVIII), while the other part
of the plant put out leaves and grew in a normal manner. After inocu¬
lation the twig darkened slightly, but no further external development
took place. No pycnidia were formed as usual, even after the growing
season commenced.

The failure of part of the inoculations was probably due to the time
of inoculation, as it was found that the twigs are the most susceptible
when the new shoots are just coming out. Practically all the inocula¬
tions made at this time were effective, but after two weeks from the
time of leafing-out the susceptibility lessened greatly, only a small
percentage made from that time on having any effect.

In some cases after the dying of the tip the branch put out new shoots
below and apparently overcame the injurious effect of the fungus.
Inoculations from cultures of the mature stage developed somewhat
slower than those from the Macrophoma stage.

The inoculations of Quercus prinus in the field were more conclusive.
Fifty inoculations were made on May 8, 1912, and 28 of these were
effective. Twenty-six were made in the usual manner by a slight inci¬
sion in the bark and the inserting of a portion of the mycelium into
the wound. Fifteen were made by inoculating with spores. Of the
latter, 10 were made by placing the spores in the incision and 5 by
puncturing the bark with a needle and spraying the injured part with
spores suspended in corn-meal infusion. Four inoculations were made
by binding the mycelium on the surface of the uninjured twigs. Five
leaves were pricked slightly with a needle and sprayed with the spores —
one on both upper and lower surface, two on the upper surface only,
and three on the lower only. Checks of both leaves and twigs were
treated in the same manner. The leaves all healed normally and were
not affected by the fungus. Three of the twigs that were sprayed with
spores withered and died, while the two others healed normally. Four
of the twigs inoculated with spores by a slit in the bark withered from
the point of infection out to the tip; the others were uninjured by the
fungus and put out new leaves and shoots. Of the 26 twigs inoculated
with mycelium on wounds, 21 showed the effects of the fungus, most of
them dying completely from point of inoculation outward ; those
unwounded showed no effects whatever but grew in a normal manner.
The inoculations were made partly on small saplings and partly on the
small branches of larger trees. The largest sapling which died com-

Jan. io, 1914

Twig Blight of Quercus Prinus

343

pletely was about 8 feet high and the main trunk about il/i inches in
diameter. After two weeks the ends of the twigs withered and the leaves
dried up. The twigs showed the darkening of the cambium for a dis¬
tance of 6 inches from the tip. Sections across the twig also showed
pustules of the fungus just beneath the bark. After three months the
pycnidia had broken completely through the bark, spores of both types
being present in the pycnidium. On June 1, 1912, 20 other inocula¬
tions were made in the field by the wounding of the bark and inserting
a portion of mycelium. Checks were treated in like manner. Of these
only 7 were effective, as the twigs were by that time older and possibly
more resistant. In no case were there any large limbs killed, only the
small branches and tips.

CULTURE WORK

The fungus grows well in culture, but does not fruit readily, and then
only on solid media. Fresh twigs of Quercus alba and 0. prinus were
brought in from the field and
sterilized by wiping with mer-
curic-chlorid solution and rins¬
ing with distilled water. The
bark was then pricked in several
places, and portions of agar con¬
taining mycelium were spread
over these portions. These
were then put in test tubes with
sufficient moisture. In one
week discolored areas appeared
on the twigs, and in three
weeks the small black pustules
of the fungus appeared. On
examination these proved to
be the Macrophoma stage.

Twigs of the same species were also used, sterilizing them by the use
of the autoclave. The growth on these was almost entirely superficial,
the mycelium completely covering the twigs in a grayish green, felty
mass. Occasional humps or tufts of mycelium were present in which a
few pycnidia containing spores of the Macrophoma type were found. After
six months no further development had taken place. As a medium
the autoclaved twigs proved to be much inferior to the unheated twigs.

Of the agars corn meal and prune gave the best vegetative growth
and were used to the exclusion of others in securing pure cultures and in
germination studies. Portions of the mycelium were transferred to
corn-meal flasks or other solid media to secure the formation of pycnidia.

A number of different kinds of media were used: Potato, prune, beef,
and corn-meal agars, —15; potato and beef agars, +15; corn-meal and

Fig. 3. — Diplodia longxspora: Sclerotial bodies formed in
artificial media.

344

Journal of Agricultural Research

Vol. I, No. 4

prune agars, + 11, Puller's scale; Raulin’s fluid, malt, and string-bean
agars; and cylinders of Irish potato, sweet potato, parsnip, and carrot,
banana, orange, prune, and apple.

Fig. 4 .—Diplodia longispora ; A section showing grouping of pycnidia.

The Irish potato and the sweet potato gave the best results for the
vegetables. The fruits gave an abundance of mycelial growth, but few

pycnidia. In several media, espe¬
cially apple, peculiar sclerotial bodies
(fig. 3) were formed in abundance.
An extremely acid or extremely alka¬
line medium was not as satisfactory
as a nearly neutral one, and starchy
media in general gave the best re¬
sults. On all artificial media which
produced pycnidia, a dense stroma
was produced and the spores were
borne in locules in the stroma. This
is not the case on the host, where,
while the pycnidia are usually
grouped (fig. 4), a typical stroma is
never present. On all media the
colonies are at first hyaline, later
becoming grayish green, and finally
almost black.

GERMINATION STUDIES

The spores germinate readily in
distilled water, corn-meal infusion,
Raulin's fluid, and corn-meal, prune,
or potato agar. If a diseased twig
is placed in a damp chamber many
spores will germinate inside the
pycnidium. When placed in a liquid medium without being subjected
previously to a moist atmosphere, the time varies from three to six hours.

Fig. 5. — Diplodia longispora: Types of germination.
A, B, Germ tubes from end of spore; C, germ
tube from side of spore.

Jan. io, 1914

Twig Blight of Quercus Prinus

345

Usually the germ tubes are sent out from the long axis of the spores
(fig. 5, A and B) and occasionally from the sides (fig. 5, C). As many as
six tubes have been observed from a single spore.

At first the tubes *
are nonseptate, but
the cross walls grad¬
ually begin to ap¬
pear in from two to
five days from time
of germination. The
hyphae show a
marked tendency to
coalesce (fig. 6) , and
often unite to form meshes. Soon after the formation of septse the
mycelium begins to darken, taking on a grayish green hue. The hyphae
become constricted, and peculiar chlamydosporelike bodies are formed

(fig. 7) intercalary in the hyphae. When
a number of spores are sown at one
time, some of them undergo a further
development, instead of germinating as
above described. The spore turns a dark
olive brown in color, and a central, transverse septum is formed. Occa¬
sionally two septae are present (fig. 2, C), but this is not typical.

Fig. 6. — Diplodia longispora: A portion of mycelium showing the coalescing
of the hyphae.

Fig. 7. — Diplodia longispora: A portion of my¬
celium with chlamydosporelike bodies.

DETERMINATION OF THE FUNGUS

In order to determine definitely whether the Macrophoma and Diplodia
types of spores were really stages in the life history of the same fungus,
a number of single spores of each were planted in agar plates, and carefully
marked colonies of each from single spores were then transferred to corn-
meal flasks. Each first produced the Macrophoma stage and later the
Diplodia stage. Numbers of diseased twigs were brought in from the
field and carefully examined the following winter after being attacked, in
the hope of finding a perfect stage, but without success. According to
Saccardo, this fungus should be called a Botryodiplodia, as the pycnidia
are usually grouped. However, since the characters which separate it
from the genus Diplodia may be produced artificially on culture media
and vary with the amount of moisture present, it seems advisable to
place it in the latter genus.

A number of species of Diplodia have been described on Quercus,
mostly from European countries. All of them are described either from
the immature stage, or insufficient morphological characters are given for
a positive identification, the spore measurements in several being absent.
Only one species has been found described from America — Diplodia
longispora C. and Ell. on Quercus coccinea from New Jersey. It is the

346

Journal of Agricultural Research

Vd. I, No. 4

only species which is described with mature spores and in which the
spore measurements are given. The morphological characters given
agree very well, but, according to the measurements given, the spores
are uniformly longer and narrower, being 30 to 35 by 7/z in comparison
with 23 to 32 by 8 to 1 2/j. of the species under discussion.

However, since there is much variation in this genus and since the
perfect form of this fungus may eventually be found, the species herein
described is referred to Diplodia longispora C. and Ell. While, as men¬
tioned above, the spore measurements do not exactly agree, the variation
being considered by some sufficient to warrant a new species, it was not
thought desirable to add another species to the already cumbersome
and much confused nomenclature of this genus. None of the species
described are recorded as causing any disease of the host.

SUMMARY

A fungus which is referred to Diplodia longispora C. and Ell. is the
cause of a destructive twig disease of Quercus primes, also of several other
species of Quercus and of Castanea dentata.

Large trees are not killed outright, but they may eventually die as a
result of the weakened condition caused by losing the young branches,
and particularly the cumulative effect of the attacks of several years.
Saplings are often killed outright.

Infection takes place through wounds in the bark and will not take
place through an unbroken surface. The fungus does not extend into
the leaves, as no mycelium is present in the leaf tissues.

DESCRIPTION OF PLATE

Pirate XXXVIII. An oak ( Quercus gambelii) inoculated with Diplodia longispora
at X when dormant. No leaves developed above the point
of inoculation.

XXXVIII

NEW POTATO WEEVILS FROM ANDEAN SOUTH AMERICA

By W. Dwight Pierce,

Agent and Expert , Investigations of Insects Affecting Southern Field Crops , Bureau of

Entomology

During the year 1913 a number of shipments of South American pota¬
toes for experimental propagation by the Department of Agriculture
have been intercepted by Messrs. E. R. Sasscer and H. L. Sanford,
inspectors of the Federal Horticultural Board, because of more or less
serious infestations by weevils. In most of the shipments the weevils
were alive. Those received early in the summer were partly immature,
while in later shipments they were all mature. When the material was
shipped it was supposedly free of insect pests, and in fact it is quite
possible to find a potato apparently whole which contains a weevil
within. Mr. C. H. T. Townsend, the Entomologist of Peru, writes that
the work of the weevils is often undetected until the potatoes are cooked
and served on the table. It can therefore be seen how readily a shipment
of South American potatoes received for planting purposes might be
passed by quarantine officers and perhaps be the source of a very danger¬
ous pest to the American potato industry.

As a result of the finding of weevils in many shipments of potatoes,
the Federal Horticultural Board has taken action excluding South
American potatoes from the United States. This article has therefore
been prepared with the view of assisting the inspectors in their work and
also to place on record descriptions of the weevils in question.

The three species of weevils so far found are very different in appearance
and can be readily identified from the illustrations published herewith.

A notice of the finding of a species of weevil known as Rhigopsidius
tucumanus Heller in potatoes shipped by Mr. W. F. Wight from points
in Peru, Bolivia, and Chile has been published.1 Since the publication
of this note two other species, each representing a new genus and a new
species, have been discovered.

The second species found in shipments of potatoes from Peru was
obtained alive on July 9, 1913, by Mr. Sasscer in a potato sent by Mr.
Wight from the mountain districts of Peru. The adult weevil was
found just under the skin of the potato in a small cell which had evi¬
dently served as a feeding cell for the larva. From the material received
it is judged that the larva does not bore extensively in the potato.

1 Sasscer, E. R., and Pierce, W. Dwight. Preliminary report of the finding of a new weevil enemy of the
potato tuber. Proc. Ent. Soc. Wash., v. 15, no. 3, p. 143-144, pi. 4-5, Oct. 2, 1913.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(347)

Vol. I, No. 4
Jan, 10, 1914

348

Journal of Agricultural Research

Vol. I, No. 4

This weevil (PI. XU, figs, i and 2 ; and text figs. 1 and 2) forms the type
of a new genus in the family Brachyrhinidae, subfamily Entiminae, tribe
Ophryastini, to which our North American genera Ophryastes, Eupago-
deres, Amydrogmus, and Tosastes belong. In Eacordaire's group
“ Eeptopsides vrais” it is to be placed near Bastactes and Catasarcus, from
both of which it differs by many characters. The
descriptions which follow will serve to identify it.

PREMNOTRYPES, new genus.

Name derived from npkpvov (root) and rpunaxo (to bore),
meaning a root borer. Type of genus. — P. solani, n. sp.

Upper surfaces roughly sculptured throughout and closely
squamose. Beak longer than head, enlarged at alae, more
or less distinctly depressed on the median line and at the
sides; scrobes broadened behind and then flexed downward
far from eyes; mandibles beneath not acutely toothed.
Eyes vertical, elongate oval, pointed beneath. Antennae
with scape clavate, not greatly overlapping the anterior
edge of the eyes; funicle 7-jointed, with first two joints
elongate, the others shorter but not transverse; club elongate
oval. Prothorax very tuberculate above and at sides;
anterior lobes without vibrissae, almost completely cover¬
ing the eyes; base truncate, apex convex. Elytra with
humeri rounded; striation irregular, with alternate inter-
Body wingless. Thorax beneath with all parts short; meso-
thoracic side pieces unequal; metepimera broad. Intercoxal process broad; first two
abdominal segments occupying over half the abdomen; first suture arcuate; second
segment at least as long as the two following; fifth segment as long as the two preced¬
ing. Femora and tibiae stout; tibiae mucronate; tarsi with third joint bilobed and a
little wider than the preceding joints, pubescent beneath ; claws simple. The posterior
tibiae have the point of attachment of the tarsi terminal and
close to the mucro. The apical surface is divided by a ridge
into two unequal disks, the inner being the larger. The
ridge passes just outside of the corbel.

Fig. i .—Premnotrypes solani
Pierce: Lateral view of
prothorax and beak.

vals multi tuberculate.

Color brown, with

Premnotrypes solani, n. sp.

Length, 7 mm.; breadth, 3.75 mm.
bronzy scales.

Beak longer than head and narrower than eyes, being nar¬
rowest at about the middle, where the flare of the scrobes
begins to widen it. Alae strongly flared, making apical por¬
tion of scrobes open above. Head with small tubercles
above the eyes. Median line sharply defined, deepened at
frontal fovea, then bifurcate to form a median ridge. The fine
median line begins again on this ridge and extends to the apex.

Beginning even with the front edges of the eyes the lateral impressions extend half
the length of the beak. Apex of beak shining black, raised in an arcuate band, which
causes the shining semielliptical nasal plate to stand obliquely. Mandibles shining
black, with at least two inner teeth and with a long, shining, acute, deciduous piece
with sharp inner edges. The right-hand deciduous piece has a tiny tooth on the
inner edge before the middle. Antennal scrobes strongly flexed downward; scape
clavate; funicle with all joints longer than wide, gradually decreasing in size toward

Fig. 2. — Premnotrypes so-
lani Pierce: Frontal view
of beak.

Jan. io, 1914

Potato Weevils from South America

349

apex; club elongate, with the first two joints occupying over half the bulk. Head,
beak, and scape densely clad with fine, silky, bronzed scales; funicle sparsely pubes¬
cent; club minutely pubescent.

Prothorax basally truncate, apically sinuate, strongly lobed over eyes, lobes with¬
out vibrissae; coarsely punctured, finely squamose with yellowish to golden metallic
scales; median line punctate, strongly impressed; surface with six basal, two discal,
and four apical tubercles; widest behind middle at points of lateral basal tubercles.

Elytra at base no wider than thorax; humeri rounded; sides rounded, rough, wider
than prothorax. Scutellum minute, triangular, depressed. Surface densely minutely
scaly; striae irregular, with small definite punctures; entire surface rough, but the
third, fifth, and seventh intervals especially are raised by a series of small tubercles,
which give the striae a wavy direction.

Prostemum strongly arcuately emarginate, not more than one-half as long as pro-
notum. Anterior coxae contiguous. Mesostemum taken up almost entirely by the
coxae, which are narrowly separated; side pieces unequal. Metastemum also short.
Undersides and legs densely squamose.

Type. — Cat. No. 16689, U. S. National Museum.

The third species also belongs to a new genus quite closely related to
Premnotrypes and belonging in the same tribe. Several specimens in a
more or less perfect condition were found by Mr. Sanford in cells in
potatoes received October 9, 1913, from Cuzco, Peru. This species breeds
in a manner closely resembling that of the Premnotrypes solani.

This species (PI. XIT, fig. 3; text fig. 3) may be identified from the
following technical descriptions.

Trypopremnon, new genus.

Name derived from rpunaa) (to bore) and npkpvov (root), signifying a root-borer. The
name is simply " Premnotrypes J ’ reversed, because the two genera belong side by side.
Type of genus. — T. latithorax , new species.

Upper surfaces roughly sculptured throughout and closely
squamose . Beak longer than head , enlarged at alae , not impressed
on median line except at frontal fovea and near apex; scrobes
broadened behind and abruptly truncate; mandibles beneath
sharply toothed. Eyes vertical, elongate oval, pointed beneath.

Antennae with scape clavate, not greatly overlapping the anterior
edge of the eyes; funicle seven- jointed, joints 1 and 2 elongate,
the others progressively shorter and the last three transverse,
moniliform ; club elongate oval . Prothorax very roughly molded ;
median line deeply impressed ; anterior lobes without vibrissae,
almost completely covering the eyes; base truncate ; apex sinuate.

Elytra with humeri rounded; striation irregular, with alternate
intervals rough and raised. Body wingless. Thorax beneath
with all parts short; mesothoracic side pieces unequal; mete-
pimera elongate, moderately broad, Intercoxal process broad;
first two abdominal segments occupying over half the abdomen ;
the first suture arcuate; the second segment as long as the
two following; fifth segment as long as the second. Femora and tibiae stout; tibiae
mucronate; tarsi pubescent beneath, with third joint strongly bilobed, the lobes
much wider than the preceding joints; claws simple. The posterior tibiae have the
point of attachment of the tarsi terminal and close to the mucro. The apical surface
is divided by a ridge into two almost equal slanting disks, like a roof. The ridge runs
directly to the middle of the corbel.

Fig. 3. — Trypopremnon
latithorax Pierce: Lat¬
eral view of thorax
and beak.

350

Journal of Agricultural Research

Vol. I, No. 4

Trypopremnon latithorax, n. sp.

Length, 6 mm.; greatest breadth, 2.75 mm. Beak longer than head and narrower
than eyes except at alae; the dorsal sqnamose portion being gradually narrowed from
the eyes to the apex. Alae strongly flared, making the apical portion of the scrobes
open above. Head very slightly tumid above the eyes. Median line distinct only
to the frontal fovea, which is deeply depressed and very faintly indicated beyond
this point. The lateral depressions on the beak are quite faint. Apex of beak shin¬
ing reddish, with the nasal plate polished, ogival, and raised at apex. Mandibles
shining, reddish; deciduous piece long, shining, acute, arcuate, with sharp edges
and with a strong, acute, erect ventral tooth. Antennal scrobes strongly flexed down¬
ward, very much broadened and evanescent behind; scape clavate; funicle with
first two joints elongate, the others progressively shorter and the last three transverse,
moniliform; club elongate oval. Head, beak, and scape densely clad with fine,
silky, bronzed scales; funicle sparsely pubescent; club minutely pubescent.

Prothorax basally truncate, apically sinuate, with very strong supraocular lobes,
which are without vibrissae; coarsely irregularly punctured, finely squamose with
golden metallic scales; median line strongly impressed; surface very uneven with two
basal and two discal elevations and with the sides very irregular, sinuate or bitumid;
widest at posterior lateral tumidities.

Elytra at base narrower than thorax ; humeri rounded ; sides feebly convex. Scutel-
lum triangular. Surface densely, minutely scaly; striae irregular, with strong punc¬
tures, entire surface rough, but the third, fifth, and seventh intervals especially are
raised by a series of tubercles, which give the striae a wavy direction.

Prostemum strongly arcuately emarginate, hardly half as long as the pronotum.
Anterior coxae contiguous. Mesostemum taken up almost entirely by the coxae,
which are narrowly separated; side pieces unequal. Metasternum also short. Under¬
sides and legs densely squamose.

Type. — Cat. No. 16690, U. S. National Museum.

Differs from Premnotrypes solani in the sculpturing of the beak, the shape of the
scrobes and mandibles, and of the nasal plate, the absence of distinct tubercles on
the head, the shape and sculpture of the prothorax, and the elytral striation. The
third tarsal lobes are also much more distinct.

The weevil Rhigopsidius tucumanus Heller (PI. XL) is, according to
present information, more widely distributed than either of the other
species. It was originally described by Heller 1 from Tucuman, Argen¬
tina, and was recorded in the note by Sasscer and Pierce,2 quoted above,

1 Heller, K. M. Neue Riisselkafer aus Central- und Sudamerika. Bnt. Ztg. Stettin, 1906. Bd. 67
(Heft 1), p. 7-9, pi. i.t figs. 3. 3^, and 3b.

a This weevil (PI. XT) belongs to the family Psaliduridae, subfamily Rhytirhininae, tribe Rhytirhinini.
The nearest North American insects are the species of the genus Thecesternus in the tribe Thecesternini of
the same subfamily.

The following description, taken from Sasscer and Pierce (op. cit.), will identify this species.

Length, 9 mm., yellowish or purplish brown, with thickly matted vestiture of a cinereous shade mottled
with black dots. Head concealed from above by prothorax and eyes, almost covered by the lateral pro-
thoracic lobes. Beak moderately short, usually reposing in a deep pocket of the prothorax, which is pos¬
teriorly limited by the anterior coxae. Beak medianly and laterally carinate to a cross carina between the
bases of the antennal scapes. Scrobes deep and narrow from apex near tip of beak almost to eyes, then
sharply deflected and broader in front of eyes. Scape stout, clavate. Funicle 7-jointed, the last joint
apparently a part of the club. Club 4-jointed. Head at base sinuately impressed, with swellings above
the eyes. Prothorax very irregularly sculptured but with a deep median furrow widened angularly at
middle and also behind, Strial punctation deep but irregular. Intervals tumid behind. Tegs stout.
Tarsi with third j'oint not widely bilobed; tarsal claws simple. First and second abdominal segments long;
third and fourth shorter than fifth.

Jan. io, 1914

Potato Weevils from South America

35i

in shipments received May 24, 1913, from Mr. Wight, who collected the
material at Cuzco, Temuco, and Arequipa, Peru; Oruro, Bolivia, and
Ancud or San Carlos and Castro Islands, Chile. In many instances the
injury occasioned by these weevils was quite noticeable. A few of the
tubers which superficially appeared to be sound were found, on being
opened, to be infested with one and sometimes two larvae or adults. Mr.
Sasscer succeeded in keeping two adults alive from May 24 to September
6, during which period they fed but little and then only on foliage of
potato. The injury of this species consists of tunnels throughout the
potato, as shown in Plate XXXIX, and the work of the two other
weevils is very similar.

DESCRIPTION OF PLATES

Plate XXXIX. Injury caused by potato weevils. Fig. i. — A section of a potato
from Peru, showing the larva of Rhigopsidius tucumcmus in its
burrow.

Fig. 2. — A section of a potato, showing the burro wings of Rhigop¬
sidius tucumanus. The work of the two other weevils is some¬
what similar.

XL. Rhigopsidius tucumanus Heller. Fig. i. — Dorsal view.

Fig. 2. — Ventral view. Both views are much enlarged; natural
size, 9 mm.

XLI. Figs, i and 2. — Premnotrypes solani Pierce (much enlarged; natural
size, 7 mm.).

Fig. 1. — Dorsal view. In this drawing the beak, scape, and tibiae
are foreshortened, which gives an idea of even greater differences
from the succeeding species than really exist.

Fig. 2. — Ventral view.

Fig. 3. — Trypopremnon latithorax Pierce (much enlarged; natural
size, 6 mm.). Dorsal view. In this drawing the scape and the
tibiae are not foreshortened as much as in the other species. The
different attitude of the beak gives a sense of greater divergence
than occurs, as can be seen from the side view of the head and
prothorax (see text figs. 1 and 3). The ventral view resembles
very closely that of Premnotrypes solani.

The drawings which accompany this article were made by Mr. Harry B. Bradford.

(352)

AN UNDESCRIBED SPECIES OF GYMN OSPORANGIUM

FROM JAPAN

By W. H. Long,

Forest Pathologist , Investigations in Forest Pathology , Bureau of Plant Industry

INTRODUCTION

In the Annual Report of the Connecticut Agricultural Experiment Sta¬
tion for 1912 (pt. 5, p. 350), Dr. Clinton reports the introduction of
Gymnosporangium japonicum Syd. on Juniperus chinensis. The rust was
found on both stems and leaves of a form known as J. compacta , while on
a seedling of /. chinensis called J. virginalis the rust occurred only on
the leaves. The plants showing rust only on the leaves were planted in
an isolated place. The following spring they were found to be free from
rust.

Through the kindness of Dr. Perley Spaulding the writer was able to
examine some of the infected material from Dr. Clinton's herbarium
containing both types of the rust. The rust on the woody stems seems
to be Gymnosporangium japonicum Syd., but that on the leaves or young
twigs differs in most of its microscopic and macroscopic characters from
G. japonicum . According to the report, the rust on the leaves or young
twigs is apparently an annual, while the other, G. japonicum , is a perennial ;
one is found on the leaves and green twigs, the other on the woody stems ;
one causes no deformation of the host, the other produces fusiform
enlargements 4 cm. in length or longer. The microscopic characters of
the two differ as widely as the gross characters mentioned above.

The writer has found in most species of Gymnosporangium three types
of teliospores in the same sorus. One type has very thick colored walls;
one, moderately thick colored walls; and the third, thin and colorless
walls. These three types usually differ from each other also in shape
and size of the spore as a whole or in the individual cells of each spore.
Constant specific characters may occur in one type, often in the thin
colorless-walled spores, while they are absent in the other two types or
are not so pronounced. For this reason the characters of at least the
two extreme types of spores should be given for each species under dis¬
cussion. In the following descriptions the two extreme types are fully
described for two of the species and the three types for the third one. As
a matter of convenience in comparing the three species brief descriptions
of G. japonicum and G. haraeanum are also given.

(353) Vol.I, No. 4

Jan. 10, 1914
G — 11

Journal of Agricultural Research,

Bept. of Agriculture, Washington, D, C.

354

Journal of Agricultural Research

Vol. I, No. 4

DESCRIPTION OF SPECIES OF GYMNOSPORANGIUM

Gymnosporangium chinensis, n. sp.

.Ecia unknown.

Telia epiphyllous or caulicolous, appearing on the very small green twigs between
the leaves, not causing a fasciation of the young shoots; scattered; usually hemis¬
pheric; about i mm. in diameter; hazel in color when desiccated.

Teliospores 2-celled; spores with colored walls, oval to broadly ellipsoid, 19 to 22
by 35 to 40 jtz (average for 10 spores, 21 by 36.7 jtz), slightly but plainly constricted at
septum. The two cells are usually subequal; spores rounded at both ends, walls
thin, about 1 to 1.5 ji, pedicel cylindric; pores, one to two in each cell near septum,
or rarely only one in upper cell and apical.

Spores with thin colorless walls, ellipsoid, 17 to 19 by 47 to 52 fi (average for 10 spores,
18 by 49 /z), plainly constricted at septum. The two cells are unequal, the lower being
from 3 to 7 n longer than the corresponding upper cell; apical cell rounded or only
slightly narrowed toward apex, lower narrowed toward base; wall thin, colorless,
about 1 fi thick; pores, one to two in each cell near septum, or usually only one in
upper cell and apical.

Host plant. — On Juniperus chinensis in the Elm City Nursery, Westville, Conn.,
March 28, 1911, on stock just imported from Japan. In same packet with Gymno¬
sporangium japonicum on the same host. From the herbarium of Dr. G. P. Clinton.

Gymnosporangium japonicum Syd.

Telia caulicolous on fusiform enlargements, 4 cm. or more long, of the woody stems,,
irregular tongue or wedge shaped, about 3 mm. or more long, often in rows.

Teliospores 2-celled, occasionally 3-celled; spores with thick colored walls, ellipsoid,
22 to 24 by 48 to 63 /z (average size for 10 spores 22 by 54.7 /z), cells subequal or lower
longer and more narrowed at base, not constricted at septum, narrowed at both ends;
walls 1.5 to 2 g thick, pores near septum, two in each cell.

Spores with thin colorless walls, elliptic fusiform to linear oblong, 16 to 19 by 57 to
70/z (average for 10 spores 16.8 by 65 /z), walls 1 jtz thick, not constricted at septum;,
pores, two in each cell near septum.

Host plant. — On Juniperus chinensis from Japan.

Gymnosporangium haraeanum Syd.

(Sydow, H., and Sydow, P. Novae fungorum species — VIII. Ann. Mycol., v. 10,
no. 4, p. 405, 1912.)

Telia epiphyllous or caulicolous on the very small green twigs, not causing a fascia¬
tion of the young shoots; scattered; hemispheric to short conic; one-half to 1 mm.
in size.

Teliospores 2-celled; spores with very thick colored walls, ellipsoid, 25 to 28 by
35 to 44 fi (average size for 10 spores 25.7 by 39 fi), not or but very slightly constricted
at the septum; spores rounded or somewhat narrowed at both ends; the two cells
subequal or the lower often larger $nd more narrowed toward the base than the upper
one ; walls very thick, 3 to 4 ft; pores, two in each cell near septum ; pedicel cylindrical.

Spores with walls moderately thick and colored, elliptic oblong, 22 to 26 by 48 to 57 /z
(average for 10 spores 23.6 by 52 fi), not or but slightly constricted at septum; spores
usually much narrowed at both ends; upper cell often with a mammillate apex;
lower cell often longer than upper; walls 2.5 to 3 /z thick; pores, two in each cell near
septum.

Spores with walls thin and colorless, oblong to oblong fusiform, 16 to 19 by 48 to 57 fi
(average size for 10 spores 17 by 51 jtz), rarely constricted at septum; cells subequal,
rounded or narrowed at both ends; pores, two in each cell near septum; walls about
1 fi thick.

* Host plant. — On Juniperus chinensis from Japan.1

1 This description was made irom a portion of the type material which Dr. Sydow kindly sent to the
writer. *

Jan. io, 1914

An Undescribed Species of Gymnosporangium

355

The three types of spores are described in full, and their diagnostic
character can readily be seen when Gymnosporangium haraeanum is com¬
pared with the other two species given in this paper. In Gymnosporan¬
gium japonicum and G. chinensis the spores with thick and moderately
thick colored walls for each species are so similar that the two kinds are
described as one; therefore, only two types of spores, thick and thin
walled, are described for each of these two species. G. chinensis and G.
haraeanum are so closely related that the writer would not publish the
former as a new species until he had examined the type material of the
latter. After a careful examination, however, the conclusion was
reached that the two were distinct, as they differ in certain fundamental
microscopic characters. These differences are shown in the description
given of each species. The most marked difference between these two
species is the position of the germ pores in the colorless thin-walled
teliospores. In G. chinesis they are plainly apical in the upper cell, while
in G. haraeanum they are just as certainly situated only at the septum
in both cells.

According to Dr. Clinton, the telia of Gymnosporangium chinensis occur
on the leaves, but in the very meager herbarium material examined by
the writer they arose between the leaves rather than on them. The telia
are therefore stated in the above description to be either caulicolous or
epiphyllous.

The three types of spores mentioned in the above descriptions are
usually more evident in herbarium material than in fresh, as the obstruct¬
ing colored contents of the spores fade in drying, thus permitting a
clearer view of the spore walls.

The value of taking into consideration at least two types of spores, the
thick and thin walled ones, is very evident when the corresponding kinds
for each species are compared. For instance, the oval thick-walled
spores of Gymnosporangium chinensis , with equal cells rounded at both
ends, are in marked contrast to the ellipsoid, thick- walled spores of G.
japonicum , with unequal cells sharply contracted at both ends; while the
long, narrow, linear-oblong, thin-walled spores, with equal cells of G.
japonicum , are very different from the shorter thin- walled spores, with
unequal cells of G. chinensis. Again, many of the thick-walled spores of
G. japonicum are so sharply attenuated at both ends that they become
trapezoid in shape, while the apical cells often have a distinctly mammil-
lated apex. Neither of these characters is present in the thick-walled
spores of G. chinensis .

Through the kindness of Dr. Shirai the writer has been able to examine
some of the material of Gymnosporangium japonicum collected in 1900.
It was probably a part of the material used by him in his inoculation
experiments with this species.1 The specimens sent consist of two

1 Shirai, M. Uber den genetischen Zusammenhang zwischen Rostelia koreaensis P. Henn. und Gym-
nosporangium japonicum Sydow. Ztschr. Pflanzenkrank., Bd. io, Heft i, p. 1-5, pi. 1-2, 1900.

170730— r4 - 6

356 Journal of Agricultural Research voi. 1, no. 4

infected branches. One lesion is on a woody stem 6 mm. in diameter;
the other is on a much younger branch 1.5 mm. in diameter. No telia
were found on the leaves or very young twigs. The telia and telio-
spores were similar to those found on the woody stems of the imported
Juniperus chinensis from Connecticut, but had nothing in common with
the telia of G. chinensis, which were found on the very young twigs and
leaves of this imported juniper.

ADDITIONAL COPIES of this publication
may be procured from the Superintend¬
ent op Documents, Government Printing
Office, Washington, D. C., at 25 cents per copy

Subscription, per year, 12 numbers - - $2.50

JOURNAL OF MCdTlAL RESEARCH

DEPARTMENT OF AGRICULTURE

Vol. I Washington, D. C., February 16, 1914 No. 5

THE PRESENCE OF SOME BENZENE DERIVATIVES

IN SOILS

By Edmund C. Shorey,

Scientist in Soil-Fertility Investigations , Bureau of Soils
INTRODUCTION

The isolation of organic compounds from soils may have an interest
other than that of the purely scientific nature attached to any increase
in our knowledge of the composition of soils. This is true, not only
when the compounds are known to be readily reactive with other com¬
pounds or to have an effect on the microflora of the soil or the growth
of higher plants, but also when their constitution indicates that they
may have such an effect. Recently three organic compounds have been
isolated from soils that seem to be of this nature.

These compounds, rather closely related in constitution, are benzoic
acid, metaoxytoluic acid, and vanillin. They were obtained from
samples of sandy soil from Florida at present devoted to orange culture.
These soils are composed of quartz sand ranging in color from light gray
to brown and contain very little organic matter. For the most part
this organic matter is deposited in a thin layer on the grains of sand, so
that when the soils are treated with dilute alkali and the film of organic
material is thereby dissolved or loosened pure white quartz sand remains.
The samples, about 90 kilograms in each case, were from eight loca¬
tions, the top soil and subsoil being represented by separate samples.

BENZOIC ACID

Benzoic acid was obtained from but one of these samples— a subsoil.
There was no indication of its presence in the corresponding surface soil,
and, although indications were obtained of its presence in other subsoils
of this series, it could not be isolated in a pure form in sufficient quantity
for identification.

The method by which benzoic acid was obtained from this soil was as
follows :

The soil was treated at room temperature with a 2 per cent solution of
sodium hydroxid for six hours, allowed to stand several hours, and the

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(357)

Vol. I, No. s
Feb. 16, 1914
H — 2

358

Journal of Agricultural Research

Vol. I, No. 5

colored extract siphoned off. This extract was acidified with sulphuric
acid, filtered, and the acid filtrate shaken out several times with ether.
The ether extracts were combined and the ether evaporated on the sur¬
face of a small quantity of warm water. The water was then heated to
boiling and filtered hot, when, on cooling the filtrate, crystals separated.
A further yield of crystals was obtained on concentrating the mother
liquor from the crystals first obtained. The compound obtained in this
way was purified by recrystallizing from water, and finally by sublima¬
tion, when a pure white product was obtained. About 2 grams were
obtained from 25 kilograms of soil.

This compound had all the properties of benzoic acid. It crystallized
in the leaflets characteristic of benzoic acid. It was readily soluble in
alcohol, ether, and chloroform, sparingly soluble in cold water but much
more readily in hot water, and melted at 12 1° C. An aqueous solution
was acid in reaction, and when neutralized and treated with a neutral
solution of ferric chlorid a dirty-brown precipitate was formed that was
insoluble in acetic acid. The compound sublimed readily and when
heated strongly gave off the irritating fumes characteristic of benzoic acid
when so treated. Finally it gave Mohler’s reaction.1

The appearance and properties of the compound obtained from the
soil, its behavior with ferric chlorid, and response to Mohler’s test are
sufficient to establish its identity as benzoic acid.

METAOXYTODUIC ACID

Metaoxytoluic acid was obtained from several samples of the series
examined, but in quantity only from subsoils. The method by which it
was obtained was exactly the same as that just outlined for benzoic add
up to the point of obtaining a water solution of the ether extract. If no
benzoic add or other compound separated on cooling the filtrate, it was
concentrated nearly to dryness and allowed to stand, when oxytoluic add,
if present, crystallized out. The compound so obtained was purified by
repeated recrystallizations from water and was finally dried on a porous
plate. This product retained persistently a slight tinge of color, and it
was only after many treatments and much loss of material that it could
be freed from color. Where traces of benzoic add accompanied it, as
seemed to be the case in some instances, it could be freed from benzoic
add by sublimation, this acid being much more readily sublimed than
oxytoluic add. About 10 grams of pure material were obtained from 25
kilograms of soil.

1 Mohler’s test is carried out by heating the substance to be determined with sulphuric add until charring
takes place, sulphobenzoic acid being formed if benzoic acid is present in the original material. On -treat¬
ing with potassium nitrate this "will be transformed into metadinitro-benzoic acid. On adding excess of
ammonia to this add and then a few drops of a colorless solution of ammonium sulphid a red color will
be obtained.

Offidal and provisional methods of analysis, Association of Official Agricultural Chemists. U. S. Dept.
Agr. Bur. Chem. Bui. 107 (rev.), p. 181, 1908.

Mohler, E. Recherche de l'adde benzoique dans les substances alimentaires. Bui. Soc. Chim., Paris,
s. 3, t. 3, p. 414-416, 1890.

Feb. 16, 1914

Benzene Derivatives in Soils

359

On elementary analysis this compound gave the following results (0.200
gram were used for each analysis) :

Analysis i. Analysis 2.

Per cent. Per cent.

Carbon.... . 63.33 63.02

Hydrogen . 5.39 5.66

Oxygen . 31.28 31.32

This corresponds with the composition of oxytoluic acid, C8H803,
which contains 63.15 per cent of carbon, 5.26 per cent of hydrogen, and
31.50 per cent of oxygen. There are 10 isomeric oxytoluic acids, all of
which have been described. The compound obtained from the soil has
the properties of metaoxytoluic acid, with the carboxyl, hydroxyl, and
methyl radicals in the 1. 3. 5. positions, respectively.

This compound crystallizes in plates or, when the quantity is small, in
groups of radiating needles. It is rather soluble in cold water, but more
so in hot water. It is soluble in alcohol and ether, sublimes unchanged,
and melts at 208° C. On the addition of a solution of ferric chlorid, an
aqueous solution of the compound gives a brown precipitate which dis¬
solves to a brown solution when the reagent is added in excess. On the
distillation of the dry compound with lime, metacresol is formed. The
identity of the metacresol obtained from the soil compound in this way
was established by transforming it into 2. 4. 6. trinitrocresol, a yellow com¬
pound melting at 106° to 107° C. This was effected by dissolving the
metacresol in strong sulphuric acid, pouring into a mixture of nitric and
sulphuric acids, heating, and then cooling. The nitro product was fil¬
tered off, washed with dilute hydrochloric acid, recrystallized, and dried.

Metaoxytoluic acid was made from sulphotoluic acid according to the
method of Jacobsen,1 and its properties were compared with the com¬
pound obtained from the soil, the two agreeing in every respect. When
the artificial product and the soil compound were mixed, the melting
point was unchanged. The agreement in composition and properties
mentioned is sufficient to establish the identity of the compound obtained
from the soil as metaoxytoluic acid.

VANILLIN

In the course of investigations of the organic matter of soils carried on
for the past six years soil extracts having the odor of vanillin and giving
some of the reactions of that compound have been encountered from time
to time, but its presence could not be confirmed by isolation in pure
form. In investigating the soil samples from Florida the isolation was
accomplished.

The method of isolation, as with the compounds just described, was
begun by making an alkaline extract of the soil. This extract was acidi¬
fied and filtered and was then shaken out with several portions of ether.

1 Jacobsen, Oscar. Oxytoluylsauren und Oxyphtalsauren. Ber. Deut. Chem. Gesell., Jahrg. 14, Juli-
Dez., p. 2357-2359, 1881.

360

Journal of Agricultural Research

Vol. I, No. s

The combined ether extracts were shaken with a strong solution of
sodium bisulphite, which treatment removes from the ether compounds
of an aldehyde nature. After separating the bisulphite from the ether
it was acidified with enough sulphuric acid to decompose all the bisul¬
phite, was freed from sulphur dioxid by blowing air through it, and was
again shaken with ether. On evaporating the ether extract at room
temperature a more or less oily, viscous residue remained which had the
odor of vanillin and gave the reactions of that compound. Crystals usu¬
ally separated from this residue after standing several days.

When these crystals were obtained in sufficient quantity, they were
purified by the method recommended for the examination of vanilla
extracts.1 The oily residue was treated several times with warm water
and filtered, and the filtrate treated with a solution of lead acetate as
long as a precipitate formed, and was then again filtered. The filtrate
was shaken out several times with ether, and the combined ether extracts
were evaporated. The residue usually crystallized readily, although it
still contained traces of resinous matter. This resinous material could
not be removed by taking up in ammonia, acidifying, and again shaking
with ether as recommended in the method for vanilla extracts, but by
recrystallizing from water several times and finally drying on a porous
plate pure crystals were obtained.

These crystals had a strong odor of vanilla, were in the form of needles
or small prisms, and melted at 8o° to 8i° C., the melting point of vanillin.
They were readily soluble in ether or alcohol, but were sparingly soluble
in water. An aqueous solution gave the following reactions character¬
istic of vanillin or the group of compounds to which vanillin belongs:

The addition of a solution of ferric chlorid gave a blue-violet color.
Colors ranging from blue to violet are given by many hydroxy-benzene
compounds.

When boiled with resorcinol and hydrochloric acid, a red color was
formed. This reaction is given by a number of aldehydes, including
some sugars.

When the crystals were treated with equal quantities of sulphuric and
hydrochloric acids and with the addition of a drop of a dilute solution of
acetone and the mixture then heated to ioo° C. for 15 minutes, a violet
color was developed.

On adding an excess of bromin water followed by the addition of ferrous
sulphate, a blue-green color wras formed. This reaction is regarded as
characteristic of vanillin, but does not seem applicable for colorimetric
determination.

When the reagent of Folin and Denis 2 wras added and the mixture made
alkaline with an excess of sodium carbonate, a clear blue color was

1 Official and provisional methods of analysis, Association of Official Agricultural Chemists. U. S. Dept.
Agr. Bur. Chem. Bui. 107 (rev.), p. 156, 1008.

2 Folin, Otto, and Denis, Wr. On phosphotungstic-phosphomolybdic compounds as color reagents.
Jour. Biol. Chem., v. 12, no. 2, pp. 239-243, 1912.

Feb. 16, 1914

Benzene Derivatives in Soils

361

slowly developed. This reaches a maximum in a short time and remains
constant for several hours, furnishing a reliable colorimetric method for
the determination of vanillin.1

The method of isolating the compound from the soil, its crystalline
form, odor, melting point, and the fact that it gave the characteristic
reactions of vanillin are sufficient to establish its identity as vanillin.

The quantity of pure vanillin obtained by this method from any of the
soils examined was but a few milligrams from 25 kilograms. It was
possible to obtain the vanillin in pure crystalline form from 4 of the 16
samples. From some of the other samples crystals were obtained that
gave the reactions of vanillin, but there were not enough for the separa¬
tion and determination of the melting point. Residues were obtained
from all the samples having the odor of vanillin and giving two or more
reactions for it. In each case where it was possible to separate vanillin
in pure form the sample was a surface soil.

An application of the colorimetric method of Folin and Denis was made
to two samples, those from which the most vanillin had been obtained by
alkaline extraction. One hundred grams of soil were finely ground and
thoroughly extracted with warm alcohol that had been freshly distilled.
The alcohol was evaporated, and the residue was taken up with warm
water and filtered. Lead acetate was added to the filtrate as long as a
precipitate formed. The solution was then filtered and the filtrate
treated with the reagent of Tolin and Denis (a mixture of phosphotungstic
and phosphomolybdic acids), followed by an excess of sodium carbonate.
It was again filtered and made to a definite volume. The resulting blue
solution was read in a colorimeter against a solution prepared in the same
way from a standard solution of vanillin. Sample No. 1 gave 0.0010 per
cent of vanillin, or 10 parts per million, while sample No. 2 showed
0.00048 per cent, or 4.8 parts per million.

Vanillin contains the radical methoxyl OCH3. In a previous paper2
it was shown that the methoxyl radical is present in many soils in suffi¬
cient quantity to be determined by the Zeisel method.3 A determination
of the methoxyl in samples Nos. 1 and 2 by this method gave, for sample
No. 1, 0.065 per cent of methoxyl calculated to vanillin, and for sample
No. 2, 0.050 per cent.

Methoxyl is contained in a number of organic compounds and is a con¬
stant constituent of the lignocellulose of plants. The quantity obtained
from these soils when calculated to vanillin is so much in excess of that
actually obtained in the isolation from an alkaline extract, or that

1 Folin, Otto, and Denis, W. A new colorimetric method for the determination of vanillin in flavoring
extracts. Jour. Indus, and Engin. Chem., v. 4, no. 9, pp. 670-672, 1912.

2 Shorey, E. C., and Eathrop, E. C. Methoxyl in soil organic matter. Jour. Amer. Chem. Soc., v. 33,
no. 1, p. 7S“78» 1911.

8 Zeisel, S. fiber ein Verfahren zum quantitativen Nachweise von Methoxyl. Monatsh. Chem.,
Bd. 6, 1S85, p. 989-996, 1 pi. 1886.

1 ■-1—' Zum quantitativen Nachweise von Methoxyl. Monatsh. Chem., Bd. 7, 1886, p. 406, 409. 1887.

362

Journal of Agricultural Research

Vol. I, No. 5

indicated by the Folin-Denis method, that it seems evident that a con¬
siderable portion of it must be derived from compounds other than
vanillin.

These three compounds, benzoic acid, metaoxytoluic acid, and vanillin,
although not related in the sense that they are readily derived from or
transformed into one another, are related as shown by the following
structural formulas :

Benzoic acid.

C;— COOH
H — C C—H

H— C C— H

Nc/

Benzoic acid is a naturally occurring product obtained from certain
gums and balsams, wherein it exists as an ester. It is also present in
some fruits, such as plums and cranberries, and has been found among
the oxidation products of casein and gelatin. Its presence in soil might
then result from the decay of plant tissues containing it or from oxida¬
tion of more complex compounds through the activity of microorgan¬
isms. The most remarkable fact in connection with its occurrence in
the soils examined is that it was found in appreciable quantity in but
one sample, although they were of the same general character. In the
absence of accurate information regarding previous natural vegetation
on these soils and other data that can be obtained only in the field, any
attempt to explain this fact is out of place here.

Metaoxytoluic acid, so far as known, is not a natural product, and its
method of preparation in the laboratory does not suggest any process by
which it might be formed in the soil from plant products or other com¬
pounds known to occur in soils.

Vanillin has its chief natural source in the so-called vanilla beans, or
seed pods, of Vanilla pompona. It has also been reported as found in
small quantities in a number of other plants or plant products, and it
probably is more widely distributed in the vegetable kingdom than has
been supposed. At present there is no information indicating its forma¬
tion from other compounds through the agency of microorganisms, and
the small quantity found in soils may possibly be regarded as an un¬
changed residue of plant debris.

Using the maximum figures for quantities obtained in these investi¬
gations and calculating to the acre-foot of soil, the following approxi¬
mate quantities are obtained: Benzoic acid, 350 pounds; metaoxytoluic
acid, 800 pounds; and vanillin, 40 pounds to the acre-foot.

Metaoxytoluic acid.

COOH

H— C C — H

1 1

CH— C C— OH

\c/

Vanillin.

0— COH
H — C C— H

h4 C— OCH

Feb. 16, 1914

Benzene Derivatives in Soils

363

In the case of the two acids the method involved considerable loss of
material and the actual quantity present in the soil is undoubtedly in
excess of these figures.

The question as to the form in which these compounds exist in the
soil is one deserving some consideration, although one not easily an¬
swered satisfactorily. It is true of most organic compounds that have
been obtained from soils through extraction with dilute alkali that they
are not readily obtained as such by water extraction of the soil. In
many soils this can be explained, in part at least, by the fact that much
of the organic matter in soils is of a resinous nature wholly insoluble in
water, and compounds which when separated are easily soluble in water
are so incased or protected by the resinous or vamishlike coating effected
by this resinous material that they are very slowly dissolved, if at all,
when the soil is leached. This effect is quite apart from any absorptive
effect and is quite marked in extreme types, such as the sands of Florida
and some peats, where either fine grinding or previous treatment with
alcohol will render soluble in water organic material that before this
treatment was so little soluble as to escape notice.

In the case of vanillin, grinding the soil and extracting with alcohol
gave more of the compound than was obtained by extraction with alkali,
and from the known properties of vanillin it seems unlikely that the
quantity found is in the soil in any other form than free vanillin.

Treatment of the soil with hot alcohol after grinding gave extracts
from which reactions for both benzoic acid and metaoxytoluic acid could
be obtained, but in the absence of colorimetric methods applicable to
small quantities and owing to the fact that the extractions with alcohol
were made with much smaller quantities of soil than the extraction with
sodium hydroxid, no comparative figures can be given. It is fair to
conclude, however, that in some of the soils examined some portion of
both acids is present as free acid.

INDICATOR SIGNIFICANCE OF VEGETATION IN
TOOELE VALLEY, UTAH

By T. H. Kearney,1 L. J. Briggs,2 H. L. Shantz,3 J. W. McUane,4 and

R. L. Piemeisel,5
Bureau of Plant Industry

INTRODUCTION

In the arid portion of the United States the different types of native
vegetation are often very sharply delimited, the transitions being so
abrupt that they can not be attributed to climatic factors ; this has sug¬
gested the possibility of correlating the distribution of the vegetation
with the physical and chemical properties of the soil. If such correla¬
tions can be made, they may be utilized in the classification of land
with respect to its agricultural capabilities.

One of the writers 6 has described the correlations which exist in the
Great Plains between the different types of vegetation and the physical
characteristics of the corresponding types of land and has pointed out
how the native growth may be used in that region to determine the
suitability of the land for dry farming.

The results obtained in the Great Plains made it desirable to undertake
similar investigations in the Great Basin region, or that portion of the
United States lying between the Rocky Mountains on the east and the
Sierra Nevada and Cascade Ranges on the west. The problems to be
solved were: First, what types of vegetation indicate conditions of soil
moisture favorable or unfavorable to dry farming, and, second, what
types indicate the presence or absence of alkali salts in quantities likely
to injure cultivated crops. For the purpose of this investigation it
was necessary to find a locality where both dry farming and irrigation
farming are practiced, where much of the land is still covered with the
original native growth, and where some of the soils contain an excess
of alkali salts.

1 Physiologist in Charge, Alkali and Drought Resistant Plant Investigations.

1 Biophysicist in Charge, Biophysical Investigations.

1 Plant Physiologist, Alkali and Drought Resistant Plant Investigations.

4 laboratory Assistant, Biophysical Investigations.

6 Scientific Assistant, Alkali and Drought Resistant Plant Investigations.

8 Shantz, H. I*. Natural vegetation as an indicator of the capabilities of land for crop production in the
Great Plains area. U. S. Dept. Agr.t Bur. Plant Indus. Bui. 201, 100 p., 33 fig., 6 pi. 1911.

With the exception of the valuable work of Hilgard in Mississippi and of Hilgard and his associates
in California (see Hilgard, E. W., Soils, New York, 1906, p. 487-548, figs. 77-89), very little had previously
been done in the United States toward a scientific study of native vegetation from the indicator point of
view. In Europe, however, the subject has been much investigated, especially as regards “lime-
loving" and “lime-avoiding" plants.

Vol. I, No. 5
Feb. 16, 1914
G — 12

Journal of Agricultural Research
Dept, of Agriculture, Washington, D. C.

(365)

366

Journal of Agricultural Research

Vol. I, No. 5

After a reconnoissance trip through portions of Wyoming, Utah,
Idaho, and Oregon in August, 1911, the Tooele Valley in central Utah
was selected for the following reasons: (1) several very distinct types of
vegetation are found within a small area, (2) the soils show a great
diversity in their moisture conditions and salt content, (3) the greater
part of the area retains its original plant cover, while examples of crop
production both with and without irrigation exist on different types
of land.

Detailed studies of the vegetation of Tooele Valley in relation to the
moisture conditions and salt content of the soil were carried on in 1912.
The work was begun near the close of the rainy season (end of May)
and was terminated during the first week of August, when the summer
drought had reached its height. Additional data were obtained during
a third visit to the valley in the latter part of August, 1913.

The distribution of the native vegetation was found to depend in a
marked degree upon the physical and chemical properties of the soils,
factors which also influence crop production. So far as this particular
area is concerned, the vegetation can unquestionably be used with
advantage in classifying land with respect to its agricultural value.
To what extent the correlations established in Tooele Valley hold good
in other parts of the Great Basin region remains to be determined by
future investigation.

The writers desire to acknowledge the helpful cooperation of Director
E. D, Ball, of the Utah Agricultural Experiment Station, and of Prof.
L. A. Merrill, formerly of that station. The writers are indebted for
the determination of the plants collected to Mr. Ivar Tidestrom, of the
Office of Economic and Systematic Botany, Bureau of Plant Industry.

METHODS OF RESEARCH

The methods used in classifying and describing the types of vegetation
are well known to ecological plant geographers and are best described in
setting forth the results. Some explanation of the methods used in
investigating the moisture conditions and salinity of the soils, however,
is desirable.

Samples of the soil were taken in the midst of the areas occupied by
each vegetation type. Where the boundaries between two types were
well defined, samples were also taken on both sides of the line, in order
to determine the limiting conditions for each type. The measurements
of moisture content, moisture equivalent, electrical resistance, and salt
content which were made upon these samples served as a basis for con¬
clusions regarding the physical conditions indicated by the presence of
each important type of vegetation.

Coelecting Soil Samples. — The samples of soil were in all cases
collected with the aid of a sampling tube, which prevents the admix¬
ture of surface material with the subsoil. Each sample consisted of a

Feb. 16, 1914

Indicator Significance of Vegetation

367

composite of four cores. The soils were usually sampled to a depth of 4
feet and occasionally to a greater depth, the cores being taken in i-foot
sections.

Determining the Soil-Moisture Content. — Numbered tin boxes of
uniform weight and with tight-fitting covers were used to receive the
soil samples directly from the sampling tubes. The whole sample was
used as a basis for the moisture determinations and after the initial
weighing was dried in a water oven to constant weight. The moisture
content is in all cases expressed as a percentage of the dry weight of the
sample.

Determining the Moisture Equivalent. — In studying soil moisture
in relation to plant growth it is important to have some standard for
measurement of the retentivity of the soil for moisture. As two of the
authors have previously shown,1 this may be conveniently accomplished
by the method of moisture equivalents. This method consists in sub¬
jecting a moist sample of soil to a constant centrifugal force equal to
1,000 times that of gravity until the moisture content of the soil is
reduced to the point where it is in equilibrium with the centrifugal
force employed. The residual moisture content of the soil is then
determined. This value, expressed as a percentage of the dry weight
of the sample, is the moisture equivalent. A direct measure of the
retentiveness for moisture of the various soils is thus obtained, and,
since the same force is employed throughout, all of the determinations
are directly comparable.

Determining the Wilting Coefficient. — It has been shown by two
of the authors 2 that the moisture equivalent serves as a useful indirect
means of determining the wilting coefficient. The latter term designates
(as a percentage of the dry weight of the soil) the quantity of water
remaining in the volume of soil occupied by the active roots of a plant
which is beginning to wilt.3

These determinations (moisture equivalent and wilting coefficient)
serve to give an idea of the texture of the soils occupied by the different
plant associations, as indicated by their retentiveness for moisture. By
subtracting the wilting coefficient from the actual moisture content a
measure is obtained of the percentage of moisture available for the active
growth of plants at the time the soil samples were taken.

Determining the Salt Content. — The total salt content of each
soil sample was determined by the electrical- resistance method developed

1 Briggs, L- J., and McLane, J. W. Moisture equivalents of soils. U. S. Dept. Agr., Bur. Soils Bui. 45,
23 p., 1 fig., 1 pi. 1907.

Briggs, L- J., and McLane, J. W. Moisture-equivalent determinations and their application. Proc.
Amer. Soc. Agron., v. 2, 1910, p. 138-147, pi. 6. 1912.

3 Briggs, L. J., and Shantz, H. L. The wilting coefficient for different plants and its indirect determi¬
nation. U. S. Dept. Agr., Bur. Plpnt Indus. Bui. 230, 83 p., 9 fig., 2 pi., 1912.

8 “ Wilting” in this case must be understood as permanent wilting — i. e., a condition from which the
plant can not recover its turgidity until the soil receives additional moisture, no matter how great the
humidity of the atmosphere.

368

Journal of Agricultural Research

Vol. I, No. 5

in the Bureau of Soils.1 The method is simple and rapid and the meas¬
urements can be readily made in the field, which is a great advantage in
studying the distribution of vegetation in relation to the salt content
of the soil. The method is, however, necessarily an approximate one,
owing to the variation in the composition of the soil solution and to the
fact that the salts found in soils differ greatly with respect to their molec¬
ular weight and ionic migration velocity. To interpret the observed
resistance, a calibration curve was prepared, based upon the observed
relationship between the electrical resistance and the salt content, gravi-
metrically determined, of a number of soils from different parts of the
valley. (See fig. i.)

In making the gravimetric determinations, the usual practice was followed of digest¬
ing ioo grams of dry soil with 500 c. c. of water, filtering, and evaporating an aliquot

Fig. i. — Curve showing the relation between the salt content (in percentages of the dry weight of the
soil) and the specific electrical resistance (in ohms) of the soil when saturated with water.

portion of the filtrate to dryness. A number of the samples examined were rich in
gypsum, and in digesting such soils with an excess of water the total quantity of gyp¬
sum which goes into solution is greatly in excess of the quantity dissolved when the
soil is simply saturated with water. The gravimetric determination of the salt content
of soils which are rich in gypsum is consequently too high, and this accounts in part at
least for the outlying points above the calibration curve. (Fig. 1.)

By means of a suitable centrifugal apparatus it is possible to remove and collect a
portion of the soil solution in an unsaturated soil. From the concentration of this
solution and the initial moisture content of the soil, the salt content of the soil can be
calculated. This method gave results more nearly in accord with those indicated by

1 Whitney, Milton, and Means, T. H. An electrical method of determining the soluble salt content of
soils. U. S. Dept. Agr., Div. Soils Bui. 8, 30 p., 6 fig. 1897.

Briggs, E. J. Electrical instruments for determining the moisture, temperature, and soluble salt content
of soils. U. S. Dept. Agr., Div. Soils Bui. 15, 35 P-» 12 fig. 1899.

Davis, R. O. E., and Bryan, H. The electrical bridge for the determination of soluble salts in soils.
U. S. Dept. Agr., Bur. Soils Bui. 6i, 36 p., 7 fig. 2 pi. 1910.

Feb. 16, 1914

Indicator Significance of Vegetation

369

the electrical resistance. Therefore, in the case of soils containing gypsum the elec¬
trical-resistance method may be considered to be more reliable than the excess-solvent
method. The probable error of determinations by the electrical-resistance method
is approximately 10 per cent of the actual salt content.

CLIMATE OF TOOELE VALLEY

Tooele Valley is dry, having a mean annual precipitation of 16 inches.1
The average monthly distribution of the precipitation at Tooele is shown
in figure 2. No precipitation records are available for other parts of the
valley, save fragmentary records at Grantsville for two years, which
indicate that the western side of the valley receives decidedly less prer
cipitation than the eastern slope. During the first nine months of 1912,
the total precipitation recorded at Grantsville was 7.6 inches, as com¬
pared with 13 inches
at Tooele. The con-
dition of the native
vegetation and of the
crops grown without
irrigation also indi¬
cates that the western
side of the valley is
much drier than the
eastern side.

In view of the im¬
portance of the soil-
moisture conditions in
explaining the distri¬
bution of the different
types of vegetation in
Tooele Valley, it is in¬
teresting to consider
the precipitation of the period immediately prior to that during which
the field work was carried on. The precipitation during the months
from October to May, inclusive, probably furnishes all of the stored
soil moisture available for the growth of plants during the following
summer. The total precipitation at Tooele during the period from
October, 191 1, to May, 1912, was 13.5 inches, or 0.9 inch above the normal
(12.6 inches) for the locality. Hence, it may be assumed that at least the
normal quantity of moisture was present in the soil on the date when
field operations were begun in the valley (May 28). As regards the
season of active growth in 1912, the precipitation of the month of May
was about 0.5 inch below the normal for Tooele, while that of June was
very nearly twice the normal. For the remaining summer months the
precipitation was about normal.

-Monthly distribution of precipitation at Tooele, Utah (mean
for is years).

1 Based upon 15 years’ measurements at the town of Tooele. From data furnished by the U. S. Weather
Bureau, through the courtesy of Mr. A. H. Thiessen, Section Director.

370

Journal of Agricultural Research

Vol. I, No. s

While no evaporation data are available for Tooele Valley, evaporation
measurements1 have been made during the last five years at Nephi,
about 60 miles south of Tooele. These measurements show that the
monthly evaporation during June, July, and August is at least double
that of April and October. (See Table I.)

Table I. — Evaporation from a free-water surface at Nephi , Utah , during the months of

April to October , 1908 to 1912.

Year,

April.

May.

June.

July.

August.

Sep¬

tember.

October.

TQO& . ...

Inches.

7. 87
8.81
IO. 90

8. 69

9. 28

Inches.
10. 52

9- 47
9.98

8. 72

9. 24

Inches.

9 ■ 34
7- 03
10. 09
10.47

8. 89

Inches.

6. 23

5* 59

6. 01

6. 69

6. 16

Inches.

I9°9 .

1910 .

1911 .

1912 .

Normal .

3- 64
5- 82
4.93

3- 54

5-99
7.46
8. 41

6. 30

4- 43
3- 72
3- 65
2. 98

4. 48

7.04

9. II i

9- 59

9. 16

6. 14

3- 70

Therefore, while the summer months are by no means rainless in this
locality, the great increase in the rate of evaporation is such that the
light precipitation can have but little effect upon vegetation. In those
parts of the valley where the ground water is beyond the reach of the
plant roots the vegetation becomes dormant after the moisture stored in
the soil by the winter and spring rains has been exhausted. Herbaceous
plants ripen and die, at least to the ground, while the woody species,
losing much of their foliage and reducing their transpiration to a mini¬
mum, enter a resting condition which is nearly as complete as that which
is brought about by the low temperatures of winter. Where there is a
greater depth of readily permeable soil in which moisture can be stored
than is ordinarily the case in this valley, the beginning of summer dor¬
mancy is longer postponed. On the sand hills the larger shrubs may con¬
tinue growing more or less actively throughout the summer. In the
lower part of the valley, where the ground-water table is high and the
soil is moist throughout the summer nearly or quite to the surface,
active growth continues until it is terminated by frosts.

GEOLOGY AND TOPOGRAPHY OF TOOELE VALLEY

Geologically, Tooele Valley is of exceptional interest because of its
occupancy at one time by a bay of Lake Bonneville, a Pleistocene lake,
the beach lines of which are strikingly in evidence upon the sides of the
surrounding mountains. The highest of these terraces is 1 ,000 feet
above the present surface of Great Salt Lake. An exhaustive study
of the region has been made by Gilbert.2 *

1 Measurements by the Office of Biophysical Investigations in cooperation with the Office of Cereal
Investigations, Bureau of Plant Industry, and with the Utah Agricultural Experiment Station.

2 Gilbert, G. K. Lake Bonneville. 438 p., 51 illus., 51 pis. Washington. 1890. (U. S. Geol. Survey

Monograph 1.)

Feb. 16, 1914

Indicator Significance of Vegetation

37i

Tooele Valley is broadly U-shaped in cross section, the mountains on
either side rising somewhat abruptly from the valley floor. This abrupt
change from valley plain to mountain is characteristic of many of the
valleys of the region and is due to the extensive deposition of alluvium
during some epoch prior to the Bonneville period.

Tooele Valley is bounded on the east by the Oquirrh Mountains and
on the west by the Stansbury or Aqui Range. The southern boundary
is formed by a spur of the Stansbury Range and by the great Stockton
embankment, which is composed of sand and water-worn gravel thrown
up by the waters of Lake Bonneville and which separates Tooele Valley
from Rush Valley. (The Stockton embankment is shown in extreme
background of PI. XLV, fig. 3.) The summit of this embankment
coincides with the highest shore line on the adjacent mountains. To the
north the valley slopes downward to the southern shore of Great Salt
Lake. The axis of the valley thus lies approximately on a north and
south line, the land rising gradually from near the center to the mountain
ranges on the east and west sides. The width of Tooele Valley at the
northern end is about 18 miles, at the southern end it is about 13 miles,
and its greatest length is approximately 16 miles. The total area of the
valley floor is, roughly, 250 square miles.

The slope of the valley from the sides and from the southern end to a
line marked approximately by the highway from Salt Lake City to
Grantsville is decidedly steep, as is indicated by the fact that the town
of Tooele has an elevation above sea level of 4,900 feet, while Grants¬
ville, although less than 5 miles farther north, is 680 feet lower.1 North
of this line the slope becomes very gentle and the surface of this portion
of the valley is plainlike.

SALINE CONDITIONS OF TOOELE VALLEY

The soils of Tooele Valley show a wide range in salinity, or, to use the
more familiar term, in “alkali” content. The soils in the upper end of
the valley and along the base of the foothills at either side, including a
large alluvial fan northeast of the town of Tooele, are characterized by a
low salt content. The other extreme is found in the flats adjacent to
the lake, which in some cases contain such an excess of soluble salts as
to prevent the development of a plant cover. The soils occupying the
central portion of the valley are, as a rule, relatively free from salts in
the surface foot, but the salinity of the subsoil is usually such as to
exclude all deep-rooted plants except those that are salt-tolerant to a
marked degree. The saline material in solution in the nearly saturated
soils of the flats, like that of the lake itself, is made up largely of sodium
chlorid. In fact, these flats have probably not infrequently been sub¬
merged by the rise of the lake, since records made by the United States

1 The elevation of the surface of the water of Great Salt Take is about 4,200 feet.

372

Journal of Agricultural Research

Vol. I, No. s

Geological Survey show that within the last 40 years the lake has under¬
gone a fluctuation in level of 16 feet.

The following determinations of the composition of the saline material
in Great Salt Lake, which are quoted from a compilation by Clarke,1
are therefore of interest in showing what may be regarded as the typical
composition of the saline material in this part of the area.

Table II. — Analyses of water from Great Salt Lake.1

Cl .

Br .

S5-99

Trace.

6- 57

56. 21

55- 57

56- 54

55- 69
Trace.

6. 52

JS-2 5
Trace.
6- 73

55- 11

53- 72

so4 .

C03 .

6. 89
.07

6.86

5-97

6. 66

5- 95

hi .

Trace.

33-15

1. 60

•17

2. 52

. 01
32. 92

1. 70
1.05

2. 10

. 01

Trace.
34- 65
2. 64
. 16
■ 57

Na .

K .

Ca .

Mg .

(Fe203,AJ203,
SiOo). .. .

33-45

. 20
3. 18

33-17
59
. 21

2. 60

33- 39

1. 08
.42

2. 60

32.97
3- 13
• i7

1. 96

32. 81
4. 99
•3* *
2. 22

Salinity, per
cent .

100. 00

14. 994

IOO. OO

13. 79O

100. 00

15. 671

100. 00

i9- 558

100. 00

*23. 036

100. 00

27. 72

100. 00

22. 99

100. 00

17. 68

1 More correctly 230.355 grams per liter.

“A. By O. D. Allen, Rept. U. S. Geol. Expl. 40th Par., vol. 2, 1877, p. 433. Water collected in 1869. A
trace of boric acid is also reported, in addition to the substances named in the table. Allen also gives analy¬
ses of a saline soil from a mud flat near Great Salt Lake. It contained 16.40 per cent of soluble matter much
like that of the lake water.

“B. By Charles Smart. Cited in Resources and attractions of the Territory of Utah, Omaha, 1879. Anal¬
ysis made in 1877.

“C. By E. von Cochenhausen, for C. Ochsenius, Zeitschr. Deutsch. geol. Gesell., vol. 34, 1882, p. 359. Sam¬
ple collected by Ochsenius April 16, 1879. Ochsenius also gives an analysis of the salt manufactured from
the water of Great Salt Lake.

“D. By J. E. Talmage, Science, vol. 14, 1889, p. 445. Collected in 1889. An analysis of a sample taken
in 1885 is also given.

“E. By E. Waller. School of Mines Quart., vol. 14, 1892, p. 57. A trace of boric add is also reported.

“F. By W. Blum. Collected in 1904. Recalculated to 100 percent. Reported by Talmage in Scottish
Geog. Mag., vol. 20, 1904, p. 424. An earlier paper by Talmage on the lake is in the same journal, vol. 17.
1901, p. 617.

"G. By W. C. Ebaugh and K. Williams, Chem. Zeitung, vol. 32, 1908, p. 409. Collected in October, 1907.

“H. By W. Macfarlane, Science, vol. 32, 1910, p. 568. Collected in February, 1910. A number of other
analyses, complete or incomplete, are cited in this paper by Ebaugh and Macfarlane.”

It will appear from these analyses that sodium and chlorin together
constitute about 90 per cent of the total soluble material. The quan¬
tity of chlorin is, in each analysis, slightly greater than that necessary
to satisfy the basic requirements of sodium. The rest of the soluble
material is made up almost wholly of potassium, magnesium, and the
sulphate radical. Concerning these analyses Clarke 2 says :

Although the salinity of the lake is very variable and from four to seven times as
great as that of the ocean, its saline matter has nearly the same composition. The

* Clarke, F. W. Data of geochemistry. U. S. Geol. Survey Bui. 491, ed. 2, p. 144. 1911.

* Clarke, F. W. Op. cit.

Feb. 16, 1914

Indicator Significance of Vegetation

373

absence of carbonates, the higher sodium, and the lower magnesium are the most
definite variations from the oceanic standard; but the general similarity, the identity
of type, is unmistakable. * * *

All the waters tributary to Great Salt Lake, so far as they have been examined,
contain notable quantities of carbonates, which are absent from the lake itself. These
salts have evidently been precipitated from solution, and evidence of this process
is found in beds of oolitic sand, composed mainly of calcium carbonate, which exist
at various points along the lake shore. The strong brine of the lake seems to be inca¬
pable of holding calcium carbonate in solution.

The analyses as given in Table II report the presence of carbonates
in solution in the lake water in only one instance.1 It is in this respect
that the saline matter of the soils more distant from the lake differs most
markedly from the type just considered. Calcium carbonate was
found widely distributed in the soils of the valley. Sodium carbonate
was often found also, usually in small amounts (0,05 to 0.10 per cent
of the dry weight of the soil), but occasionally samples were collected
containing as high as 0.25 per cent. Sodium carbonate was found most
frequently in the samples collected in areas where Kochia was growing.
These soils were also highly calcareous. The available data on the dis¬
tribution of sodium carbonate do not, however, indicate that it can be
correlated with the presence of any particular plant community.

The composition of the salts of Great Salt Lake would lead one to
expect that the chlorids would prove to be the most common and widely
distributed of the saline constituents of the Tooele Valley soils, and such
has been found to be the case. In the course of the work a quantita¬
tive examination for chlorids was made of 162 samples of soil, and all but
13 samples showed the presence of measurable quantities of chlorids.
Of these 13 exceptions 12 were samples from Artemisia tridentata (sage¬
brush) areas which are characterized by a very low total salt content.
The sodium-chlorid content of the areas examined, all of which were
occupied by vegetation of one type or another, ranged from a trace in
the land occupied by Artemisia to over 2 per cent in land occupied
by Allenrolfea occidental is. Outside of the sagebrush areas the sodium-
chlorid content of most of the samples fell between 0.4 and 1.3 per cent.
In a large majority of the samples examined sodium chlorid constituted
more than one-half of the total water-soluble material.

Sulphates are usually present in the soils containing an excess of salts.
Of 122 samples examined 96 showed the presence of sulphates. It is
well recognized through the researches of Hilgard and others that cal¬
cium sulphate is a corrective for the soluble “black alkali” (sodium car¬
bonate) , the reaction between these salts resulting in the formation of the

1 F. K. Cameron has shown, however, that while the lake water at its normal concentration does not
give an alkaline reaction with phenolphthalein, this reaction will develop simply by diluting the lake
water with distilled water. At the normal concentration of the lake, the dissociation of the sodium car¬
bonate is held back through the great number of sodium ions resulting from the dissociation of the sodium
chlorid. The lake does, therefore, carry a slight amount of sodium carbonate. (Gardner, F. D., and
Stewart, John. A soil survey in Salt Take Valley, Utah. U. S. Dept. Agr., Div. Soils Field Operations,
Rpt. 64, 1899, p. 104-105. 1900.)

24395 14 - 2

374

Journal of Agricultural Research

Vol. I, No. 5

relatively insoluble calcium carbonate and neutral sodium sulphate. It
is evident that a similar reaction would take place if magnesium sulphate
were present, since magnesium also forms an insoluble carbonate. It
consequently seemed desirable to examine the carbonate and sulphate
measurements with a view to determining to what extent the absence of
soluble carbonates was accompanied by the presence of sulphates. Of
122 samples examined for carbonates and sulphates 13 contained neither
carbonates nor sulphates, while 13 others contained carbonates but no
sulphates, leaving 96 samples containing sulphates. Of these, 78 sam¬
ples were free from carbonates, 2 samples contained both carbonates and
sulphates in measurable quantities, while in the remaining 16 samples
traces only of both sulphates and carbonates were present.

VEGETATION OF TOOELE VALLEY

The plant covering of the area under consideration is typical of a large
portion of the Great Basin, several of the most important types of vege¬
tation of that region being represented in Tooele Valley. Striking features
of this vegetation are (1) the great extent of the areas occupied continuously
by a single type of vegetation, (2) the sharpness of the boundaries between
the areas occupied by each type, and (3) the great predominance of one or
very few species in each type.1 2

CLASSIFICATION OF THE TYPES OF VEGETATION 3

The principal types of vegetation of Tooele Valley, with the names of
the species which are dominant in each, are listed in Table III.

1 These are common characteristics of the vegetation of arid regions. Thus, Borszczow, as quoted by
Ove Paulsen (Studies on the Vegetation of the Transcaspian Lowlands. Copenhagen, 1912, p. 22-23), states:

"Here, as throughout the whole of Aralo-Caspia, it is a few specially characteristic forms which prevail;
they repeat themselves continually so that the country has a very monotonous appearance. Other species
are only subordinate to these. Where the character of the soil changes, these predominant species some¬
times change very quickly and give place to others, which in turn prevail until the soil changes again.
This monotony and this repetition of certain species over vast areas is the third characteristic of the vegeta¬
tion of the Aralo-Caspian countries. It is no doubt a direct consequence of the uniformity of the climate,
which again is mainly dependent on the slight vertical relief of the surface. * * *

"In the Aralo-Caspian lands the soil in particular has such a great influence on the vegetation that a
change of soil — other conditions remaining the same — often alters the physiognomy totally and almost
abruptly without any gradual transitions. ”

2 In view of the fact that the ecological plant geography of the Great Basin region is as yet but little under-
stood, it seems inadvisable at this time to attempt to refer the plant associations of this valley to formations.

The term "plant association, ” as used in this paper, signifies an assemblage of plants occupying a rela¬
tively uniform environment, having an easily recognizable appearance or "physiognomy " and characterized
by the predominance of one or few species.

Feb. 16, 1914

375

Indicator Significance of Vegetation

Table III. — Types of the vegetation in Tooele Valley , Utah , and their dominant species .

Name of association or other plant community.1

Dominant species.

Sagebrush association .

Artemisia tridentata.

f Artemisia tridentata.

| Juniperus utahensis.

[Chrysothamnus nauseosus albicaulis.

Sand-hill niiv^d association. .

Kochia association

Kochia vestita.

Shadscale association .

Greasewood-shadscale association

Grass-flat communities .

Salt-flat communities

Atriplex confertifolia.

Sarcobatus vermiculatus.

Atriplex confertifolia.

(Distichlis spicata.

Sporobolus airoides.

Chrysothamnus graveolens glabrata.

{Allenrolfea occidentalis.

Salicomia utahensis.

Salicornia rubra.

1 Further investigation of the vegetation of the Great Basin region is needed before definite ecological
rank can be assigned to the grass-fiat and the salt-fiat communities.

DISTRIBUTION OF THE TYPES OF VEGETATION

The distribution and relative area in Tooele Valley of the different
types of vegetation is shown on the map (PI. XLII).

Nearly all of the dry land free from alkali salts which retains the
original plant covering is occupied by the sagebrush association. (PI.
XTIV.) This type of vegetation covers the southern end of the valley
and also extends northward in a narrow fringe along the base of the
Stansbury Range to within about 5 miles of Great Salt Lake. Few
vestiges of the original cover remain on the eastern side of the valley,
but there can be little doubt that sagebrush formerly occupied the bench
lands and alluvial fans at the foot of the Oquirrh Range. The dominant
species of this association is also found along gullies and in depressions,
in the midst of areas otherwise occupied by the Kochia and shadscale
associations. It is probable that most of the land now occupied by the
sagebrush association was laid bare before the waters of Lake Bonne¬
ville had become strongly saline.

South of the center of the valley a rather extensive area of sand hills is
covered by what may be designated the sand-hill mixed association.
In this association also sagebrush is the dominant plant, but there is a
plentiful admixture of Utah juniper and certain species of rabbit brush
(Chrysothamnus), together with many herbaceous plants more or less
peculiar to sandy soils. Botanically, this is the most varied and inter¬
esting type of vegetation occurring in Tooele Valley.

376

Journal of Agricultural Research

Vol. I, No. s

The middle portion of the valley resembles the upper portion in the
dryness of the soil and subsoil during the summer, but differs in the high
salt content of the subsoil. This territory is divided between two types
of vegetation, the Kochia (PI. XL VI) and the shadscale (PL XLVII,
fig. i) associations. The former occupies a sharply defined interrupted
belt extending well across the valley just south of the sagebrush
area and also penetrates the latter in the form of tongues and islands,
which occur here and there far toward the head of the valley. (PI.
XLHI, fig. 2.) Lying just below the main Kochia belt an extensive tract
is occupied by the shadscale association, which on the western side of
the valley is prolonged in a gradually narrowing strip to the north end
of the Stansbury Range. While the boundary between the sagebrush
and Kochia associations is often very sharp (PL XLVI, fig. i), that
between the Kochia and shadscale associations is much less distinct.
It is probable that the water of Lake Bonneville had become strongly
saline before the areas now occupied by the Kochia and shadscale asso¬
ciations were laid bare and that the subsequent precipitation has been
too small to leach the salts then deposited to a greater depth than i or 2
feet.

As the elevation of the land diminishes, the pure shadscale is gradually
replaced by an association of greasewood and shadscale. The frontier
between the two associations is not sharply defined (Pl. XLVII, fig. 2),
scattered greasewood plants appearing first along gullies or draws and
gradually, as Great Salt Lake is approached, mingling everywhere with
the shadscale. This association extends to the edge of the lake, covering
the summits of the low ridges and hummocks which are interspersed
among the salt flats. In Tooele Valley greasewood scarcely occurs in a
pure association, but is practically everywhere mingled with shadscale.

Between the main greasewood-shadscale area and the salt flats occur
the grass flats, a nearly level expanse, marshy in places, covered largely
, with grasses and with a species of Chrysothamnus. (Pl. XLVIII, fig. 3.)
1 Near the present margin of the lake basinlike areas are found, many
of which are doubtless under water at times. (Pl. XLHI, fig. 1 ; PL
XLVIII, fig. 1.) The larger of these appear in summer as bare expanses
covered with a glistening crust of white salts. Near their margins,
however, and often covering the entire surface of the smaller depressions
certain very salt-resistant plants occur, either scattered over the other¬
wise bare ground or forming rather dense colonies. The most important
of these plants are Allenrolfea occidentalis (Pl. XLVIII, fig. 1), which is
most at home on the slightly higher margins of the basins, and two
species of glasswort (Salicornia) — one perennial (S. utahensis) (Pl.
XLVIII, fig. 2), the other annual (5. rubra).

Feb. 16, 1914 Indicator Significance of Vegetation 377

To recapitulate, the dry, well-drained, nonsaline land in the upper
part of the valley is occupied chiefly by the sagebrush association; the
dry saline land near the center is covered with a vegetation of Kochia or
of shadscale; the land in the lower part of the valley, which is often dry
on the surface but has a moist subsoil, bears a mixed vegetation of
greasewood and shadscale; while the lowest areas near the lake shore,
where the soil is strongly saline to the surface and where during much of
the year even the first foot is wet, bear the salt-flat type of vegetation.
The grass flats occupy a moist, moderately saline area lying between the
two preceding. These relationships are shown in Table XVIII, p. 413,
and are graphically represented in figure 13 on p. 412.

In the following pages descriptions are given of the several associations
and other plant communities, arranged in the order shown in Table III.

SAGEBRUSH ASSOCIATION
Topographical Relations

The sagebrush association is one of the most important types of
vegetation of the Great Basin region. In Tooele Valley (see map, PI.
XLII) it occurs chiefly on the bench lands which skirt the mountains.
The best growth of sagebrush (apart from that on the sand hills as
described later) is found on the alluvial fans which are situated near the
mouths of canyons. In such places the moisture received directly as
precipitation is probably supplemented by water from the hills. This
type of vegetation extends across the southern end of the valley and
probably at one time formed a continuous belt, although fire and cultiva¬
tion have greatly diminished the area originally occupied, especially on
the east side. Farther down the valley, below the main area occupied
by this association, sagebrush is found only on sand hills, along drainage
channels, and in depressions — places where the moisture conditions are
more favorable and more of the alkali has been leached out than in the
surrounding areas.

Botanical Composition

This association in its typical form is dominated by Artemisia tridentata
(PI, XUV) as almost the sole woody plant. In less typical phases two
or three species of rabbit brush (Chrysothamnus) occur.1 Many species of
perennial herbs associate with the sagebrush, especially in those portions
of the area which lie nearest the foothills. The following list includes
all shrubs and perennial herbs which were noted as belonging to the
sagebrush association.

1 These are never abundant and never attain their maximum size where they occur in the typical sage¬
brush association in Tooele Valley. They appear more at home where associated with Artemisia on the
sand hills, and at roadsides and along ditches in areas which were formerly covered with the sagebrush
association.

378

Journal of Agricultural Research

Vol. I, No. 5

PERENNIAL species op the sagebrush association 1
Common or frequent

Agropyron spicatum (Pursh) Rydb.
Eriocoma cuspidata Nutt.

Poa sandbergii Vasey
Sitanion jubatum J. G. Smith
Zygadenus paniculatus Wats.
Eriogonum ovalifolium Nutt.
Opuntia sp.

Malvastrum coccineum (Pursh) Gray
Phlox longifolia Nutt.

Castilleja linariaefolia Benth.

Artemisia tridentata Nutt.

Chrysothamnus marianus Rydb.
Chrysothamnus nauseosus albicaulis (Nutt.)
Rydb.

Chrysothamnus pumilus Nutt.

Erigeron pumilus Nutt.

Guiierrezia saroihrae (Pursh) B. and R.
Senecio uiniahensis A. Nels.

Less frequent or rare

Stipa comata Trin. and Rupr.

Atrip lex canescens (Pursh) James
Delphinium burkei Greene
Cowania stansburiana Torr.

Astragalus arietinus Jones
Astragalus beckwitkii T. and G.
Astragalus utahensis T. and G.

Anogra pallida (Lindl.) Brit.

Gaura parviflora Dougl.

Pachylophus marginatus (Nutt.) Rydb.
Lappula caerulescens Rydb.

Lappula occidental is (Wats.) Greene

Thalesiafasciculata (Nutt.) Brit.
Antennaria dimorpha (Nutt.) T, and G.
Balsamorrhiza hirsuta Nutt.
Balsamorrhiza sagittata (Pursh) Nutt.
Chaenactis douglasii H. and A.
Chrysopsis villosa (Pursh) Nutt.

Crepis occidentals Nutt.

Layia glandulosa H. and A.

Leucelene ericoides (Torr. ) Greene
Ptilocalais nutans (Geyer) Greene
Tetradymia inermis Nutt.

Numerous annual and biennial plants occur in this association. By
far the most abundant of these are two introduced species, Bromus iectorum
and alfilaria (. Er odium cicutarium) , which in many places cover the
ground among the “sage” bushes with a dense mat of vegetation. The
more abundant or otherwise conspicuous annual and biennial plants of
the sagebrush association are given in the following list :

ANNUAL AND BIENNIAL SPECIES OF THE SAGEBRUSH ASSOCIATION

Bromus iectorum L.

Festuca octoflora hiriella Piper
Arabis longirostris Wats.

Draba sp.

Sophia filipes (Gray) Heller
Sophia pinnata (Walt.) Brit.
Erodium cicutarium L’Her.
Mentzelia dispersa (Wats.) A. Nels.

Mentzelia laevicaulis (Dougl.) T. and G.
Anogra albicaulis (Pursh) Brit.

Phacelia linearis (Pursh) Holz.
Cryptanihe sp.

Lappula cupulata (Gray) Rydb.
Lappula subdecumbens (Parry) Nels.
Amsinckia tessellata Gray

Appearance

The characteristic appearance of the sagebrush association is illus¬
trated in Plate XLIV, figure i. During the early summer, when their
maximum growth is taking place, the sagebrush plants present a silvery

1 la this and all following lists of species the families are arranged in the sequence of Engler and Prantl
(Die Natiirlichen Pflanzenfamilien), while the genera are arranged alphabetically under each family.

Feb. 16, 1914

Indicator Significance of Vegetation

379

appearance, due to the hairy covering of the young leaves. From the
middle of summer to the following spring the plants having lost many of
their leaves and the dark stems being more in evidence, the appearance
of the vegetation is decidedly different. Still another aspect is that of
the early fall when the Artemisia plants are in flower and give a yellow¬
ish color to the vegetation. The contrast between the comparatively
vivid and varied appearance of the vegetation in early summer and its
monotonous aspect during the rest of the year is heightened by the
fact that nearly all of the flowering herbs belonging to this association die,
at least to the ground,
long before the close
of the summer.

In some parts of
the valley, especially
where the soil is sandy,
the plants of sagebrush
are tall, vigorous, and
stand close together.

In other and more
extensive areas, where
the moisture condi¬
tions are less favor¬
able, they are scat¬
tered and stunted, and
the proportion of new
growth to old wood is
small.1 The plants, in
fact, look as if they
were slowly dying in
such areas. By far
the best growth of
Artemisia tridentata? is
found on the sand hills
and along irrigating
ditches. In the greater part of the area occupied by this association the
plants are from 2)4 to 4 feet high. Their frequency is indicated in figure
3, which represents a quadrat 2 platted early in the month of August in a
typical portion of this association as it occurs in Tooele Valley.

The associated herbs, although of many species, are not sufficiently
numerous individually nor sufficiently large in size to materially affect
the aspect of the vegetation, even when they are at the height of their

Fig. 3. —A representative 10-meter quadrat of the sagebrush association,
showing the location of each individual of Artemisia tridentata (A)
and of Gutierrezia sarothrae (G), these being the only woody species
present. The figures show the number of main branches of the Arte¬
misia plants and hence indicate their relative size. The absence of a
figure indicates that there was only one large stem. A circle around
the A indicates a dead plant of Artemisia. Bromus teetorum was
very abundant around the Artemisia bushes, and Sitanion was also
present. These two grasses are not shown on the quadrat.

1 These slow-growing plants reach a considerable age without attaining a large stem diameter. Twenty-
three annual rings were counted in a stem barely -2% inches in cross section.

2 For descriptions of the method of quadrats, see Clements, F. E., Research Methods in Ecology, Lincoln,
1905, p. 161-176; and also his Plant Physiology and Ecology, New York, 1907, p. 202-310.

380

Journal of Agricultural Research

Vol. I, No. s

growth and blossoming. Moreover, they are apt to be partly hidden,
owing to their habit of growing close among the stems of the sagebrush.
After midsummer most of the herbaceous species die, at least to the
ground, and during the rest of the year typical areas when viewed from
a little distance appear to contain no species other than Artemisia
tridentata .

Physical Conditions Indicated

The soils occupied by the sagebrush association, which consist largely
of products of erosion deposited upon the bed of the ancient Take
Bonneville, are rather coarse in texture and often contain much gravel.
All available data concerning the moisture conditions and salt content
of the soil in typical portions of this association as it occurs in Tooele
Valley are given in Table IV.

Table IV. — Sagebrush association: Moisture conditions and salt content of the soil in

typical areas.1

Date of collection.

Item.

Depth of
(feet.)

June.

August.

Aver¬

age.

No. of sample . . .

104

f r

13* 1

15. 9

16. 7

12. 6

15. 7

18. 4

12. 1

9 3
0- 0

14* 3

Moisture equiv¬

J a

15. 6

17. 4

17. 7

22. 9

14. 7

19. 1

8.9

ic 6

alent .

1 3

23. 8

13- 3

22. 2

24. $

15* 2

7- 5

0* 0

16* c

19. 6

11. 6

23. 4

7* w

8. 7

TP. 8

f I

7. 1

8.6

9. I

6.8

8. 5

10. 0

6.6

Wilting coeffi¬

J ^

8. 5

9. 4

9. 6

12. 4

8.0

jo. 4

4.8

0 * v

A* O

4* 4
& 5

cient .

1 3

12. 9

7. 2

12. 0

13. 3

8.2

4* 1

h* y

Am O

1 4

i c. 6

6. 3

12. 7

4* y

4*7

8,6

Moisture con¬

i I

— , 7

— i* 4

— 3. 0

— 1. 2

— 3* 7

”5* 0

-—2. *

tent above or

J *

+3. 1

H- .4

— 1. 5

-(- . 3

—2* 5

— 4* I

* j

— .6

below the wilt¬

1 3

—4. 4

+ • 7

— 4* 4

+3. 7

— 3. 2

— I* 3

ing coefficient .

( 4

— 1. 0

+4* O

+x. 0

( 1

0.03

o. 04

•03

.04

•03

■03

• 03

■03

O.03

•03

*03

•03

*03

■03

.04

. 06

•03

.03

•03

*03

•03

. 02

•03

Salt content .

< 3

. 08

• 03

. 12

• °S

. 02

* 03

* 05

1 4

. 10

*03

. 12

•OS

• 07

l s

*05

. 10

*07

1 All data in this table are stated in percentages of the dry weight of the soil. The moisture contents
with a plus sign (+) represent moisture available for growth (above the wilting coefficient), while those
with a minus sign (—) represent a corresponding deficit of available moisture (below the wilting coefficient).

Soil, Moisture). — Typical sagebrush land is characterized by a rather
light texture of the soil, as is indicated by the relatively low moisture
equivalent. In such soil water penetrates readily to considerable depths,
and the run-off must be small. Consequently, although the moisture¬
holding capacity is low, the total quantity of water available to deep-
rooted plants is considerable.

The rapid growth of the sagebrush plants in the early part of the
summer results in a speedy exhaustion of the moisture available for
growth, and in most years the water content of the soil to the depth
reached by the roots is probably reduced by midsummer to below the

Feb. 16, 1914

Indicator Significance of Vegetation

38i

wilting coefficient in much the greater part of the area occupied by this
association. That this is the case is strongly indicated by the fact that
most samples of soil collected during the month of June, 1912, showed
very little or no moisture above the wilting coefficient to a depth of 4
feet (Table IV).

In places where the conditions for the reception of water and removal
of alkali salts are more than usually favorable — e. g., along drainage
channels, in depressions, and in places where the soil has been loosened
by burrowing animals — there is probably available moisture within reach
of the roots during a longer period. Artemisia tridentata was not seen
growing under natural conditions where the water table is near the sur¬
face of the soil.

Fig. 4. — Artemisia tridentata (sagebrush): A, Detail showing the wedge-shaped, 3-toothed leaves by
which this plant is easily recognized; B, a small plant growing where hard pan occurred, showing the
deflection of the taproot from a vertical to a horizontal direction after reaching a depth of 5 inches.

Optimum soil-moisture conditions for the growth of Artemisia triden¬
tata are rarely realized in Tooele Valley. This is shown by the much
larger size and more vigorous appearance of the plants which grow on
sand hills and along irrigating ditches. In many places the maximum
depth reached by the roots is only from 1 8 to 30 inches and is marked by
the presence of a hardpan consisting of coarse gravel cemented by cal¬
careous material. The depth at which this hardpan is formed probably
represents the limit of penetration of the rain water, and consequently
most of the roots of the sagebrush do not penetrate farther. The shal¬
lowness of the moisture-holding layer of the soil greatly reduces the
absolute quantity of moisture available for growth. The effect is shown
in the thin stand and in the small size and sickly appearance of the
plants (fig. 4). The eastern part of the valley, where most of the dry-

382

Journal of Agricultural Research

Vol. I, No. 5

farming area is situated, was in all probability once covered with sage¬
brush vegetation, although few traces of it now remain. Here the condi¬
tions were probably more favorable for the growth of this plant than in
most of the area still occupied by it.

Salinity. — Reference to Table IV shows that in the typical sage¬
brush land of Tooele Valley the salt content of the soil is extremely low —
lower, in fact, than in many soils of humid regions. Near the lines of con¬
tact with other associations, however, Artemisia iridentata frequently
grows where much salt is present at a depth of 30 or 40 inches. In such
places the saline subsoil is an effectual barrier to the penetration of the
roots, the depth of soil from which the plant must extract its entire
supply of water is correspondingly limited, and as a result the plants are
scattered, very small, and give every appearance of suffering from
drought.

An excellent example of this condition was observed on the west side
of the valley, where in a spot of considerable size the plants of Artemisia
were widely spaced, rarely more than 2 feet high, and had many dead
branches. Samples of soil collected in this spot on June 3 gave salt
contents and moisture contents as follows:

Tabi,E V. — Salt content and moisture content ( above or below the wilting coefficient) at
different depths of the soil where the sagebrush was small and suffering .

Depth.

Salt content.

Moisture con- i
tent above or j
below wilting j
coefficient. !

Feet .

Per cent.

Per cent .

O. 05

— 2. O

. 18

— 2. 2

• 53

“3-4

4 '

. 64

+ .8

• 59

The roots of the plant alongside this boring penetrated to a depth of
only about 2 feet, at which point the taproot had died, and development
was continued by horizontal laterals. The feeding roots were mostly
confined to the first foot of the soil.

The most extreme condition as regards salinity which was noted at any
point in Tooele Valley where Artemisia iridentata grew was in a small
pocketlike depression among the sand hills where salts had accumulated
as a result of seepage from the surrounding dunes and where very small
sickly plants of sagebrush grew in company with greasewood (Sarcobatus)
and Kochia. The salt contents were as follows: First foot, 0.16 per cent;
second foot, 0.51 per cent; third foot, 0.67 per cent; fourth foot, o .66
per cent. The presence here of living plants of sagebrush is doubtless
explained by the fact that large seed-producing plants of this species
were growing on the surrounding dunes and that the salt content of the

Feb. 16, 1914

Indicator Significance of Vegetation

383

surface soil was not high enough to prevent the germination and seedling
growth of the Artemisia.

Summary of Physical Conditions. — The observations made in
Tooele Valley lead to the conclusion that in this area a good stand and
growth of sagebrush indicates (1) a rather coarse textured, readily
permeable soil, with low run-off and good underdrainage (water table
low); (2) a depth of soil of at least 3 feet, in which water can be
stored and into which the roots of plants may easily penetrate; (3) at
least 3 feet of soil free from alkali salts in quantity sufficient to injure
ordinary crop plants.

Adaptations to the Physical Conditions

The herbaceous species of the sagebrush association are for the most
part shallow rooted, and, hence, are dependent upon the moisture of the
upper soil. The great majority of them grow so rapidly during the
spring and early summer that they are able to complete their develop¬
ment and ripen seed before the water content of the first foot or two of
the soil has been exhausted to the wilting coefficient. When this occurs,
they die, at least to the ground.1 After the middle of July few living
plants except sagebrush are visible in typical areas of this association.

The dominant species, Artemisia tridentata (sagebrush) is able to
continue growth during a longer period. As shown in figure 5 and in
Plate XLIV, figure 2, it possesses a “generalized” type of roots2 — i. e.,
a highly developed system of laterals in the upper soil and also a deeply
penetrating taproot. The former are admirably adapted for securing
the moisture which penetrates only to a small depth during light rains
and for which in spring and early summer this plant must compete with
the numerous associated annual and perennial herbs. By means of its
taproot the plant can also avail itself of moisture stored at greater
depths3 in the readily permeable soils which are preferred by ‘this
association.

The great development of superficial lateral roots favors rapid growth
so long as abundant moisture is present in the upper soil, while the deep
penetration of the taproot permits the plant to continue growth at a
slower rate long after most of the herbaceous species of this association
have withered away. In typical areas of sagebrush vegetation as repre¬
sented in Tooele Valley (PI. XLIV, fig. 1) the available moisture is
probably exhausted before the end of the summer in all depths of soil

1 They are for the most part " drought escaping ’ ' rather than " drought enduring.” See Kearney, T. H.,
and Shantz, H. L., The water economy of dry-land crops, U. S. Dept. Agr., Yearbook, 1911, p. 354-357.
1912.

2 See Cannon, W. A., The Root Habits of Desert Plants, Washington, p. 87, 1911. (Carnegie Inst.,
Washington, Pub. 131.)

3 Plate XLIV, figure 2, reproduced from a photograph taken in the vicinity of Nephi, Utah, shows the
taproot of Artemisia tridentata extending vertically to a depth of over 15 feet. The root penetration of this
plant under optimum conditions was not studied in Tooele Valley, but it is unlikely that in most of the
area there occupied by this association the roots reach so great a depth. In this locality the deepest rooting
plants are doubtless those which grow on the sand hills.

384

Journal of Agricultural Research

Vol. I, No. 5

reached by the Artemisia roots. The plants then lose many of their
leaves and make no further growth until the following spring.1

The total transpiring surface is small in proportion to the size of the
plant, especially where the physical conditions are least favorable, and
this helps to prevent rapid exhaustion of the available soil moisture.
The limited amount of new growth made during exceptionally dry seasons

diminishes the dan¬
ger of death from
drought. Another
circumstance which
serves as a protection
from this danger is
the thinness of the
stand. Even on the
sand hills, where the
conditions are most
favorable for their
growth, the sage¬
brush plants are
rarely crowded. In
proportion as the
soil-moisture condi¬
tions depart more
and more from the
optimum for this
species, the plants
are farther and far¬
ther apart. Each in¬
dividual (PI. XLIV,
fig. 1) is surrounded
by a space of ground
which is bare during
the greater part of
the year, although producing a few shallow- rooted herbaceous plants in
spring and early summer. The wide spacing of the plants is indicated in
figure 3.

Effects of Disturbing Factors: Successions

During the summer and autumn large areas of sagebrush are often
burned over. The fire consumes the dry herbaceous growth and the
sagebrush plants are usually burned to the ground. They do not sprout
up from the old stumps, and the result is usually the complete removal
of the Artemisia. In the following year a mat of herbaceous vegetation,
composed chiefly of Bromus tectorum and Erodium cicuiarium , covers the

1 Sagebrush is therefore to be classed as a “ drought-enduring’ ' species. See Kearney and Shantz, op.
dt., p. 354, 355-

Fig. 5. — A small plant of sagebrush ( Artemisia tridentata), showing the
deeply penetrating taproot and good development of superficial lateral
roots typical of this species.

Feb. i6, 1914

Indicator Significance of Vegetation

385

ground among the blackened stumps. After a few years Gutierrezia is
likely to become the dominant plant on these burned-over areas. (PI.
XLV, fig. 1.) This, in turn, is followed by the sagebrush, which gradu¬
ally reestablishes itself.

In sagebrush land which has been plowed up and subsequently aban¬
doned, the removal of the shrubs favors the development of various annual
and biennial weeds, such as Bromus tectorum , alfilaria (Er odium cicuta-
rium), pigweeds (species of Amaranthus), Sunflower (. Helianthus annuus ),
wild tomato (Solanum triflorum ), vervain ( Verbena bracteosa) , etc. As
time goes on, Gutierrezia sarothraey a small, much-branched, yellow-
flowered composite, often becomes established and maintains itself
for a period which is short or long accordingly as the conditions are
more or less favorable for the reestablishment of the sagebrush. Sooner
or later the Artemisia reappears (PI. XLV, fig. 2), and unless fire or
some other disturbing factor intervenes, the territory is eventually
reconquered by the original association. (PI. XLV, fig. 3.)

The succession after either fire or breaking may be shortened, Arte¬
misia following immediately after the annual weed stage, without the
intervention of Gutierrezia. As a rule, however, the succession com¬
prises (1) a growth of annual and biennial weeds, (2) a growth of the
perennial Gutierrezia, and (3) the return of the original sagebrush
vegetation.

Grazing does not materially alter the sagebrush vegetation, although
diminishing the numbers of many of the herbaceous species. Artemisia
tridentata itself is rarely eaten and is, in fact, benefited by grazing, since
the plants which compete with it for the soil moisture are thereby
removed.

Variations from the Typical Association

Sagebrush with Kochia and with Shadscale. — Near the lower limit
of the main area occupied by sagebrush this association comes into con¬
tact with the Kochia and shadscale associations, and the dominant species
of the three associations often grow together in a mixed community. The
plants of Artemisia which push out farthest into areas occupied by these
other associations are confined to drainage channels, depressions, and
the vicinity of animal burrows. In such places the conditions as to
soil moisture are more favorable and the greater penetration of the
rain water has leached the salts into lower depths of soil than is generally
the case in Kochia and shadscale land. But along the frontiers of these
associations scattered, small, and sickly looking plants of Artemisia
mingle directly with Kochia or with shadscale.

Borings made where Artemisia tridentata and A triplex confertifolia grow
side by side invariably showed the presence of salts in the second, or,
at the deepest, in the third foot of the soil. (Table VI.) The sagebrush
roots are unable to penetrate this saline subsoil, and the total quantity
of water available for the growth of this plant is correspondingly limited.

386

Journal of Agricultural Research

Vol. I, No. 5

Table: VI. — Salt content of the soil at points where Artemisia tridentata and Atriplex

confertifolia grew side by side.

Depth of
soil.

Salt content in boring No. —

Feet.

Per cent .

o. 06
.27

Per cent.
O. 08
• 53

Per cent.
O. 12

• 53

I. 02

I. 36

"JO 0 §

Os-fik -fc. SS

Sagebrush with Juniper. — The Utah juniper ( Juniperus utahensis)
is abundant on the lower slopes of the mountains and also pushes
down into the upper part of Tooele Valley, where it occurs scatteringly
in the midst of areas occupied by the typical sagebrush association (see
background of PI. XLIV, fig. i), as well as on the sand hills. The pres¬
ence of juniper away from the sand hills usually indicates a stonier soil
than that on which the typical sagebrush association occurs.

SAND-HILL MIXED ASSOCIATION
Topographical Relations

The sand-hill mixed association covers a limited area towards the
south end of the valley, lying directly in the path of the winds from the
southwest which sweep over the low divide separating Tooele Valley
from Rush Valley. Even when the air is nearly motionless in other
parts of the valley, a sandstorm may often be seen blowing in this quar¬
ter. The sand is mostly heaped in dunes, which form more or less
continuous ridges having a general north and south trend. In places
where “ blow-outs” have taken place the ground is sometimes bare,
but for the most part it is fairly well covered with vegetation.

Appearance and Botanical Composition

As is usually the case in arid regions, the vegetation of the sand hills
is characterized by the presence of a large number of species — far more
than occur in any other plant association of Tooele Valley. The appear¬
ance of the vegetation as viewed a short distance away is determined by
the presence of a few woody species, notably sagebrush (Artemisia triden¬
tata) and juniper ( Juniperus utahensis). Sagebrush is much the most
abundant of the woody plants of the sand hills, and the individual
plants of this species which grow there are the largest and thriftiest
found anywhere in Tooele Valley under natural conditions.

The Utah juniper is fairly abundant on the sand hills. It occurs
as a large shrub or small tree, rarely exceeding io feet in height. Two
species of rabbit brush (Chrysothamnus nauseosus albicaulis and C .
pumilus) are also common, while the remaining woody species of this

Feb. 16, 1914

Indicator Significance of Vegetation

387

association, Atriplex canescens , Grayia spinosa , Sarcobatus vermictdatus,
and Purshia iridentaia , are relatively infrequent. The predominance
of woody plants distinguishes the sand-hill association of Tooele Valley
from the corresponding type of vegetation in the Great Plains east of the
Rocky Mountains.1

Next to the shrubs, perennial herbs are the most important members
of this association. Noteworthy among these are two characteristic
sand-loving species, Psoralea lanceolata and Abronia salsa. Certain
bunch grasses, Eriocoma cuspidata , Stipa comata , and Agropyron spica -
turn, are also important constituents of this vegetation. A few annual
and biennial species are to be seen during the first weeks of summer, but
the plants are too small and too short lived to greatly influence the
appearance of the vegetation.

The following list includes the more important species noted as occur¬
ring in the sand-hill mixed association:

PERENNIAL species

Juniperus utahensis (Englm.) Lemm.
Agropyron spicatum (Pursh) Rydb.
Eriocoma cuspidata Nutt.

Stipa comata Trin. and Rupr.
Eriogonum ovalifolium Nutt.
Eriogonum kearneyi Tidestrom
Atriplex canescens (Pursh) James
Eurotia lanata (Pursh) Moq.

Grayia spinosa (Hook.) Moq.
Sarcobatus vermiculatus (Hook.) Torr.
Abronia salsa Rydb.

Purshia tridentata (Pursh) DC.

Psoralea lanceolata Pursh
Gilia pungens (Torr.) Benth.

Lap pula occidentals (Wats.) Greene
Castilleja linariaefolia Benth.

Artemisia tridentata Nutt.

Chrysothamnus nauseosus albicaulis (Nutt.)
Rydb.

Chrysothamnus pumilus Nutt.

Layia glandulosa H. and H.

Senecio uintahensis A. Nels.

ANNUAL AND BIENNIAL SPECIES

Abronia cycloptera Gray
Eriogonum cernuum Nutt.
Lepidium pubecarpum A. Nels.
Erodium cicutarium I/Her.

Gilia leptomeria Gray
Crytanthe sp.

Lap pula sp.

Physical Conditions Indicated

The soil is nearly pure sand and is easily moved by the wind. The con¬
ditions for penetration of the total rainfall are excellent and the run-off
is negligible. The great depth of loose soil is favorable to storage of
water during a long period after rains. Only one soil boring in this asso¬
ciation was made (June 3), but the location was apparently in all re¬
spects a typical one and the resulting data (Table VII) probably repre¬
sent the average conditions of moisture and salt content of the soil where
this type of vegetation occurs.

1 Shantz, H. L-, Natural vegetation as an indicator of the capabilities of land for crop production in the
Great Plains area, U. S. Dept. Agr., Bur. Plant Indus. Bui. 201, p. 58-60. 1911.

388

Journal of Agricultural Research

Vol. I, No. 5

Table VII. — Sand-hill mixed association: Moisture conditions and salt content of

the soil in a typical area.1

Moisture

content

Depth

Moisture

Wilting

above or

Salt

(feet).

equivalent.

coefficient.

below the

content.

wilting

coefficient.

9.2

5*0

— 0. 1

0. 03

9*7

5- 3

+0. 7

•03

6. 2

3*4

+ 1.1

•03

5-8

3* 1

+ 1*3

.04

. 01

1 All data are in percentages of the dry weight of the soil.

If the data given in Table VII may be taken as representative, land
occupied by this association is characterized by the following soil con¬
ditions: (i) A low moisture-holding capacity, as indicated by the low
moisture equivalents, (2) available moisture present, at least during the
fore part of the summer, at a depth attainable by the more deeply pene¬
trating roots, and (3) a very low salt content.

Adaptations to the Physical Conditions

The soil-moisture conditions of the sand hills are obviously such as to
favor plants with deeply penetrating roots, and, accordingly, large woody
plants are predominant in this association. Sagebrush, the most abun¬
dant woody species, is noteworthy for the great depth reached by its tap¬
root when the conditions are favorable. Of the herbaceous species of this
association, some have a well-developed taproot, while others produce an
abundance of superficial roots. The shallow-rooted herbs, being de¬
pendent upon the moisture of the surface soil, mostly complete their
growth and ripen seed early in the summer. Certain of the perennial
herbs, notably Psoralea lanceolata , spread by slender, creeping rootstocks
and can therefore withstand frequent burial. This plant is excellently
adapted to colonizing the blow-outs and may be regarded as the pioneer
plant of the moving sands.

KOCHIA ASSOCIATION1
Topographical Relations

The Kochia association (PI. XLVI) occupies a narrow and nearly con¬
tinuous belt which extends across the valley along the lower boundary of
the sagebrush area and lies between the latter and the shadscale area.
(See map, PI. XLII.) This type of vegetation likewise occurs as islands
of greater or less extent scattered through the sagebrush zone well

1 While this plant association is one of the most important in Tooele Valley, it appears to be a much less
prominent feature of the vegetation in other portions of central and western Utah.

Feb. 1 6, 1914

Indicator Significance of Vegetation

389

toward the head of the valley. (PI. XLIII, fig. 2.) The boundaries
between the areas occupied by the Kochia and by the sagebrush asso¬
ciations are usually very sharply defined. Equally abrupt is the change
in salt content of the soil, as is well exemplified by the results of borings
which were made on either side of the line of contact shown in Plate XLVI,
figure 1. The location of the boring in the sagebrush association is
marked by the position of the soil tube shown in the illustration. The
two borings were only 20 feet apart. The results are given in Table VIII.

TABLE VIII. — Salt content {in percentages of the dry weight of the soil) on either side of
a line of contact between the sagebrush and Kochia associations .

Depth (feet).

Sagebrush.

Kochia.

0.03

O.31

•03

I.49

On the other hand, the boundaries between the Kochia and shadscale
associations are usually indefinite.

Kochia vestita , sometimes locally known as “white sage,” which is the
dominant species of the Kochia association, is also a frequent component
of the shadscale and greasewood-shadscale associations, reaching the
shore of Great Salt Lake with the latter association. But the small size
of the plants as compared with those of shadscale and of greasewood
makes Kochia an inconspicuous member of these associations.

In typical portions of this association Kochia vestita is almost the only
species of flowering plant, except that where grazing animals are kept
off the land, a small grass, Poa sandbergii , is very abundant. Eew other
species occur, and these are seldom represented by numerous individuals.
The following list includes all species of flowering plants which were noted
as occurring in typical areas of the Kochia association :

Botanical Composition

Sphaerostigma pubens (Wats.) Rydb.
Opuntia sp.

Gutierrezia sarothrae (Pursh) B. and R,
Appearance

The Kochia association is the most uniform in appearance of the types
of vegetation occurring in this valley (Pi. XLVI, fig. 2). It is virtually a
1 -species association. The height of the plants varies but slightly over
large areas and usually does not exceed 6 inches. Owing to the low
growth and the hairiness of the plants (see text fig. 7 and PI. XLVI, fig. 3),
an area of Kochia presents at a little distance the homogeneous appear¬
ance of a gray blanket. Even at a distance of several miles, the strips

Kochia vestita (Wats.) A. Nets.
Poa sandbergii Vasey
Erodium cicutarium L/Her.
Lepidium sp.

24395 —r4 - 3

390

Journal of Agricultural Research

Vol. I, No. s

and islands occupied by this plant are easily distinguishable from sur¬
rounding areas of sagebrush vegetation (PI. XLIII, fig. 2). The contrast
is especially striking in spring and early summer when the sagebrush has
a silvery green color, which is quite distinct from the dull gray of the
Kochia.

When viewed closely (PI. XLVI), the plants are found to be sep¬
arated by patches of bare ground which vary in size as the physical
conditions are more or less favorable. In a more distant view the light
ashy color of the soil occupied by this association blends with the color
of the plants themselves, and this tends to create the impression

that the plant cover¬
ing is dense. In
fenced areas occupied
by this association the
color is modified to a
yellowish or brownish
gray during a few
weeks in the early
part of the summer,
owing to the abun¬
dant fruiting heads of
a small grass, Poa
sandbergii (PI. XLVI,
fig. 3). But most of
this land is grazed by
sheep, which soon ex¬
tirpate the grass or at
least prevent its flow-

Fig. 6. — A representative io-meter quadrat of the Kochia association, erfilg, while leaving
showing the location of each tuft of Kochia ( K ) and of each matlike Kochia practicallv

colony of Poa sandbergii ( P ). * *

undisturbed. The

distribution of the plants of Kochia and of Poa in a typical, unmodified
quadrat of this association is shown in figure 6.

Physical Conditions Indicated

The type of land occupied by the Kochia association in its typical form
is uniform and well defined. The soil is remarkably homogeneous to a
depth of several feet, fine in texture, and close in structure. Unlike sage¬
brush and shadscale lands, there is usually little gravel present. The
smooth, polished condition of the surface after it has been wet indicates
that this soil puddles readily. The conditions for the penetration of
water are, therefore, unfavorable, and the run-off is doubtless high.

Feb. x6, 1914

Indicator Significance of , Vegetation

39i

TablU IX. — Kochia association: Moisture conditions and salt content of the soil in

typical areas.1 * * 4

Date of collection.

Item.

No. of sample.

Moisture equivalent.

Wilting coefficient.

Depth
of soil
(feet).

Moisture content above
or below wilting coeffi¬
cient .

Salt content.

June.

July.

August.

Aver-

age.

105

116

23. 0

35- 8

21. 7

26. 0

24. 7

23-9

2 5-&

25- 5

25* 2

30*3

26. 4

29.4

25-3

27. 0

33-3

35* O

29- 4

34*5

36. 8

32-4

33* 5

31-4

34-2

24.4

35*o

34-3

32. 5

31-9-

12. 5

19. 5

11. 8

14. 1

13. 4

13. 0

14. 0

13- 9

13-7

16. s

14*3

16. 0

13- 7

14. 7

18. 1

18. 7

17. 6

18.2

17. 0

18. 6

19. O

18. 6

17. 7

17. 3

— 1.8

— 6. 2

—5.8

— 5. 4

— i. s

— 2. 6

+ 1. 6

— 6. 1

— 1.8

— 5* 4

— 2. 0

— 2- 0

— 2. 5

— 3. 2

+ i-6

— 2. 4

•05

.08

.04

. 22

• 14

.07

.06

.18

•3i

•05

. 12

• 3°

. 16

.14

1. 20

•30

•47

•32

.80

1. 49

•3i

•55

.82

. 60

• 52

1. 14

x. 56

1. 10

.88

1.36

.92

1. 02

• 70

.78

1. 02

i*43

•97

1. 76

1. 10

1. 11

1 All data in this table are stated in percentages of the dry weight of the soil. The moisture contents
with a plus sign ( + ) represent moisture available for growth (above the wilting coefficient), while those
with a minus sign ( — ) represent a corresponding deficit of available moisture (below the wilting coefficient).

Soil, Moisture. — The moisture equivalents given in Table IX indi¬
cate that the moisture-holding capacity of the soil is much higher in
Kochia land than in sagebrush land. The moisture contents in typical

areas show that as early as the first of June, 1912, the soil to a depth of

4 feet was usually devoid of water available for plant growth. The
deficit was usually greatest in the surface foot, partly no doubt because
of surface evaporation and partly because of the shallow-rooting habit of
Kochia vesiita.

Salinity. — The soil in typical Kochia land, at least in Tooele Valley,
is usually free from an injurious quantity of salts in the surface
foot. On the other hand, the second foot is usually, and the third and
fourth feet are almost invariably, highly saline. In places where the
surface foot contains much salts the plants of Kochia are scattered and
stunted.

There is some evidence that the presence of Kochia vegetation, al¬
though in the great majority of cases associated with a strongly saline
subsoil, does not* invariably indicate such a condition. In the upper
part of Tooele Valley an island of Kochia (PI. XUII, fig. 2) several acres
in extent, in the midst of the sagebrush zone, was found to be underlain
at a depth of 30 inches by a gravelly hardpan. The soil just above this
stratum contained only about 0.2 per cent of readily soluble salts. It
would seem that here the presence of hardpan rather than of salts had
caused the Artemisia to give place to Kochia.

Vol. I, No. $

392 Journal of Agricultural Research

The conclusion seems warranted that the presence of the Kochia
association to the exclusion of sagebrush is determined by the occurrence
of 1 or at most 2 feet of soil free from an excess of salts, underlain by a
subsoil which is strongly saline or which for some other reason is un¬
favorable to deep penetration of roots.

Summary or Physical Conditions. — In Tooele Valley the land occu¬
pied by the Kochia association is distinguished from that occupied by
the sagebrush association by its finer texture, its tendency to puddle at
the surface and, hence, resist the penetration of water, and its higher
moisture-holding capacity, and also by the limitation of the depth in
which the roots can freely develop to not more than 24 inches, the
obstacle to deeper penetration being usually the high salt content of the
subsoil.

Adaptations to Physical Conditions

Since Kochia vestita is the only very important species of the Kochia
association, the structure of this plant alone need be considered in its

Rig. 7. — Kochia vestita: A, Detail, showing the narrow, hairy leaves; B, a plant showing the shallow root
system and the propagation by root shoots.

relations to the physical conditions. The underground portion of the
plant (fig. 7) is well adapted to the comparatively small depth of soil
from which the total supply of water must be obtained. Kochia vestita
spreads by means of long, slender-branching root shoots, which extend
almost horizontally for distances of 10 feet or more, and often at a depth
of only 3 inches from the surface of the soil. At intervals clusters of
vertical shoots are sent up, and, hence, the plants above ground appear as
isolated, unconnected clumps. In typical portions of this association the
feeding roots are limited to the first 12 to 18 inches of the soil, the depth
which is usually free from excessive quantities of alkali salts.

At one point where the root distribution was studied with special care,
living roots were traced to a depth of about 18 inches, and at that depth
the soil contained 0.9 per cent of salts, while at a depth of about 21
inches there was 1.6 per cent. Below this depth traces of dead roots
were observed. Excavation at another point, where the first 6 inches

Feb. 16, 1914

Indicator Significance of Vegetation

393

of the soil contained about 0.15 per cent of salts, the second 6 inches
0.36 per cent, and the second foot 1.2 per cent, showed only dead roots
below the depth of 12 inches. A possible explanation of these circum¬
stances would be that in some past period of exceptionally heavy rain¬
fall the salt had been washed down to an unusually low depth and that
the growth of the roots had kept pace. In subsequent years an upward
movement of the salts would have resulted in the death of the deeper
roots.

The total quantity of water available for growth in Kochia land is
probably less than in any other type of land in the valley. The quan¬
tity of organic matter produced is also less, and although the plants
often remain alive throughout the greater part of the summer the total
quantity of water transpired is necessarily small.

Poa sandbergii , the only other abundant species of this association, is
a shallow-rooted grass, which ripens its seeds and withers to the ground
early in the summer. It is clearly dependent upon the moisture avail¬
able in the surface soil.

Effects of Disturbing Factors: Successions

Where Kochia land has been plowed so as to completely destroy the
original vegetation and subsequently has been abandoned, the reestab¬
lishment of the Kochia probably takes place rather slowly. When the
“breaking" has been less thorough and a few plants have been left alive,
the reestablishment of the Kochia proceeds more rapidly. The interven¬
ing stage of annual weeds or of Gutierrezia, such as occurs when the
vegetation has been removed from sagebrush land, apparently does not
follow after breaking on Kochia land.

Grazing is general in Tooele Valley, where many sheep are wintered.
Kochia land is especially suitable for pasturage, being relatively level
and free from spiny shrubs. The Kochia plants themselves are usually
not much injured by grazing, but the associated grass (Poa sandbergii)
is eaten to the ground and is often almost wholly eradicated.

Variations from the Typicae Association

Kochia with Sagebrush. — As was pointed out above, Artemisia
iridentata penetrates Kochia areas along drainage channels and in other
places where the soil-moisture conditions are more favorable and the
salt content smaller than in typical Kochia land. When associated with
sagebrush, the plants of Kochia are much larger and more vigorous than
where this plant occurs in the pure association.

Kochia with ShadscaeE. — On the lower edge of the Kochia zone
plants of shadscale appear, scatteringly at first, then in greater numbers,
until finally the two species are found mingled together in approximately
equal proportions over large areas. The shadscale, being much the larger
plant, is alone visible at a short distance, even where it is numerically

394

Journal of Agricultural Research

Vol. I, No. s

not superior to the Kochia. The line of demarkation between the
Kochia and shadscale associations is never a sharp one, and this con¬
forms with the fact that the physical conditions indicated by the two
types of vegetation are similar.

shadscale association
Topographical Relations

The shadscale association is one of the most characteristic and important
of the Great Basin region In the Tooele Valley (see map, PI. XLII) it
occupies a wide belt across the middle part of the valley, just below the
Kochia belt, extending farthest northward along the base of the Stans-
bury Range. The dominant species as a constituent of the greasewood-
shadscale association extends to the edge of the grass flats and beyond
that area occupies ridges and hummocks on the salt flats which border
Great Salt Lake. Isolated small patches of pure shadscale also occur
within the area mapped as salt flat.

Botanical Composition

The most abundant plant of this association is the species of saltbush
(Atriplex confertifolia) which is commonly known as shadscale, from the
shape of the scalelike bracts which envelop the fruits. (See fig. 9.) The
number of associated species is much smaller than in the sagebrush
association, and those which occur are usually represented by fewer
individuals. The plants which were noted as occurring in this associa¬
tion are the following :

perennial species

Poa sandbergii Vasey
Sitanion minus Smith
AUium acuminatum Hook.
Atriplex confertifolia (Torr.) Wats.
Eurotia lanata (Pursh) Moq.
Kochia vestita (Wats.) A. Nels.

Opuntia sp.

Lappula occidentalis (Wats.) Greene
Artemisia spinescens Eat.
Chrysothamnus marianus Rydb.
Tetradymia glabrata Gray
Tetradymia spinosa H. and A.

ANNUAL AND BIENNIAL SPECIES

Bromus marginatus seminudus Shear
Lepidium jonesii Rydb.

Thelypodium elegans Jones

Cryptanthe multicaulis A. Nels.
Oreocarya shantzii Tidestrom
Townsendia watsonii Gray

Appearance

The general appearance of the shadscale association (PI. XLVII, fig. 1)
is due almost entirely to the dominant species. Atriplex confertifolia as
found in Tooele Valley is a rounded bush, usually about 18 inches in
height and also in diameter, with rigid, spiny branches and harsh dry¬
looking foliage. (See fig. 9.) The individual plants tend to form low
hummocks, the soil immediately about them being held by the partly

Feb. 16, 1914

Indicator Significance of Vegetation

395

recumbent, twisted branches, while the bare ground between is subject
to blowing.

The prevailing color is a dull grayish brown, turning to reddish brown
in autumn. Plants growing in depressions, where the moisture condi¬
tions are exceptionally favorable, have a bluish hue. Viewed from a
short distance the association gives the impression of extreme monotony
and lifelessness.

The distribution of the plants is indicated in figure 8, which repre¬
sents a quadrat 10 meters square, in a typical portion of this asso¬
ciation. Where the
conditions are most
favorable, the plants
have a fairly vigorous
appearance and cover
somewhat more than
half the ground, the
stand being frequently
more dense than is
usually the case in the
sagebrush association.

In much the greater
part of the area, how¬
ever, the proportion
of bare ground is
greater and the plants
seem to be having a
hard struggle to main¬
tain life, many of their
branches being dead
or dying. (Pl.XLVII,
fig. 1 . ) No other vege¬
tation in this valley
gives the impression of
being so nearly con¬
quered by the environment. Even the few species which grow on the
salt flats have the appearance of finding their habitat more congenial.

The associated species contribute scarcely at all to the general appear¬
ance of the association. Annuals are of very minor importance. The
small shrubs of the family Composite which occur here and there are
too few in number of individuals and are too much like the. shadscale in
habit of growth and dullness of color to influence materially the aspect
of the vegetation. Kochia vestita is associated with the shadscale in
extensive areas where the vegetation appears otherwise typical of the
present association. The much smaller size of the Kochia plants makes
them inconspicuous.

Fig. 8. — A representative io-meter quadrat of the shadscale association,
showing the location of each individual plant of A triplex confertifolia
(A), the only woody species present, and of Opuntia sp. { O ). The
figures show the number of main stems and, hence, indicate the size
of the plant. A circle around the letter indicates that the plant is
dead. Seedlings of Atriplex are indicated by the small a. The an¬
nual grass Bromus (not indicated on the quadrat) was very abun¬
dant around the Atriplex bushes.

396

Journal of Agricultural Research

Vol. I, No. s

Physical Conditions Indicated

The conditions in shadscale land as regards moisture and salt content
of the soil are shown in Table X, which gives the results of various
borings in typical areas.

Table X. — Shadscale association: Moisture conditions and salt content of the soil in

typical areas.1

Date of collection.

Item.

No. of sample.

Depth
of soil
(foot).

Moisture equivalent.

Wilting coefficient.

Moisture content above or below
wilting coefficient .

Salt content.

June.

0 10
•36
.40

0. 05

•05

. 22

24. 2
3i-4
34-o
33* 7

13* 1
17.0

18.5

18.3

“6. 5

-4*5
-5. 1
-4.6

•OS

.44

.88

23-4

28.8

32.4

32. 6

12. 7
IS- 6
17. 6
17. 7

-5-9
S* 2
-4-8
—3- 6

.07
• 29
.82
•94

July.

21. 7
36. 1
35- 8
29- S

11. 8
19- 6
19- S
16. o

—4- 7
5* 2
—5- 2
-6.6

.06
. 22
.88
.88

August.

22. 4
27. 9
35-8
26. 9

IS- 1
19- 5
14. 6

-5- 2
—5.8
-4.8
-7. 1

. 06
. 27
1. 06
.80

108

0.05
.42
1. 14

117

o. 12
• 54
.88
1. 14

Aver¬

age.

22. 9
31-0
34*5
30*6

12.4

16.8

18.7

16.6

-5-6
-5- a
“5-0
“5*5

.07

•32

•78

•93

1 All data in this table are stated in percentages of the dry weight of the soil. The moisture contents
with a plus sign (+) represent moisture available for growth (above the wilting coefficient), while those
with a minus sign (— ) represent a corresponding deficit of available moisture (below the -wilting coeffi¬
cient).

Comparison with the corresponding data in Table IX shows little
difference in the physical conditions of the shadscale and Kochia land.
The surface foot of soil in the shadscale association usually contains more
gravel and is of somewhat lighter texture, as indicated by the somewhat
lower average moisture equivalent. This, together with the rougher
surface of the land, would indicate more favorable conditions for the
penetration of water.1 On the other hand, the second, third, and fourth
feet show a more constant and more pronounced deficit of available
moisture than is the case in Kochia land. At first glance this would
seem to confute the assumption that the conditions for the penetration
of water are better than on Kochia land, but it must be remembered
that in shadscale land, which supports much the heavier vegetation, the
loss of water by transpiration must be greater.

The average salt content at all depths down to 4 feet is somewhat
smaller in shadscale than in Kochia land, but in this respect the dif¬
ference between the two types is of small importance.

1 In some parts of the shadscale area, especially where Kochia vestiia is abundant, a tendency to the for¬
mation of hardpan at a depth of about 24 inches was noted, but this condition appears to be exceptional
in Tooele Valley.

Feb. i6» 1914

Indicator Significance of Vegetation

397

An 8-foot boring made in a portion of the shadscale area where Kochia
vestita was also abundant gave interesting results as regards the salinity
of the soil at greater depths than were reached by any of the borings
included in Tables IX and X. The percentage of salt contents at the
successive i-foot depths were 0.10, 0.79, 1.02, 0.98, 0.92, 0.88, 0.94, and
1.02, which indicates a very uniform condition as regards salt content of
the soil below the second foot and to an unknown depth.

The differences in the averages for each physical factor as given in
Tables IX and X scarcely seem of sufficient magnitude to explain the
separate occurrence of A triplex confertifolia and Kochia vestita in distinct
associations alternating over large areas, especially when we note that
some of the borings in typical portions of each association show almost
identical physical conditions. It is not surprising therefore that the line
of contact between the two associations is a vague one and that the two
species mingle on equal terms over areas of considerable extent. Yet
there is a possible explanation for the alternation of these two types
which is not brought out by the data given in these tables. In Kochia
land, because of the less favorable conditions for penetration, the sea¬
sonal total of available moisture may not be sufficient to support a stand
of shadscale in competition with Kochia.

The distribution of A triplex confertifolia appears to be limited toward
the upper end of the valley by the occurrence of light, easily permeable
soil which is free from an excess of salts to a depth of 3 feet or more. In
such land shadscale can not compete with sagebrush. Toward Great
Salt Lake it is confined to the drier, better drained land of the hummocks
and ridges and is excluded from the flats where the soil is excessively
saline and is wet to the surface during much of the year.

Areas of very limited extent are found here and there in which the
shadscale plants are much larger than the average and have a green,
thrifty appearance, with a notable absence of dead wood. In such spots —
generally obvious depressions — the soil conditions are more favorable
than in most of the shadscale area, the salt content being lower and the
moisture content higher. The results of a boring in one such spot, made
on July 13, are given in Table XI.

Table) XI. — Salt content and moisture conditions of the soil in a spot where Atriplex
confertifolia was exceptionally large and healthy.1

Depth (feet).

Salt content.

Wilting coeffi¬
cient.

Moisture con¬
tent above or
below wilting
coefficient.

0. 06

20. 5

— 10. 1

•05

17.4

— 2. O

.09

18.3

+ 2.8

.09

18. I

0. 0

1 All data in percentages of the dry weight of the soil.

39«

Journal of Agricultural Research

Vol. I> No. 5

The low salt content throughout the 4 feet, the relatively high wilting
coefficients, and the presence so late in the summer of available moisture

in the third foot are worthy of note.

Summary of Physical Conditions— The presence of the typi
shadscale association as it occurs in Toode Valley ntdtcates usuaUy
(i) a soil of finer texture, having a higher moisture equivalent than
savebrush land- (2) a deficit in midsummer of moisture available fo
pit growth; (3) a high salt content of the soil below the depth of 1 or
2 feet and 4) as compared with land occupied by the Kochia associa-

or scales, which, envelop the fruits.

tion, a somewhat lighter and more gravelly texture in the first foot and
a much more uneven surface — conditions which probaby
better penetration and, hence, in a larger seasonal total of water available
for plant growth than on Kochia land.

Adaptations to Physical Conditions

The dominant species, A triplex confertifolia, need alone be c°nsl
in this connection. As shown in figure 9, this plant has a well-developed

Feb. i6, 1914

Indicator Significance of Vegetation

399

taproot, its root system being, therefore, very different from that of
Kochia vestita.

The roots of shadscale, although by no means so deep as those of
Artemisia iridentata , doubtless as a rule penetrate and obtain water
from a greater depth of soil than do the roots of Kochia.1 This would
help to explain the fact that, notwithstanding the more favorable con¬
ditions for penetration of water, the deficit during periods of drought of
moisture available for growth in the second, third, and fourth feet is
normally greater in shadscale than in Kochia land.

The moribund appearance in 1912 of the shadscale plants in most of
the area covered by this association in Tooele Valley points to the con¬
clusion that in years of not more than average rainfall the moisture
supply is inadequate. Thus, in 1912 the moisture available for growth
had been exhausted to a depth of 4 feet, and the plants had begun to
shed their leaves before the end of June. Apparently, with the normal
thickness of stand in this association, the older individual plants can not
obtain sufficient water to maintain life in all parts of their bodies. The
branches are in active competition and the plant as a whole remains
alive only by sacrificing some of its members. The death of some of the
branches in almost every plant which has passed the seedling stage
reduces by so much the transpiring surface and results in a proportionate
economy of the scanty moisture available to each individual. To an
even greater degree than Artemisia iridentata this plant has the faculty
of remaining during a great part of the year in a nearly dormant condi¬
tion, while retaining some of its foliage.

Effects of Disturbing Factors: Successions

The exact stages in the revegetation of shadscale land from which the
original plant cover has been removed by fire or by the plow remain to
be worked out. There is evidence, however, that Gutierrezia sarothrae
forms an important stage in these successions. A large area near the cen¬
ter of Tooele Valley is covered with an almost pure growth of this small,
yellow-flowered plant of the Composite. While a part of this tract was
probably once occupied by sagebrush, the greater portion occurs in the
midst of the shadscale belt and has a strongly saline subsoil.2

1 Nevertheless, the Atriplex roots do not develop well in a strongly saline subsoil. Thus, at a boring
where the first foot of the soil contained o.i and the second 0.8 per cent of salts, few living roots were found
below the depth of 12 inches.

2 This plant shows marked adaptability to varying soil conditions. In areas having a saline subsoil,
which were presumably covered originally with shadscale, the plants of Gutierrezia are scattered and
small and have a superficial root system, while in nonsaline areas, where sagebrush was probably the
original vegetation, the stand is denser, the plants are larger, and a good taproot is developed.

400

Journal of Agricultural Research

Vol. I, No. 5

Variations from the Typical Association

The shadscale area in Tooele Valley comes into contact on its upper
limit with the sagebrush and the Kochia associations and on its lower
limit with the greasewood-shadscale association. In each case mixed
or transitional communities result. The conditions under which shad-
scale mingles with sagebrush and with Kochia have been discussed on
preceding pages. The transition to the greasewood-shadscale associa¬
tion, which is a very gradual one, will be treated in connection with the
latter association.

greasewood-shadscale association
Topographical Relations

The area occupied by the greasewood-shadscale association forms an
interrupted belt (see map, PI. XLII) across the valley between the areas
occupied by the shadscale association, and by the grass flats, respectively.
It also covers the low ridges and hummocks which alternate with the
basinlike depressions and flats near the shore of Great Salt Lake (PI.
XLIII, fig. i, and PI. XLVII, fig. 3). In general, it occupies all areas
where the water table is sufficiently high to make moist soil accessible to
the deep-rooting greasewood and where at the same time 1 or 2 feet of
the surface soil are sufficiently dry to permit the growth of shadscales
Where the water table is too low this association gives place to the pure
shadscale. On the other hand, as the soil becomes wet nearer and nearer
the surface, the shadscale gradually disappears and at the edge of the
grass flats greasewood associates with Allenrolfea occidentalis and Suaeda
moquinii instead of with A triplex con jerti folia.

Botanical Composition

This type of vegetation is dominated by two shrubby species, grease¬
wood (Sarcobaius vermiculatus; see fig. 10, p. 404) and shadscale (A triplex
confertifolia; fig. 9, p. 398). In typical areas these plants intermingle in
approximately equal numbers, but on the highest ground (PI. XLVII,
fig. 2) shadscale is strongly predominant, while on the lowest land where
this association occurs greasewood is the more abundant. Kochia vestita
is abundant in much of the area occupied by this association, but the
plants are so small in comparison with the two dominant species that
they do not affect the general appearance of the vegetation. In spots of
limited size greasewood and Kochia are associated, shadscale being
absent. The soil conditions in such spots do not differ materially from
those of the typical greasewood-shadscale association. Few other species
are found, and of these the number of individuals is usually small. The
following list includes all species noted as occurring in the greasewood-
shadscale association :

Feb. 16, 1914

Indicator Significance of Vegetation

401

PERENNIAL SPECIES

Elymus condensatus Presl
Poa sp. (P. sandbergii Vasey?).
Sitanion minus Smith
Atriplex confertifolia (Torr.) Wats.
A triplex nuttallii Wats.

Kochia vestita (Wats.) A. Nels.

Sarcobatus vermiculatus (Hook.) Torr.
Suaeda moquinii (Torr.) A. Nels.
Suaeda intermedia Wats.

Lap pula occidentals (Wats.) Greene
Gutierrezia sarotkrae (Pursh) B. and R.
Tetradymia nuttallii T. and G.

ANNUAL AND BIENNIAL SPECIES

Bromus tectorum I,. Sophia pinnata (Walt.) Howell

Erysimum asperrimum (Greene) Rydb. Machaeranthera canescens (Nutt.) Gray

The local distribution of most of these plants varies greatly within the
area occupied by the association, probably because of the great diversity
in the depth to permanent moisture and at which the subsoil becomes
strongly saline.

Appearance

This type of vegetation is less monotonous in its appearance than the
sagebrush, Kochia, and shadscale associations, owing to the strong
contrast in color and usually in size between the two dominant species.
(PI. XLVII, fig. 2.) Greasewood has a bright-green color, changing to
yellowish later in the season, and appears dark when photographed
against the sun. Shadscale, on the other hand, has a dull brownish gray
hue. The former plant often reaches a height of 4 or 5 feet, while the
latter seldom exceeds 2 feet.

At the highest elevations occupied by this association there is sufficient
moisture for the growth of greasewood only along drainage channels,
and the general surface of the land is covered with pure shadscale.
Somewhat farther toward Great Salt Lake plants of greasewood are
scattered among the shadscale, although much less numerous than the
latter. Finally near the borders of the grass flats and on the ridges and
hillocks which intersect the salt flats the two species grow side by side
on more or less equal terms, and their colors blend when the vegetation
is viewed from a short distance.

Physical Conditions Indicated

The soil moisture and salinity conditions, which characterize typical
portions of the land occupied by this association, are indicated by the
data in Table XII. Comparison with Table X will bring out the differences
between this environment and that of the pure shadscale association.

402

Journal of Agricultural Research

Vol. I, No. s

Table? XII. — Greasewood-shadsca le association: Moisture conditions and salt content of

the soil in typical areas.1

Date of collection.

Depth

soil

Item.

June.

July.

(feet).

Aver-

age.

No. of sample .

22. O

26. 5

26. I

28. 5

15 -5

24. 1 -

Moisture equivalent . .

J 2

22. 9

30. 8

34- 2

32. 2

24. 8
20. 8

x4. 7

26. 6
26. 1

1 3

*3-3

35-3

31* 5

33-5

22. 2

l 4

27. 0

30. 2

31-5

31-3

17*5

24. 2

26. 9

11. 9

14.4

14.2

15 -5

8.4

13- 1

Wilting coefficient .

12. 4

16. 7

18.6

17- S
18.2

13- 5

8. 0

14.4

7.2

19. 2

17. 1

11. 2

12. 1

14* 2

14. 7

16. 4

17. 1

17.0

9-5

13- 1

14.6

— . I

- • 5

— 4-4

— 4.4

- -3

-4-7

— 2- 47

Moisture content above or below

1 2

+ 4-5

+ 6.1

+ 4-1

+ 3* 2

+ 3-1

+ ■ 4

+ 3-6

wilting coefficient .

1 3

+ 3-5

+10. 2

+ 4-1

+ 4-3

+ 8-8

+ 3*3

+ 5-7

l 4

+ 1.0

+ 8.4

+ 3*9

+ 2.8

+ 13- 1

+ 2. s

+ 5* 3

.08

0. 05

.64

.27

■ 54

. 61

• 23

•34

. 61

•38

1. 24

■ 54

1. 10

. 61

, 60

. 72

Salt content .

. 62
1. 30

.82

1.85

1.36

*65
1. 05

1. 36
I. 48

.68
• 76

1*25

1.03
i- 15

1. 96

1. 20

1. 58

1 All data in this table are stated in percentages of the dry weight of the soil. The moisture contents
with a plus sign (+) represent moisture available for growth (above the wilting coefficient), while those
with a minus sign (— ) represent a corresponding deficit of available moisture (below the wilting coefficient).

Soil, Moisture. — The moisture-holding capacity of the soil, as indi¬
cated by the moisture equivalent, is somewhat higher in the first foot,
but is lower in the second, third, and fourth feet than in the shadscale
association. It is significant that moisture available for the growth of
plants was present in considerable quantity during the months of June
and July in all but the surface foot in the greasewood-shadscale associa¬
tion, while in the shadscale association during the same months there
was a marked deficit of available water to a depth of 4 feet. The rela¬
tively high moisture content is correlated with the relatively slight
elevation above the level of water in the lake and with a consequently
high ground-water table.

Salinity. — The average salt content of each of the first 4 feet of the
soil is much higher than in the shadscale association, the difference being
especially marked in the second foot, which contains, on the average, as
much salts as does the third foot in land occupied by pure shadscale.

Summary of Physical Conditions. — In Tooele Valley the presence
of typical greasewood-shadscale vegetation indicates soil conditions as
follows: (1) A fairly high moisture equivalent; (2) the surface foot well
drained and usually dry during the summer; (3) moisture available for
the growth of plants present throughout the summer at a comparatively
slight depth ; (4) a high salt content from the second foot downward and
often in the surface foot as well. /

Feb. x6, 1914

Indicator Significance of Vegetation

4°3

Adaptations to the Physical Conditions

The two dominant species have somewhat different soil requirements,
and the land occupied by this association offers a combination of con¬
ditions which permits them to grow side by side. Greasewood prefers
an ample and permanent supply of moisture within reach of its roots,
and its strong, deeply penetrating taproot (fig. 10) enables it to reach
moisture in places where the surface soil is dry and the ground-water table
is at a considerable depth. This plant can live in soil which is moist to
the surface, although under such conditions the plants are never as
large and vigorous as where a higher elevation and a subsoil of light
texture afford better drainage. Shadscale, on the other hand, does not
thrive with its roots in wet soil, and its presence is usually a reliable
indication that at least the surface foot is dry during the greater part of
the summer.

Greasewood {Sarcobatus vermiculatus) grows in a greater variety of
habitats than any other flowering plant of the Tooele Valley. It was
found in one place or another in company with the dominant species
of all of the leading associations. In much the greater part of its range
in the valley greasewood is associated with shadscale, but there are excep¬
tions to this rule. The largest and thriftiest looking greasewood plants1
grew on the summits of dunes of pure sand, together with sagebrush,
juniper, Eriocoma, Abronia, Eriogonum, Psoralea, and other character¬
istic plants of the sand-hill mixed association. Shadscale is absent from
this community. At the other extreme greasewood occurs in company
with Allenrolfea in land which is too wet and saline for the growth of
shadscale. The widely different conditions in these two environments
are indicated by the data in Table XIII.

Table XIII. — Moisture equivalent and salt content of the soil where Sarcobatus
vermiculatus occurred — on the sand hills and with Allenrolfea.1

Depth

(feet).

Moisture equivalent.

Salt content.

On sand
hills.

with

Allenrolfea.

On sand
hills.

with

Allenrolfea.

6. 2

31.0

0. 09

2. 16

6.8

37-3

.08

2. 08

6. 1

27. 7

. 14

1. 76

7.0

25- 9

. 16

1. 25

1 All data in percentages of the dry weight of the soil.

The growth of greasewood on the sand hills makes it evident that this
plant is not an infallible alkali indicator, although in the great majority of
cases its occurrence is associated with an excess of salts in the soil, and in
its ability to endure the presence of alkali it is surpassed by few other

!The individual alongside the boring made in the sand hills (see Table XIII) was 6 feet high, iofeet
across, and had several stems which were from i to 2 inches in diameter at the surface of the ground.

Journal of Agricultural Research

Vol. I, No. 5

4°4

Fig. io. Sarcobatus vermiculatus (greasewood) : A , Detail showing the narrow, rather fleshy leaves; S, a
plant showing the excellent root development. The large, deeply penetrating taproot is characteristic
of this species.

Feb. 16, 1914

Indicator Significance of Vegetation

405

flowering plants.1 A condition which is almost always correlated with the
presence of greasewood is a permanent supply of moisture available for
growth within the depth of soil penetrated by its roots.

GRASS-FLAT COMMUNITIES 2
Topographical Relations

The grass-flat of vegetation occurs in an interrupted belt (see map,
PI. XUI), which crosses the northern part of the valley and lies
between the area occupied by the main body of the greasewood-shadscale
association and the salt flats. It covers a gently sloping or nearly level
expanse and appears to be lower in elevation than some of the ridges and
hillocks situated between it and the shore of the lake. The area is thus
somewhat analogous to a coastal lagoon and may have had a similar
origin. It is characterized during the greater part of the year by a very
moist condition of the soil, due probably in part to seepage.

Appearance and Botanical Composition

The vegetation of the grass flats shows considerable diversity.
Several plant communities can be distinguished, although the boundaries
are rarely very sharp. The two most important of these are characterized
by the dominance of (i) tussock grass, or purple top (Sporobolus airoides),
and rabbit brush ( Chrysothamnus grave o lens glabrata) , and (2) salt grass
(. Distichlis spicata ). The rabbit brush is also frequently associated with
greasewood (Sarcobatus vermiculatus) , especially along lines of contact
between greasewood-shadscale areas and the grass flats. For the most
part the vegetation of the grass flats is distinctly halophytic in character,
but in limited areas around springs and flowing wells it resembles that
of an ordinary nonsaline meadow.

A list of the species which were noted as composing the grass-flat
vegetation follows :

perennial species

Triglochin mariiima L.

Triglochin palustris L.

Distichlis spicata (L.) Greene
Poa nevadensis Scribner
Puccinellia airoides (Nutt.) Wats.
Spartina gracilis Trin.

Sporobolus airoides Torr.

J uncus balticus Willd.

Iris sp. (probably I. missouriensis Nutt.).

Halerpestes cymbalaria (Pursh) Greene
Dodecatheon sp.

Glaux mariiima L.

Aster pauciflorus Nutt.

Chrysothamnus graveolens glabrata (Gray)
A. Nels.

Crepis glauca (Nutt.) T. and G.

Iva axillaris Pursh

1 At Grand Junction, Colo., young seedlings of greasewood were found growing where the soil to a depth
of 2 inches, which was about the limit to which their roots had penetrated, gave a specific resistance of 36
ohms, indicating the presence of at least 2.5 per cent of salts.

2 The ecological status of the vegetation of the grass flats can not be determined until further investigations
in the Great Basin region shall have been made. For the present, therefore, it seems advisable to use the
general term “community” in referring to these types.

24395°— M - 4

406

Journal of Agricultural Research

Vol. I, No. 5

ANNUAL AND BIENNIAL SPECIES

Hordeum jubatum L.
Salicornia rubra A. Nels.
Suaeda erecta (Wats.) A. Nels.
A triplex spatiosa A. Nels.
Cleome serrulata Pursh

Melilotus alba Desv.

Erythraea arizonica (Gray) Rydb.
Orthocarpus tolmiei H. and A.
Carduus scariosus (Nutt.) Heller

Physical Conditions Indicated

Reference to Tables XIV and XV shows that there is much variation in
the moisture and salinity conditions of the grass-flat area, but, broadly
speaking, the soils are characterized by (i) a high moisture-holding
capacity, ascribable partly to the fine texture and partly to the large
quantity of organic matter present, (2) the presence near the surface
and usually throughout the summer of moisture available for growth
(above the wilting coefficient), and (3) a salt content sufficiently high
to be injurious to many crop plants. The soils under the salt-grass
community (Table XV), while usually much more saline than under the
Sporobolus-Chrysothamnus community (Table XIV), have an average
salinity inferior to that of the salt flats (Tables XVI and XVII).

Sporobolus-Chrysothamnus Community

The Sporobolus-Chrysothamnus community (PI. XT VIII, fig. 3) covers
a large part of the grass-flat area in Tooele Valley. In places one or the
other of the dominant species occurs where the other is absent, but they
are more often closely associated. Salt grass ( Distichlis spicata) is also
usually more or less abundant in this community.

Tussock grass ( Sporobolus airoides) forms coarse mats, which are as
a rule closely grazed by animals. In late summer the feathery purple
panicles of this grass are a characteristic feature of the vegetation of
the grass flats in such areas as are not grazed. The rabbit brush
(Chrysothamnus graveolens glabrata) is a much-branched shrub, from 2
to 4 feet high, with whiplike slender branches having green bark
and very small, narrow leaves. Its color is bright green, modified in
late summer by the numerous small heads of yellow flowers which
resemble those of goldenrod. The physical conditions where this com¬
munity occurs are indicated by the data given in Table XIV.

Feb. x6, 1914

Indicator Significance of Vegetation

407

Tabus XIV. — Sporobolus-Chrysothamnus community: Salt content and moisture
conditions of the soil in typical areas.1

Item.

Depth
of soil
(feet).

Date of collection.

June

July

29.

August 26.

Aver¬

age.

No. of sample .

18 SC.

ioo C.

SC.

sc.

Salt content .

f 1

1 3

I 5

I I

0. 22
.46

* 53
. 26

• 14

34-0
29. 1

34 -3
26. 4
31. 6

18. s
15-8
18.6
14-3

17. 1

-0.8
+4-4
+ 5.8
+9-8
+6. 1

0. 25

■ 45
.58
*30

o- 85
.40
. 20

. 20

o- 45
.40
. 20

o-35
. 60
• 5°

0. 20

•25

. 20

• 15

0. 20

•35

•30

0. 25
•45
• 15

0.35

•6s

1.05

0. 20

•45

•35

o*33

■44

.40

•23

Moisture equivalent. . . .

29-5
35-3
46. 6
36. 7

16. 0
19. 2
25-3
19-9

Moisture content above
or below wilting co-

+ 4*5
+ 11. 4

+25. 6
+ 14*3

1 All data in this table are stated in percentages of the dry weight of the soil. The moisture contents
with a plus sign (+) represent moisture available for growth (above the wilting coefficient), while those
with a minus sign ( — ) represent a corresponding deficit of available moisture (below the wilting coefficient).

The unnumbered borings were made on Aug. 26, 1913, and the letters indicate whether the vegetation
was dominated by Sporobolus without Chrysothamnus (S), Sporobolus with Chrysothamnus (SC), or
Chrysothamnus without Sporobolus (C).

Salt-Grass (Distichus) Community

DisiicMis spicata , well known as salt grass throughout the western
United States, is a low-growing, harsh-leaved grass which spreads by
creeping rootstocks. It tends to form a heavy sod, especially where the
land is grazed, and under such conditions this plant is very efficient in
adding humus to the soil.

Salt grass is more or less abundant in all parts of the grass flats and
also penetrates the salt flats (PI. XLVII, fig. 3), where in some places it
associates scatteringly with Allenrolfea and in other places forms dense
mats. In the wetter portions of the grass flats salt grass is the principal
component of a meadowlike vegetation, with J uncus baUicus, Suaeda
erecia , Puccinellia airoides , and Glaux maritime, as important associates
and with numerous other species occasionally present.

The conditions as regards soil moisture and salinity at borings where
this community occurs are stated in Table XV.

408

Journal of Agricultural Research

Vol. I, No. s

Table XV. — Salt-grass community: Moisture conditions and salt content of the soil in

typical areas.1

Depth
of soil
(feet).

Date of collection.

Item.

June.

July.

August.

Aver¬

age.

No. of sample .

ior

109

f 1

28. s
32. 6

28. 4
17- 1

3°* 9
22. 6

48.9
54-9
65-8
62. 2

34- 1
31.8

Moisture equivalent .

1 \

1 4

35- 1

19. 4

36. 1

15-4

13. 2

33* 3
49- 1

18. 5

f 1

15-5

17. 7

16. 5

26. 6

Wilting coefficient .

12. 3

29. 6

-ip «

i \

18. 1
26. 7

+ 9-6
+ 13* 5
+ 11. 4
+ 13-8

1.07
•97
• 7<>
.64

[ l

19. 1

io. 5

7. 2

35- e
33* 8

+ 16. 7
+ 24. 2

Moisture content above or below

+ 8.9
+ 6.9

wilting coefficient .

+ 11. 5

I* 3
“4-5

2. 30

1. 64
1-36
1. 14

+ 24. 2
+32- 2

• 59

• 72

Salt content .

j ;

•25
< 56

0. 57
.80
. 60
. 24

• 53

• 24

2. 18
1.85

1 4

I. 02

• 34
. 18

1 All data in this table are stated in percentages of the dry weight of the soil. The moisture contents with
a plus sign (+) represent moisture available for growth (above the wilting coefficient), while those with a
minus sign (— ) represent a corresponding deficit of available moisture (below the wilting coefficient).

salt-flat COMMUNITIES1
Topographical Relations

Along the margin of Great Salt Lake there is a belt of low land which
varies in width from about 4 miles, near the axis of the valley, to a mere
fringe on the east and west sides where the mountain ranges approach
the lake shore. Much of this area (see map, PI. XLII) is covered with
water at times, but in summer presents a dazzling white surface due to
the heavy crust of salts (Pi. XLIII, fig. 1 , and PI. XLVIII, fig. 1). These
flats are divided into shallow basins of greater or less extent, separated
by low ridges and hummocks. (See PI. XLII, detail of vegetation west
of Grants.) All but the lowest of these elevations are occupied by the
greasewood-shadscale association (see foreground of PI. XLIII, fig. 1),
while the basins and flats when not altogether devoid of vegetation
support a few extremely halophytic species (Pi. XLVIII, figs. 1 and 2),
which occur either as scattered individuals or in crowded colonies.

The two environments are ecologically quite distinct, but it is impracti¬
cable to indicate on a map of the small scale used in Plate XLII the areas
actually occupied by elevations and by depressions, with their respective
types of vegetation. Greasewood occurs not only on the higher ridges in
association with shadscale, but also on the lower hummocks and at the
edges of the depressions, in association with Allenrolfea. Shadscale, on
the other hand, is not found in the depressions, nor do the typical salt-
flat species occur on the higher ridges.

1 The ecological status of the salt-flat vegetation, like that of the grass-flat vegetation, can not be deter¬
mined without more extensive investigation in the Great Basin region. In the present paper it seems ad¬
visable to use the general term ' ‘community ” in referring to these types.

Feb. 1 6, 1914

Indicator Significance of Vegetation 409

The vegetation of the flats and depressions comprises several commu¬
nities, each characterized by the presence of a single species — Allenrolfea
occidentalism Salicornia utahensis , and *S. rubra. The first of these is by
far the most abundant and widely distributed. These three species
appear to be the most salt resistant of the flowering plants of this region,
taking possession of the land left bare by the recession of the lake as
soon as its salt content has been reduced sufficiently from the point of
saturation with the excessively saline lake water to permit the growth of
any flowering plant.

Fig. ii .—Allenrolfea occidentals : A, Detail of a fruiting branch, showing the cylindrical, fleshy, practically
leafless stems; Bt a plant showing the large taproot and rather scanty lateral roots characteristic of this
species.

Aixenrolfba Community

Appearance and Botanicae Composition. — The dominant species,
Allenrolfea occidentalism is a shrubby plant with numerous cylindrical,
jointed, fleshy, practically leafless branches and a large taproot (fig. 11).
In Tooele Valley it rarely exceeds a height of 2 feet. There is con¬
siderable variation in the habitat of this plant, but it develops most
characteristically on low hummocks on the salt flats (PI. XI/VIII, fig. 1)
and near the bases of the higher ridges, usually preferring a slightly

4io

Journal of Agricultural Research

Vol. I, No. s

better drained and less saline soil than the species of Salicornia. In
places, however, it is seen scattered over the surface of the flats, the

dark brownish green
tufts of Allenrolfea
contrasting strikingly
with the pure white of
the saline incrustation.
The thinness of the
stand is shown in fig¬
ure 12, which repre¬
sents a typical io-
meter quadrat. Often,
as shown in Plate
XLVIII, figure i, Al¬
lenrolfea forms a pure
community. On high¬
er and better drained
ground, however, it is
frequently associated
with Sarcobatus vermi-
culatus and with Sua -
eda moquinii} while in
the wetter depressions

Fig. i2.— A representative 10-meter quadrat of the Allenrolfea com¬
munity (salt-flat association), showing the location of each individual
plant of Allenrolfea occidentals , the only species present.

it often mingles with Salicornia utahensis . Plants of greasewood, when
growing with Allenrolfea, are usually stunted and sickly looking.

Physical Conditions Indicated. — The conditions as to soil moisture
and salt content at all borings where Allenrolfea occurred are given in
Table XVT.

Table XVI. — Allenrolfea community: Moisture conditions and salt content of the soil in

typical areas.1

Item.

Depth
of soil
(feet).

Date of collection.

June.

July.

j Aug.

29 6

Aver¬

age.

No. of sample.

Moisture equivalent.

Wilting coefficient . .

Moisture content
above or below the
wilting coefficient.

Salt content.

,98

31, 1
37-4

27.9
26. o

16. 9
20. 3
IS- 1
14. 1

+ 3-5
+ 12. 3
+ 6.9
+ 4-3
2. 18
2. 08
1. 64
x. 24

28. 4
17- 1
19.4
36. 1

IS- 4
9-3
10. 5
19. 6

+ 8. 9
+ 6.9
-1-3
- 4- S
2. 30
1. 64
1. 36
I. 16

25-5
25- 6
26. o
30. 2

13-9

13- 9

14- I
16. 4

- 2.0
- .8
- 2. 9
‘ 7-3
• 25

. 26
. 19
•44

24.4

19-3

17-3

26. 7
13- 2

13*2
10. s
9.4

14- 5
7. 2

+ 8. o
+ 8.6
+ 9-7

1.36

1.24

a- 30

1. 8S
1.36
.88
.88

i- 36
.91
. 98
.98

24. 6
28. 2
18. s
29- S
13-4
IS- 3
10. o
16. o

- 7.2

- 2. I

- 5-5

- 7*7

. 18
. 76
• 76

2. 18
1. 8S

26. 7
23-4
21.8
30*4

14- S
12. 7
11. 8
16.5

+2. 2
+6. 1
+4- 7
+3-7
1. 26
1. 11
i- 13
-94

1 All data in this table are stated in percentages of the dry weight of the soil. The moisture contents
with a plus sign (+) represent moisture available for growth (above the wilting coefficient), while those
with a minus sign (— ) represent a corresponding deficit of available moisture (below the wilting coefficient).

Feb. i6,i9i4 Indicator Significance of Vegetation 41 1

Several of these samples — e. g., Nos. 76 and 97 — were taken at places
where Allenrolfea grew in company with Sarcobatus and where the salt
content and moisture content of the soil were lower than in the typical
Allenrolfea community. It is clear, nevertheless, that the presence of
this plant is an almost invariable indicator that the soil (1) contains
moisture available for growth, at least below the surface foot, throughout
the summer; and (2) is excessively saline to a depth of at least 4 feet.

Salicornia Utahensis Community

Appearance) and Botanical Composition. — Salicornia utahensis1
(PI. XLVIII, fig. 2) is a nearly leafless plant with fleshy, jointed stems.
It resembles small plants of Allenrolfea, but is readily distinguished by
the light blue-green color and by the fact that the branches are opposite,
while in Allenrolfea they are alternate. It spreads by creeping rootstocks
and forms pure colonies of greater or less size which sometimes cover the
bottoms of depressions (see right end of PL XLITI, fig. 1), sometimes
occupy hummocks elevated but a few inches above the general surface of
the flats. In this case the appearance is much like the Allenrolfea hum¬
mocks (PL XLVIII, fig. 1), except that the latter are higher and the
plants are larger and darker colored. This Salicornia is also found in
association with Allenrolfea and with Distichlis.

Physical Conditions Indicated.— No determinations were made of
the moisture equivalent and moisture content of the soil where this
community occurs, but two borings carried to a depth of 30 inches and 12
inches, respectively, showed that abundant moisture was present through¬
out that depth, as would be expected from the slight elevation of the
land above the water surface of the lake. The salt contents of different
depths of the soil from the borings in question are given in Table XVII.

Table XVII. — Salt content of soil in the Salicornia utahensis community.

Depth of
soil.

Salt content.

Sample No. 1.

Sample No. 2.

Inches.

0 to 6

7 to 12

0 to 12

13 to 18

18 to 30

Per cent.

2. 20
>2- S°

Per cent.
>2. 50

2. 25

>2. 50

2. 20

Salicornia Rubra Community

This small, shallow-rooted annual species of Salicornia is found most
abundantly in pure communities along drainage channels in the salt flats.
The patches of Salicornia rubra are very conspicuous late in the summer,

1 This species was recently described from specimens collected by the writers in Tooele Valley by Mr.
Ivar Tidestrom. (A new Salicornia. Proc. Biol. Soc. Wash,, v. 26, p. 13, 1913.)

412

Journal of Agricultural Research

Vol. I, No. s

owing to the bright-red color then assumed by the plants. Scattered
individuals of this species were also observed far out on the otherwise
bare salt flats.

CORRELATIONS BETWEEN THE TYPES OF VEGETATION AND THE
CHARACTER AND PRODUCTIVITY OF THE LAND

CORRELATIONS WITH PHYSICAL CONDITIONS

The natural vegetation of Tooele Valley consists of a few easily recog¬
nizable plant communities, the distribution of which is largely determined
by the moisture relations and the salt content of the soil. The areas
occupied by each community are rather sharply delimited, although

Fig. 13. — Diagram showing the characteristic root development of the dominant species of each of the
principal types of vegetation of Tooele Valley, and indicating the average conditions of soil moisture and
salinity of the corresponding types of land. The double hatching indicates soil containing an excessive
quantity of salt (more than 0.5 per cent) and containing moisture available for growth (above the wilting
coefficient) during the summer. The single hatching indicates soil containing more than 0.5 per cent 0/
salts and no moisture available for growth during the summer. No hatching indicates soil containing
less than 0.5 per cent of salts and no moisture available for growth during the summer. A, Artemisia
tridentata; B, Kochia vestita; C, Atriplex conferti folia; D, Sarcobatus vermiculaius; E, Allenrolfea occiden -
talis; F, Salicornia utahensis ; G, Distichlis spicata.

along their boundaries, where the soils are of an intermediate character,
the vegetation is more or less mixed. Where, as a result of the removal
of the original vegetation by fire or by the plow, secondary plant commu¬
nities have developed, the correlations between the vegetation and the
physical properties of the underlying soils are not always well marked.
But with these exceptions, which have been sufficiently discussed on
preceding pages, all important variations in the character of the soil are
clearly expressed in the appearance and botanical composition of the
plant covering. In other words, the principal types of vegetation, where
typically developed, are reliable indicators of the physical conditions of
the environment. These correlations are stated in Table XVIII, which
follows, and are graphically represented in figure 13.

Feb. i6t 1914

Indicator Significance of Vegetation

4i3

Table XVIII. — Principal types of vegetation of Tooele Valley in relation to average soil
moisture and salinity conditions.1

Plant community.

Sagebrush.

San d-h ill
mixed.

Shadscale . —
Kochia .

Greasewood-
shadscale . . .

Grass flat .

Salt flat ,

Moisture and salinity conditions.

Source of moisture.

Direct precipitation .

.... do .

. . . .do .

.... do .

Direct precipitation
and high water
table.

Direct precipitation,
high water table,
springs and irriga¬
tion.

Direct precipitation
and high water
table.

Surface foot of soil.

Nonsaline, usually
dry in summer.

.... do .

do.

.do.

Saline or nonsaline ,
usually dry in
summer.

Moderately saline,
moist.

Saline, moist.

Soil below surface foot.

Nonsaline, usually dry
in late summer.

Nonsaline, usually (?)
moist in summer.

Saline, usually dry in
late summer.

Saline, moist.

Moderately saline,
moist.

Saline, moist.

1 The term' ‘ dry ” as here applied to the soil indicates that its water content is below the wilting coefficient.
The term ' ' moist ” implies that moisture available for plant growth (above the wilting coefficient) is present.

The average conditions as respects moisture and salinity of the soil
which characterize the land occupied by each of the more important
types of vegetation are stated in Table XIX. The data for the different
samples upon which these averages are based are given in full under the
respective associations (Tables IV, IX, X, XII, XV, and XVI). Only
typical areas of each plant community have been taken into account in
computing the averages.

Table XIX.— Moisture conditions and salt content of the soil in typical areas occupied
by the principal plant communities.1

Plant community.

Soil depth (feet).

Sagebrush
( Artemisia
iridentata).

Kochia
( Kochia
vestiia).

Shadscale
(A triplex
confertifolia).

Greasewood-
shadscale
(Sar cobatus
and Atri-
plex).

Grass flat.

Salt flat.

Salt grass
(Disiichlis
spicata).

Allenrolfea

occidentalis.

Moisture Equivalent.

14- ^

25. 8

22. 9

24. 1

34- 1

26. 7

I5- 6

27. 0

31* 0

26. 6

31.8

23-4

16. 5

33- 5

34- 5

26. 1

33*3

21. 8

4 .

15.8

31- 9

3°. 6

26. 9

49. 1

30.4

1 All data are given as percentages of the dry weight of the soil.

414

Journal of Agricultural Research

Vol, I, No. s 1

Table XIX. — Moisture conditions and salt content of the soil in typical areas occupied
by the principal plant communities — Continued.

Plant community.

Soil depth (feet).

Sagebrush

Kochia

Shadscale

Greasewood-

shadscale

Grass flat.

Salt flat.

{Artemisia

tridentata).

( Kochia
vestita).

( Atriplex
confertifolia ).

(Sarcobatus
and Atri¬
plex).

Salt grass
{Distichlis
spicata).

Allenrolfea

occtdentalis.

Wilting Coefficient.

7* 7

14. 0

12.4

13* 1

18.5

14. 5

14.7

16. 8

14.4

17-3

12. 7

3 .

8.9

18. 2

18. 7

14. 2

18. 1

11. 8

4 .

8.6

i7-3

16. 6

14. 6

26. 7

16. 5

Moisture Content above or below Wilting Coefficient.

-2. 5

-5*4

-5.6

-2.4

+ 9*6

+2. 2

- .6

-1.8

5* 2

+3-6

+I3- S

+6.1

3 .

“1-3

-2. 5

“5-0

+5-7

+11. 4

+4-7

4 .

+1.0

-2.4

-5- 5

+5- 3

+13-8

+3-7

Average. .. .

- .8

-3.0

-5-3

+3- 0

+ 12. I

+4. 2

Salt Content.

0. 03

0. 12

0. 07

0. 34

1.07

1. 26

•03

• 55

•32

.72

•97

1. 11

3 .

*°5

1. 02

.78

. 76

i- 13

4 .

.07

1. 11

•93

i- 15

.64

•94

Average. .. .

. 04

. 70

• 52 ^

.81 i

.86

1. 11

correlation s with crop production.

The capabilities of the principal types of land in Tooele Valley for crop
production with or without irrigation are summarized in Table XX.

Table XX. — Crop-producing capabilities of the land as indicated by a normal growth of

the different types of vegetation.

Type of vegetation.

Is land capable of crop production?

Without irrigation.

With irrigation.

Sagebrush .

Yes .

Yes.

Yes; if alkali can be re¬
moved.

Yes; after alkali is re¬
moved.

Possibly, with drainage.

No.

Kochia .

Precariously in years of
rainfall above the nor¬
mal.

Precariously; conditions
rather more favorable
than on Kochia land.

No .

Shadscale .

Grease wood-shadscale .

Grass flats .

Probably not .

Salt flats .

No .

Feb. 16, 1914

Indicator Significance of Vegetation

4i5

Sagebrush Land. — This is the only type in Tooele Valley which is
well adapted to dry farming. Practically all of the area devoted to
wheat in this valley was doubtless originally occupied by the sagebrush
association. Most of this area is situated on the eastern side of the
valley, where the rainfall is heavier than on the western side. But
the presence of sagebrush does not necessarily indicate good conditions
for dry farming. Where the stand is thin and the plants are small and
unthrifty, the depth of good soil is too slight for profitable crop pro¬
duction without irrigation. Sagebrush vegetation of this character indi¬
cates the presence of gravelly hardpan, or else of an excessive quantity
of alkali salts, at a depth of only 2 or 3 feet.

A good growth of sagebrush also indicates the best land for farming
under irrigation. Because of the low water table and the absence of
alkali salts, such land is not likely to require artificial drainage.

Kochi a Land. — Dry farming is sometimes attempted on Kochia land,
rye being the crop which is usually grown. The yields obtained are very
small, at least in years of only normal rainfall, the depth of good soil
being narrowly limited by the strongly saline subsoil. Whether Kochia
land is suitable for irrigation farming is somewhat doubtful, since the
rather impervious character of the soil might make it difficult to leach
the salts to a sufficient depth to insure profitable crop production.

Shadscale Land. — Dry farming is precarious on this type of land.
On the other hand, it seems probable that most of the shadscale land in
Tooele Valley would produce crops under irrigation, if water for this
purpose were available, since as compared with Kochia land the soil is
more permeable and there is greater likelihood that the salts could be
leached out of the subsoil.

Greasewood-Shadscale Land. — One or two attempts at crop pro¬
duction without irrigation on this type of land were observed, but the
results seemed to be no better than on Kochia and on shadscale land.
The reason doubtless is that while the moisture conditions are more
favorable than on the latter types the salt content of the soil at only a
slight depth is too high to permit crop plants to make a satisfactory
root development.

On the other hand, greasewood-shadscale land when irrigated and
reclaimed produces good crops of alfalfa, grain, and even of orchard
fruits. Artificial drainage, however, would probably be required in case
an extensive area of this type of land were under irrigation, the water
table being already high and the subsoil strongly saline.

Grass-Flat Land. — This type of land affords pasturage to horses and
cattle and is therefore by no means negligible as one of the agricultural
resources of the valley. Drainage would probably be necessary in order
to fit it for crop production.

Salt-Flat Land. — Most of the area occupied by this type of vegeta¬
tion is too wet and too saline for crop production and offers little prospect
of successful reclamation.

416

Journal of Agricultural Reserach

Vol. I, No. 5

SUMMARY

In Tooele Valley the different types of native vegetation indicate the
conditions of soil moisture and salinity of the land on which they are
found and thus afford a basis for estimating its capabilities for crop pro¬
duction. These correlations are stated in Table XVIII (p. 413), Table
XIX (p. 413), and Table XX (p. 414).

The sagebrush (Artemisia tridentata) association covers the land
nearest the mountains where the soil is of rather light texture, perme¬
able, rather low in moisture-holding capacity, and free from an excess
of alkali salts and where under natural conditions the moisture available
for growth is usually exhausted early in summer. A good stand and
growth of sagebrush indicates land that is well adapted to both dry
farming and irrigation farming; but where the stand is thin and the
growth poor the depth of good soil is usually too small for profitable
crop production, at least without irrigation.

The Kochia ( Kochia vestita) association covers areas lying just below
the sagebrush belt and also occupies islands in the midst of the latter
vegetation. The soil, which is remarkably homogeneous, differs from
that of sagebrush land in its finer texture, relative impermeability,
higher moisture-holding capacity, and the high salt content of the sub¬
soil. The first foot of soil is usually free from an injurious quantity of
alkali salts. Moisture available for growth is usually wanting during
the summer to a depth of at least 4 feet and probably to a much greater
depth. Dry farming is precarious on such land, owing to the small
depth of soil free from alkali. Even under irrigation the relatively
impervious nature of the soil might hinder washing out the salts to a
depth which would permit profitable crop production.

The shad scale (A triplex confertifolia ) association occupies the land
next below the Kochia belt. The soil is similar, in the main, to that
where Kochia occurs, but frequently contains much gravel, is usually
even drier during the summer months, and has on the average a some¬
what smaller salt content. Dry farming is nearly as precarious on
shadscale land as on Kochia land, but where water is available for irri¬
gation the salts could probably be leached to a greater depth than on
Kochia land, the soil being more permeable.

The greasewood-shadscale (Sarcobatus vermiculatus and Atriplex con -
fertifolia) association occupies a belt lying between the pure shadscale
vegetation and the salt flats and also crowns the ridges and knolls which
intersect the latter. The soil differs from that of any of the foregoing
associations in usually containing, during the summer, moisture available
for growth at all depths below the surface foot. It is also strongly
saline below the depth of 1 foot, and even the surface foot often con¬
tains a considerable quantity of salts. Land of this type is not suitable
for dry farming, but can be made to produce good crops under irriga¬
tion, especially when drainage is provided.

Feb. 1 6, 1914

Indicator Significance of Vegetation

417

The presence of the grass-flat (Sporobolus, Distichlis, Chrysothamnus)
vegetation indicates a soil which has a high moisture capacity, is more
or less saline, and is moist to the surface during a great part of the year.
Such land produces a coarse natural pasturage, but is not suitable for
crop production unless it is drained.

The salt-flat (Allenrolfea, Salicornia) vegetation occupies land which
is extremely saline and is wet to the surface during a great part of the
year. This type of land is not adapted to crop production.

The correlations above outlined are yet known to apply only in Tooele
Valley. Further investigation is needed in order to establish their
applicability in the classification of agricultural land in other parts of
the Great Basin.

Plate XUI. Sketch map showing the distribution and relative areas of the different
types of vegetation in Tooele Valley, with detail showing depressions
covered with salt-flat vegetation alternating with ridges bearing
grease wood-shadscale vegetation.

Indicator Significance of Vegetation

PLATE XLU

Journal of Agricultural Retearch

Plate XLIII. Fig. i. — Salt-flat vegetation bordering Great Salt Lake with a grease-
wood-shadscale ridge in the foreground, a pure stand of Salicornia
utahensis at the right and hummocks covered with Allenrolfea
occidentalis in the background.

Fig. 2 . — Sagebrush association (the darker areas) and islands of
Kochia vegetation (the lighter areas) in the upper part of Tooele
Valley. The sagebrush is encroaching upon the Kochia (at left).

Agricultural

Vegetatior

Plate XLIV

Research

Plate XLIV. Sagebrush ( Artemisia tridentata ). Fig. i. — A good stand and growth,
showing the typical appearance of this association where the
conditions are relatively favorable. Juniperus utahensis in the
background.

Fig. 2. — Plants showing the root habit; photographed at the edge of
a deep ' ‘ arroyo 5 ' where the soil had recently caved in . The exten¬
sive development of the lateral roots in the upper soil and the
penetration of the taproot to a depth of about n feet is illustrated.

24395 — *4 - 5

Plate XLV. Fig. i. — Sagebrush land which has recently been burned over,
showing scattered, dead plants of Artemisia tridentata (no living
ones), a dense growth of the annual grass Bromus tectorum, and
scattered plants (dark colored in the picture) of Gutierrezia saro-
ihrae.

Fig. 2. — An advanced stage in succession on sagebrush land which
has been under cultivation, with numerous young plants of
Artemisia tridentata and a dense herbaceous covering of Bromus
tectorum and alfilaria (Erodium cicutarium ).

Fig. 3. — Sagebrush reestablished on land which has been in cultiva¬
tion (right) and the original, undisturbed sagebrush vegetation
(left). The Stockton embankment in the background.

i&nificance

Plate XLVI. Fig. i. — Line of contact between the sagebrush association (right
hand) and the Kochia association (left hand), showing the char¬
acteristically sharp demarcation of the two types. Soil samples
collected at each side of this line, at points only 20 feet apart,
showed that in the Kochia land there was ten times as much salt
in the first foot and seventy-five times as much in the second foot
as in the sagebrush land.

Fig. 2. — A typical view of the Kochia association, with plants rather
far apart, and very uniform in size and appearance. This land
has been pastured, which has resulted in the removal of practi¬
cally all grasses and other species which occur in this association
when protected from grazing animals.

Fig. 3. — Plants of Kochia vestita , 4 or 5 inches high, and the grass
Poa sandbergii, which is usually associated wTith the Kochia in land
that is not grazed.

Plate XL VI I. Fig. i. — Typical shadscale vegetation, consisting of a nearly pure
stand of Atrip lex confertifolia, showing much dead wood, as is
usually the case, but the stand is denser than in much of the area
occupied by this association.

Fig. 2. — Transition area between the shadscale and the grease wood-
shadscale types of vegetation. Scattered (larger and darker col¬
ored) plants of greasewood ( Sarcobatus vermiculatus) in an area
occupied chiefly by shadscale.

Fig. 3. — Salt grass ( Distichlis spicata) covering the whole of the
depression to the right with the exception of a colony of Allenrolfea
in the middle distance . The higher land to the left is occupied by
greasewood (very dark in the illustration) and shadscale.

Plate XLVIII. Fig. i. — Salt-fiat vegetation, Allenrolfea community. The ground
between the hummocks is covered with a white crust of salts,
mostly sodium chlorid.

Fig. 2 . — Salt-flat vegetation, showing plants of Salicornia utahensis .

Fig. 3. — Grass-flat vegetation, Sporobolus-Chrysothamnus commu¬
nity, showing a species of rabbit brush, associated with tussock
grass.

CITROPSIS, A NEW TROPICAL AFRICAN GENUS ALLIED

TO CITRUS

By Walter T. Swingle, Physiologist in Charge , and Maude KELLERMAN, Botanical

Assistant , Crop Physiology and Breeding Investigations , Bureau of Plant Industry

INTRODUCTION

Missionaries and pioneer explorers of equatorial Africa long ago re¬
ported the finding of wild oranges and wild lemons. If the fruits were
green, they resembled small limes and lemons; if ripe, their sweet and
agreeable flavor caused them to be classed as oranges.

These fruits are from 2 to 3 cm. in diameter and are borne, two to five
or more in a cluster, in the axils of the leaves. Because of this pecul-

Fig. i. — Citropsis Schweinfurthii: A branch showing 3-foliate and 5-foliate leaves, leaflike petioles, and
rachis segments; also paired and single spines in the axils of the leaves. From a plant in greenhouse of
the Department of Agriculture grown from seed from Budongo Forest, Uganda, Africa. (C. P. B. No.
2902 . ) One-fourth natural size.

iarity they may be called African cherry oranges. The leaves are odd-
pinnate, usually with five leaflets, but often trifoliate. The petioles and
the segments of the rachis are so broadly winged that in some species
they look not unlike leaflets. (See fig. i.)

As early as 1870 Schweinfurth, the veteran African explorer, had col¬
lected leafy twigs of one of these plants, but no flowers or fruits, in the

Vol. I, No. 5
Feb. 16, 1914
0-13

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(419)

420 Journal of Agricultural Research vqh.no. 5

“Galleriewaldungen” at Uando, near the divide between the Congo and
the Bahr-el Ghazal drainage basins. In 1880 Soyaux collected specimens
of another species in Gabun (French Congo). In 1882 Pogge collected
material at Lulua in Congo proper, and in 1890 Preuss found still another
very distinct species on the shores of Elephant Lake in Kamerun. Early
in 1895 Prof. Adolph Engler described four new species of Limonia to
include these plants.1 2 In November of the same year he segregated these
African species of Limonia as a new section, Citropsis, in contradistinction
to the true Limonias of the Asiatic mainland.3

Since then several additional species have been described from tropical
Africa, and it is now clear that these plants occur not uncommonly
throughout central Africa from the Ivory Coast in the west to Uganda
in the east.

In connection with a study of the plants related to Citrus, these African
species of the Citropsis section of Limonia have been carefully examined.
The material of this section in the principal European collections of
African plants has been studied and a number of representative speci¬
mens secured, through the generosity of M. Emile de Wildeman, of
Brussels, and M. Auguste Chevalier, of Paris. Mr. B. T. Dawe, formerly
Forest Administrator of Uganda, who had discovered a new species
(Limonia ugandensis Baker) in the forests bordering the north shore of
Victoria Nyanza, sent to the Department of Agriculture at Washington
in 1910 both good herbarium specimens and viable seed.

As a result of these investigations, which have been in progress some
three years, it is now clear that these plants have been wrongly placed
in the Asiatic genus Limonia. Instead of constituting a section of this
genus, they are in reality only remotely related to the type species from
Asia (Limonia acidissima L-) and are, on the other hand, closely and
clearly related to Citrus.

The Limonia acidissima (Hesperethusa crenulata (Roxb.) Roem.) of
India has small, globose fruits only 12 mm. or less in diameter, becoming
a purple-black, bitterish berry when ripe. Each of the four cells of the
fruit contains a single seed surrounded with mucilage. There are no
pulp vesicles. The fruits are, thus, of an entirely different structure
from Citropsis and are like those of many Asiatic genera, such as Lavanga,
Triphasia, Severinia, etc., which constitute a natural group.

Besides the very important differences in the structure of the fruit,
Limonia acidissima differs from Citropsis in having free-spreading stamens
with slender filaments. None of the other Asiatic species usually referred
to Limonia are any more closely related to Citropsis than is Limonia
acidissima .

1 Engler, A, Diagnosen neuer Arten. In Notizbl. K. Bot. Gartens u. Mus. Berlin, Bd. i, No. i, p. 28-29.
Jan. 2, 1895.

2 Engler, A. Rutacese. In Engler, Adolf, and Prantl. Natiir lichen Pflanzenfamilien. T. 3, Abt. 4, p*

189-190, fig. 109, E-H. Leipzig, 1895*

Feb. 16, 1914

Citropsis , an African Genus Allied to Citrus

421

That the African species of Limonia constituting the section Citropsis
are related to Citrus rather than to the Asiatic species of Limonia is a
conclusion, based at first on a study of herbarium and living material,
that has since been confirmed in gratifying manner by the results of
experiments in grafting, which show that the African species belonging
to the section Citropsis can be budded easily and grow well on all the
commonly cultivated species of Citrus.

TECHNICAL DESCRIPTION OF CITROPSIS

It seems necessary to establish a new genus to include these African
cherry oranges. This is best done by raising to generic rank the section
Citropsis of Engler.1 2

Citropsis (Engler) Swing, and M. K.

The genus Citropsis resembles Citrus in the general structure and appearance of
the flowers and fruit, as well as in the texture, venation, and general type of the
leaves. It differs from Citrus in having 4- or rarely 5- merous ovaries, with only a
single ovule in each cell ; fruits with sessile pulp vesicles which are broad at the bases
where they are embedded in the endocarp; the stamens only twice as numerous as
the petals; large compound leaves; and spines usually occurring in pairs. The leaves
are odd pinnate, 5- or rarely 7-foliate, trifoliate, or sometimes unifoliate, subcoriaceous,
pellucid punctate. The spines are paired or single in the axils of the leaves. The
flowers occur in few- or many-flowered axillary clusters and are perfect, 4- or rarely 5-
merous. The stamens are twice as numerous as the petals, free but flattened, and
arranged to form a staminal tube surrounding the pistil much as in Citrus. The disk
subtends and is slightly larger than the base of the ovary. The ovaries are 4- rarely
5-celled with one ovule in [each cell. The style is long and deciduous; the stigma
is large, subglobose, more or less 4- rarely 5-lobed. The fruit is globular or sub-
globular, small (2 to 3 cm. in diameter), with a fleshy skin like that of a lime, dotted
with oil glands. The pulp is vesicular, either sweet and edible or waxy. The pulp
vesicles are not stalked as in Citrus, but are broad at the base where they are embedded
in the endocarpic lining of the cells and taper gradually toward the pointed tips. In
some species they are full of juice, in some they contain a waxy substance, and in
some they dry up as the fruit develops. The seeds are large, 10 by 6 by 4 mm., with
a hard, parchmentlike testa having a foramen at the tip. The cotyledons remain
hypogeous in germination. The first two foliage leaves are opposite, as in Citrus.3
(See fig. 2.)

1 Citropsis, gen. nov. (Limonia, § Citropsis, Engler). — Genus Citro affinis, foliis pinnatis, staminibus
paudoribus (staminum numero petalorum duplo nunquam quadruplo), ovariis 4- rarius 5- locularis,
loculis monospermis.

Folia imparipinnata, trifoliata vel rarius unifoliata, subcoriacea, pelluddo-punctata. Spinae in axillis
foliorum geminae vel singulae. Paniculae axillares, pauciflores. Flores hermaphroditi, 4- vel rarius 5-
meri. Stamina 8 vel 10 (numero petalorum duplo). Discus ovarii basim subtendens. Ovarium 4- vel
rarius 5- locularis, stylus longus, deciduus, stigma plus minusve quadrilobum, ovulo in loculo singulo.
Fructus globosus vel subglobosus, cortice ut in Citro carnoso, glandulis oieiferis instructo, pulpa vesiculari,
duld et eduli, vel cerea, vesiculis fusiformibus, ad basin in endocarpio immersis. Semina tnagna, testa
dura, pergamena, foraminea. Cotyledones in germinatione hypogaeae.

Arbor parva vel arbuscula, spinosa.

Species typica, Limonia Preussii Engler.

2 In Citropsis Schweinfurthii the first two postcotyledonary leaves are opposite, broadly oval, and short

stalked; the next two or three leaves are simple, with short petioles; then follow unifoliate leaves with
winged, longer petioles; then trifoliate leaves; and finally pinnately 5-foliate leaves. (See fig. 2.)

422

Journal of Agricultural Research

Vol. I, No. 5

Shrubs or small trees; native to tropical Africa.

The type species is Limonia Preussii Engler, from Kamerun.

Citropsis is related to Citrus on the one hand and to Atalantia on the
other. It differs from both in its compound leaves and broad-based
pulp vesicles partly embedded in the endocarp and from Atalantia in

Fig. 2. — Citropsis Schweinfurihii : Young seedlings germinated in
Washington, I). C., from seed from Budongo Forest, Uganda,
Africa (C. P. B. No. 2902). A, Young seedling, showing the first
pair of leaves, succeeded by alternate simple leaves, and finally
& unifoliate leaves (one-half natural size); B and D, young seed¬

lings, showing the first foliage leaves, which are opposite (natural
size); C, a single one of the pair of first foliage leaves (natural size).
Drawn by Theo. Holm.

having only a single ovule in each of the four
or five cells of the ovary. In spite of the fact

_ j that Poncirus 1 has trifoliate leaves it seems to

be less closely related to Citropsis than is Citrus.
Poncirus differs from both Citrus and Citropsis in its deciduous leaves,
sessile solitary flowers, clawed petals, spreading stamens, stalked* pulp
vesicles with external, branched, secreting hairs, and in having in ger¬
mination the first postcotyledonary leaves in the form of alternate
cataphylls.

1 Poncirus Raf. includes Citrus trifoliata L,., the type species, and as yet the only one known. See Swingle,
Walter T. Poncirus (and Citrus). In Sargent, C. S. Plantse Wilsonianae. pt. 5. Cambridge, 1914.

Feb. 16, 1914

Citropsis , an African Genus Allied to Citrus

423

CHARACTERS WHICH DISTINGUISH SPECIES OF CITROPSIS

The principal diagnostic characters of the species of Citropsis are
found in the flowers, leaves, and fruits. The size, shape, and propor¬
tions of the pistil and in particular of the style are of great importance.
The smoothness or hairiness of the filaments and the shape of the ovary
are also important characters, as is the length of the pedicel and peduncle
in relation to the length of the pistil. The shape, size, and proportions
of the leaflets, segments of the rachis, and petioles are not only obvious
but necessary characters for use in distinguishing the species. Finally,
the nature of the fruit, whether dry or pulpy, and if pulpy, whether
juicy and sweet, or waxy, is useful in distinguishing the species. Owing
to the number of species of Citropsis and the variability due to their
wide range, it is usually necessary to have at least good flowers and
leaves to be able to determine the species with any certitude, and in
some cases fruits also are necessary.

Inasmuch as none of the original descriptions of the African species
of Limonia now referred to Citropsis included both flowers and mature
fruits, it is obvious that it is a matter of much difficulty to determine the
affinities of some of these species based on imperfect material.

Citropsis Preussii (Engler), n. comb.

Limonia Preussii Engler, 1895, in Notizbl. K. Bot. Gartens u. Mus. Berlin, Bd. 1, p. 28.

Illus., Engler, 189s, in Engl, and Frantl, Pflanzenfam. , T. 3, Abt. 4, p. 189, fig. 109, E-H.

The following specimens1 have been consulted: Kamerun. — Preuss (No. 548),
September 19, 1890, Barombi Station on Elephanten See (Dahlem Herbarium 2; Kew
Herbarium). Standt (No. 747), November 29, 1896, Johann Albrechtshbhe (Dahlem
Herbarium, fragment in National Herbarium, Washington, D. C.; British Museum
Herbarium). BttSGEN (No. 37), November 18, 1905, Johann Albrechtshohe (Dahlem
Herbarium). Ledermann (No. 1455), December 1, 1908, Bare (Dahlem Herbarium;
fragment in National Herbarium, Washington, D. C.). 3

The type of the genus, Citropsis Preussii, was first collected by Preuss at
Barombi Station on the south shore of Elephanten See in Kamerun on
September 19, 1890. Of his original collection (No. 548) three specimens,
all showing good flowers, have been studied by the writers. (See fig. 3.)
Two of these are preserved in the herbarium at Dahlem, near Berlin.
The third was sent to Kew Gardens (April, 1894) before the species was
published and evidently was not used by Prof. Engler in drawing up the
original description, as the species is described as having trifoliate leaves,
while those of the Kew specimen are 5-foliate.

Besides this original material there are three excellent sheets in the
Dahlem Herbarium and one at South Kensington of material collected
by Standt (No. 747) on November 29, 1896, at Johann Albrechtshohe,

1 All of the specimens cited from European herbaria were photographed by one of the writers in 1911-12,
and prints enlarged to natural size have been filed in the National Herbarium at Washington, D. C.

2 The sheet to which the original label is attached is the type specimen.

3 Eedermann’s specimens have been designated "Limonia Preussii Engl., var. micrantha Engl./’ but it
is probable that the very small flowers are due to a diseased condition of the plant and do not constitute a
true varietal difference.

424

Journal of Agricultural Research

Vol. I, No. 5

Fig. 3. — Citropsis Preussii: Flowers after petals and stamens have fallen; leaves, one trifoliate and one
having the terminal leaflet borne on a winged segment of the rachis. From paratype, Standt No. 548,
in Dahlem Herbarium. One-half natural size.

Feb. 16, 1914

Citropsis, an African Genus Allied to Citrus

425

near the original type locality on Elephanten See. These specimens show
flowers and young fruits. Finally, there is one sheet in the Dahlem
Herbarium, collected by Biisgen (No. 37) on November 18, 1905, also at
Johann Albrechtshohe near Elephanten See. This specimen shows young
fruits.

All of this material comes from the same general locality, Johann
Albrechtshohe being only 3 or 4 km. distant from Barombi Station. All
eight of these specimens show a great resemblance and undoubtedly
belong to a single species. Unfortunately all were collected in the autumn
and show only flowers and very young fruits.

A number of other specimens have been referred to Citropsis Preussii
in the Dahlem Herbarium, but some of them certainly do not belong here,
and for the present the only material certainly referable to this species is
that collected in the immediate vicinity of Elephanten See in Kamerun.

The excellent specimens with flowers and
young fruit and numerous leaves permit a
very good idea to be gained of this species.

The leaves are 3- to 5-foliate, with broadly
winged petiole and rachis. (See fig. 3.)

The leaflets are very large, 100 to 160 by
45 to 1 15 mm., broadly oval or oblong,
abruptly narrowed above into a short ob¬
tuse tip, and broadly cuneate at the base,
with very short petiolules. Petioles usually
69 to 80 by 25 to 35 mm., elongate, ellipti¬
cal, rather acute at tip and base, but some¬
times shorter and broader or even obcordate
30 to 40 by 25 mm. The segments of the
rachis are elongate elliptical, 50 to 70 by
15 to 25 mm. Spines usually single, 16 to 28 mm. long, rarely wanting.
Flowers 15 to 18 mm. long in the bud, 20 to 25 mm. in diameter when
open, in dense many-flowered clusters borne in the axils of the leaves,
very short pediceled (3 to 5 mm.), usually 4-merous, ovaries 12 to 15 mm.
long, with a long, slender style broadening at the base and merging
gradually into the ovary. Only young fruits are known as yet. These
are short-stalked or nearly sessile, slightly apiculate.

Citropsis Preussii is readily distinguished from its congeners by the
broadly oval or oblong leaflets, and by the short-stalked flowers with
very long styles broadened at the base and not sharply delimited from
the tip of the pointed ovaries. Citropsis mirabilis resembles this species
in the shape of the leaves, winged petioles, and rachis, but differs in the
longer stalked flowers, which have a shorter more slender style which is
not broadened at the base and consequently is more sharply delimited
from the tip of the more rounded ovary. (See fig. 4.)

Fig. 4. — Pistils of four species of Citrop¬
sis. Ay Citropsis Preussii (Standt
No. 747); Bt Citropsis m irabilis (Cheva¬
lier No. 21609); Ct Citropsis Schwein-
furthii (C. P. B. 2902); and D , Citrop¬
sis gabonmsis (Klaine No. 2260).
Twice natural size.

426

Journal 0} Agricultural Research

Vol. I, No. 5

Citropsis Schweinfurthii (Engler), n. comb.

Limonia Schweinfurlkii JUngler, 1895, in Notizbl. K. Bot. Gartens u. Mus. Berlin, Bd. 1, p. 29.

(?) Limonia ugandensis Baker, 1907, in Jour. Bot. [London], v. 45, p. 61.

(?) Limonia Poggei Engler, 1895, in Notizbl. K. Bot. Gartens u. Mus. Berlin, Bd. 1, p. 29.

The following specimens have been consulted: Sudan. — SchwEinFurth (No. 3656)
April 25, 1870, Uando (Dahlem Herbarium,1 clastotype in National Herbarium,
Washington, D. C., see fig. 7; Kew Herbarium, clastotype). Stuhlmann (No. 2641),
August 24, 1891, Ituri Ferry (Dahlem Herbarium). Uganda. — Bagshawe(No. 1007), 2
April 25, 1906, Mpanga Forest, Toro (British Museum Herbarium); (No. 1365), Decem¬
ber 17, 1906, Ngusi River, Albert Edward Myanza, altitude 950 meters (British Museum
Herbarium; National Herbarium, Washington, D. C.). DawE (No. 399), South
Buddu (Kew Herbarium); (No. 809), 1905, Budongo Forest (Kew Herbarium);
(No. ?) March 17, 1910, Budongo Forest (National Herbarium, Washington, D. 0.);
(No. ?, C. P. B. No. 2902) April 17, 1910, Budongo Forest (National Herbarium;
greenhouses, Department of Agriculture, Washington, D. C. See fig. 1 and PI.
XLIX). Milbraed (No. 2394) January 1, 1908, Fort Beni (Dahlem Herbarium);
(No. 2880), May 1, 1908, Irumu (Dahlem Herbarium). Congo. — Pogge (No. 668), 3
June 1, 1882, Lulua (Dahlem Herbarium). Laurent (No. ?), November 24, 1903,
Ibaka (Brussels Herbarium); (No. ?), January 2, 1904, Bolombo (Brussels Herbarium;
National Herbarium, Washington, D, C.); (?) French Congo. — Thollon (No. 1049),
June, 1888, on Niari River from Komba to Bounanza (Museum, Paris, Herbarium).

In 1895 Engler published Citropsis Schweinfurthii, which was based on
a single unbranched twig without flowers or fruit collected by Schwein-
furth (No. 3656) in April, 1870, in the “ Galleriewaldungen ” at Uando
(altitude 700 to 800 m. ; lat. 40 18' N., long. 28° 22' E.), about 260 km.
northeast of Albert Nyanza. The twig was originally some 33 cm. long,
with 12 intemodes. The basal intemode, with a trifoliate leaf, was sent
to Kew Herbarium in February, 1878, where it is now preserved. The
rest of the specimen is in Prof. Schweinfurth’s herbarium in the Dahlem
Museum and is the type upon which Prof. Engler based the species.
In the original description of the species the leaves are said to be trifo¬
liate, but in this specimen one of them, the fifth from the tip of the twig,
is pinnately 5 -foliate with a well-developed, broadly winged rachis
between the first and second pair of leaflets. One of the lateral leaflets
of the terminal pair is missing, but the shape and position of the terminal
leaflet show clearly that it was present during the life of the plant and
was probably lost after the specimen was dried, as has happened to seven
or eight leaflets belonging to other leaves of this same specimen.

The discovery of this pinnate leaf on the type specimen is of impor¬
tance in justifying the reference to this species of a number of pinnate¬
leaved specimens from the eastern part of equatorial Africa.

A fruiting specimen was collected at a ferry of the Ituri River about
60 km. WNW. of Albert Nyanza in latitude 20 55' N. (altitude 900 meters)
by Dr. F. Stuhlmann (No. 2641) on August 24, 1891, in his journey
around the great lakes of equatorial Africa. Stuhlmann mistook the

1 This is the type specimen.

2 Type specimen of Lhnonia ugandensis.

* This is the type specimen of Limonia Poggei.

Feb. 16, 1914

Citropsis , an African Genus Allied to Citrus

427

broadly winged segments of the rachis for leaflets sprouted out of each
other.1 His specimen is preserved in the Dahlem Herbarium and has
been referred to Limonia Schweinfurthii by Engler.2

The original label has a note by Stuhlmann to the effect that the fruit
is orangelike, light yellow in color, shows two seeds, and has a sweet pulp
without acid. A sketch on the label shows a 4-celled fruit with two seeds.
Most of the leaves are pinna tely 5 -foliate, though the specimen is in bad
condition and many leaflets have been lost. Both the leaves and spines
are much like those of Schweinfurth’s original specimen from Uando,
and it is very probable that both belong to the same species.

Misled by the statement in the original description of this species that
the leaves are trifoliate, Baker described a new species, Limonia ugan -
densis, in 1907, which he says differs from Limonia Schweinfurthii (known
to him only from the description) in having 5-foliate instead of 3-foliate
leaves.

The, type of Limonia ugandensis was collected by Mr. A. G. Bagshawe
(No. 1007) on April 25, 1906, at Toro, in the Mpanga Forest, to the
east of Albert Nyanza, in western
Uganda, at an altitude of 1,550
meters. The type specimen shows
flower buds and has single spines and
mostly 5-foliate leaves, but appar¬
ently a few 3-foliate leaves also.

The petioles and segments of the FIG. 5. — Citropsis Schweinfurthii: Nearly mature

rachis are broadly winged and vary fruit; a, side view, showing calyx and disk; s,

, . section showing four cells with pulp vesicles

from narrowly elliptical to obovate in and three seeds. Bagshawe No. 1365, in

outline. Because of the absence of National Herbarium, Washington, D. c.

_ Natural size.

mature flowers the description of the

stamens is erroneous in giving the filament as about equaling the anther
in length. In a fully open flower the filaments would undoubtedly be
much longer. A specimen of this species which was collected by Mr.
A. G. Bagshawe (No. 1365) at Ngusi River, Albert Edward Nyanza, at
an altitude of 970 meters, shows good fruits (see fig. 5).

Aside from the usually but not universally broader winged petiole and
rachis segments these specimens can scarcely be distinguished from
Citropsis Schweinfurthii , and unless the flower and fruit characters prove
to be different, Limonia ugandensis will doubtless have to be considered
to be a synonym of C. Schweinfurthii.

Besides the specimens from Uganda hitherto referred to Limonia
ugandensis Baker, there are two specimens in the Dahlem Herbarium,

1 “In dem dichten Untesholz fiel uns vor allem ein kleiner Busch mit dornigen Aesten auf. Von seinen
lederharten Blattern spriesst eines aus dem anderen heraus. Seine Frucht ist eine kleine Orange mit
mehreren Abtheilungen, aber nur zwei Kernen. Von unseren Eimonen unterscheiden sie sich durch den
susslichen, jeder Saure entbehrenden Geschmack. ” (Stuhlmann, Franz. Mit Emin Pascha ins Herz von
Afrika. p. 406. Berlin, 1894.)

2 Engler, Adolf. Die Pflanzenwelt Ost-Afrikas. . . . Teil C, p. 229. Berlin, 1895.

428

Journal of Agricultural Research

Vol. I, No. 5

collected by J. Milbraed in 1908, which seem to be referable to Ciiropsis
Schweinfurihii. One specimen (No. 2394) is from Fort Beni, in extreme
western Uganda, on the Semliki River, about half way between Albert
Nyanza and Albert Edward Nyanza. This specimen consists of a single
twig with 5-foliate leaves, single spines, and two young fruits. The other
(No. 2280) is from Kikufu, near Irumu, in the Ituri River valley, only a few
kilometers south of the ferry where Stuhlmann crossed the Ituri and
collected his No. 2641. This second specimen of Milbraed consists of
two twigs with mostly 5-foliate leaves, but one of them has a trifoliate
leaf almost exactly like those of Schweinfurth’s original specimen from
Uando.

Limonia Poggei Engler, which the writers have referred doubtfully to
Ciiropsis Schweinfurihii , was based on a single specimen collected by Pogge
(No. 668) June 1, 1882, at Lulua, latitude 6° S., on the Lulua River, an
affluent of the Kasai River. The type specimen preserved in the Dahlem
Herbarium shows a single twig with 11 or 12 intemodes, but with only
one 5-foliate leaf remaining attached. There is also one loose leaf and
a single fruit. Pogge's original label notes that the fruit is yellow. An
examination of the fruit preserved with the type specimen at Dahlem
shows it to possess distinct pulp vesicles. There is nothing in the
specimen or in the description to distinguish it from Ciiropsis Schwein¬
furihii , and as it occurs at a considerable altitude, 660 meters, and only
500 km. west from the nearest of the great African lakes, while Uando,
the type locality, was some 250 km. west, its geographic range is not
such as to render its inclusion in the species improbable.

It is interesting to note that all the reported localities of Ciiropsis
Schweinfurihii are above 660 meters altitude, the highest reported being
1,550 meters at Toro, Mpanga Forest, Uganda.

There is, however, a specimen in the herbarium of the Museum
d’Histoire Naturelle at Paris, collected by Thollon (No. 1049) in June,
1 888, in French Congo on the Niari River between Bounanza and
Komba, that can scarcely be distinguished by its leaf characters from
Ciiropsis Schweinfurihii. Bounanza is only 250 km. from the Atlantic
Ocean and at an altitude of only 130 meters. Thollon states on his
original label that this plant occurs in all the woods from Komba to
Bounanza. If this material proves to be Ciiropsis Schweinfurihii , it will
give this species the greatest range both in distance and in altitude of
any yet known in the genus Citropsis.

There is a specimen 1 in the herbarium of the botanic garden at
Brussels, collected by Messrs. Em. and M. Laurent below Ibaka, on the
Sankuru River, Congo, on November 24, 1903, and also a specimen in
the National Herbarium at Washington, D. C., collected by Messrs.
Laurent below Bolombo, on the Sankuru River, on January 2, 1904.

1 This specimen seems to have been referred to Limonia Demeusei by M. ftmile de Wildeman. See his
Mission £mil Laurent (1903-4). v. 1, p. 238. Brussels, 1905-1907.

Feb. 16, 1914

Citropsis , an African Genus Allied to Citrus

429

Both of these specimens, as well as one in the herbarium at Brussels,
collected in Congo by Messrs. Laurent in 1903-4, but without exact
locality or date, have trifoliate leaves, long, slender leaflets, with the
terminal one disproportionately long, being no to 165 by 30 to 45, while
the adjacent lateral leaflets are 65 to 90 by 25 to 45; thus the terminal
leaflet is from two-fifths to one-third longer than the lateral. The only
species to which these specimens can be referred at present is Citropsis
Schweinfurthii , but in the absence of flowers and fruits and because of
the rather unusual appearance of the leaves such reference must be
merely provisional.

Citropsis Schweinfurthii is a spiny shrub or small tree with 3- to 5-
foliate leaves. The flowers are borne in clusters of 4 to 10 in the axils
of the leaves. (See fig. 6.)

They are 4- or rarely 5-
parted with strap-shaped
petals, a short, thick style,

6 to 9 mm. long, scarcely
narrower than the stigma
but rather sharply set off
from the rounded tip of the
ovary, and broad flattened
filaments with a subulate
apex where the anther is
attached. The leaves are
pinnately 5 -foliate or tri¬
foliate. The petioles are
broadly winged, 40 to 75
by 18 to 35 mm., narrowly
obovate or elliptical, usu¬
ally rounded at the tip
and bluntly pointed at ,the
base. The segments of the
rachis are 35 to 65 by 15 to 25 mm., usually elliptical, bluntly pointed at
both ends but more rounded (sometimes rather broadly rounded) at the
tip. The leaflets, 55 to 125 by 15 to 50 mm., are broadly lanceolate,
narrowed from the middle to the long, cuneate base and into an acute
or subacute tip, with strongly marked serrations. (See fig. 7.) The ter¬
minal leaflet is often much larger than the adjacent lateral leaflets, some¬
times one-third longer, usually from one-fourth to one-eighth longer. The
spines, 12 to 30 mm. long, are usually paired in the axils of the leaves.

Citropsis Schweinfurthii differs from all its congeners in having a short,
thick style (shorter than any other species except C. gabunensisy which
has very small flowers, with a slender style) and slender, broadly
lanceolate leaflets, narrowing from the middle into a long, cuneate
acute base.

Fig. 6 —Citropsis Schweinfurthii: Cluster of flowers, showing
stamens arranged to form a staminal tube. From a plant
growing in greenhouse of the Department of Agriculture,
grown from seed from Budongo Forest, Uganda, Africa.
(C. P. B. No. 2902.) Natural size.

24395 — 14-

-6

430

Journal of Agricultural Research

Vol. I, No. $

Citropsis gabunensis (Engler), n. comb.

Ltnumia gabunensis Kngler, 1895, Notizbl. K. Bot. Gartens u. Mus. Berlin, Bd. 1, p. 28.

(?) Limonia Lacourtiana De Wild., 1904, in Ann. Mus. du Congo, Bot. s. 5, v. 1, p. 159-160, pi. 50.
Illus., De Wild., op. cit., pi. 50.

The following specimens have been consulted: French Congo (Gabtin). — Soyaux
(No. 105), July 25, 1880, Sibanga Farm, Munda. (Dahlem Herbarium,1 Kew Herba¬
rium; Museum, Paris, Herbarium). Kxaine; (No. 2260), July and October, 1901,

Fig. 7. — Citropsis Sckwcinfurtkii: A trifoliate leaf from the type specimen, showing double spines in the
axils and pronounced serrations of the leaflets toward the tips (Schweinfurth No. 3656); in National
Herbarium, Washington, D. C. Natural size.

near Libreville (Kew Herbarium; Dahlem Herbarium; Museum, Paris, Herbarium);
(No. 1973), March 10, 1901, May and October, 1902, Libreville (Museum, Paris,
Herbarium); (No. 2924, 2925), June, 1902, Libreville (Museum, Paris, Herbarium);
(No. 3494), May 25, 1904, Libreville (Museum, Paris, Herbarium). ButtnFR (No. 432),

1 The specimen with the original label attached is the type.

Feb. 16, 1914

Citropsis, an African Genus Allied to Citrus

43i

September, 1884, Sibange Farm (Dahlem Herbarium). Du Beu,ay (No. 4?), 1864
(Museum, Paris, Herbarium). Tessmann (No. 874), January 26, 1909. Spanish
Guinea (?)— Bebady (?); (No. 194), February 14, 1908, Nkolentagan. (?) Congo. —
GenTie (No. 93)/ May, 1903, Bombaie (Brussels Herbarium). Hendrickx (coll.
Gillet, No. 3280) Lumene (Brussels Herbarium). Daurejnt (No. ?), November 28,
1903, Bombaie (Brussels Herbarium).

Citropsis gabunensis , one of the first four species of Limonia, described
from Africa by Engler in 1895, was based on specimens collected by
H. Soyaux (No. 105) at Sibanga Farm in the Munda region near Libre¬
ville, French Congo (Gabun), on July 25, 1880. Three sheets of this
number are preserved in the Dahlem Herbarium, and on them the
species is based. The type specimen had a single fruit ; the paratypes
are sterile. The herbaria of Dahlem, Brussels, Paris, and Kew contain
numerous other specimens of this species from northern French Congo
and Spanish Guinea. This material represents a wide range of foliar
characters and shows all stages of flower and fruit development. All
these specimens seem to belong to a single species which is very distinct
from any of the others.

The type specimen of Limonia Lacourtiana was collected by L. Gentil
(No. 93), May, 1893, and is preserved along with Gentil’s original label
in the herbarium of the botanic garden at Brussels. The leaves are all
5-foliate, and in one case a terminal leaflet has a winged petiole. The
leaflets are broadly oval, more or less abruptly narrowed at the base,
and caudate at the tip. The young fruits are borne in clusters in the
axils of the leaves on pedicels 10 to 12 mm. long. In all of these char¬
acters this specimen is indistinguishable from Citropsis gabunensis .

A young fruit from this type specimen now preserved in the National
Herbarium at Washington, D. C., seems to be seedless, but shows
numerous pulp vesicles which contain a whitish granular wax.2 The
original label of M. Gentil says “ fruits delicieux,” but as the fruits in
the type specimen are very small and immature it is obvious that his
statement must apply to some other plant, doubtless not belonging to
this species. Most of the fruits of the typical Citropsis gabunensis
examined contain large seeds, often nearly filling the small fruit and
leaving very little space for the pulp vesicles, which are crowded and
often nearly obliterated by the seeds.

Whether the vesicles of a young seedless fruit of the typical Citropsis
gabunensis would show the presence of wax remains to be investigated.
In the absence of knowledge on this point it seems inadvisable to recog¬
nize Limonia Lacourtiana as a species distinct from Citropsis gabunensis ,
though future research may possibly prove it to be a good species.

1 This is the type specimen of Limonia Lacourtiana.

1 Recently, through the kindness of M. Auguste Chevalier and of Rev. J. Gillet, of Kisantu, Congo,
abundant material has been received of a species of Citropsis apparently distinct from any hitherto described,
the fruits of which are often seedless and contain abundant pulp vesicles filled with a wax, which, when
extracted, makes a yellow, fragrant mass much like beeswax in character.

43 2

Journal of Agricultural Research

Vol. I, No. s

Ciiropsis gabunensis differs from all its congeners in having very small
flowers, with hairy filaments, caudate leaflets, and a nearly dry fruit.
The flower buds are only 5 to 6 mm. long and the fully expanded flowers
are only 10 to 12 mm. in diameter. The filaments are hairy. The pistil
is very short (3^ to 4 mm.) and shows a well-marked, clavate ovary,
narrowed gradually toward the base and rounded at the tip, which is
clearly delimited from the slender style which ends in the subglobose

4- lobed stigma. (See fig. 4.) The pedicels are very long (sometimes
8 to 12 mm.), often twice as long as the pistil, and appear as branches
of a slender peduncle % to 2 cm. long.

No other species of Citropsis shows so much variation in the size of
the leaves and in the number of leaflets. They may be unifoliate,
greatly resembling orange leaves, or they may have 5 to 7 leaflets.
Very frequently the leaves are 5-foliate, with the terminal leaflet borne
at the end of a winged segment of the rachis. Such stalked terminal
leaflets are often seen in trifoliate leaves (see fig. 3) but almost never in

5- foliate leaves of other species of Citropsis. The leaflets are caudate —
unlike any of the other species.

Those of compound leaves are from 40 to 115 by 18 to 60 mm.,
mostly 50 to 100 by 25 to 45. The leaflets of unifoliate leaves are 90 to
150 by 40 to 70 mm. The winged petioles are 15 to 35 by 3 to 15 mm.,
varying from linear to narrowly obcordate, especially in unifoliate
leaves. They are usually broadly rounded at the tip and narrowed
gradually toward the base. The rachis segments vary from 20 to 45
by 4 to 10 mm. and usually have the same shape as the winged petioles.
Citropsis mirabilis (Chev.), n. comb.

Limonia mirabilis Chevalier, 1912, in Bui. Soc. Bot. France, t. 58, 1911, Mem. 8d, p. 144-145.

The following material has been consulted: Ivory Coast. — Chevalier (No. 21609),
May 21, 1909, between Sanrou and Quode on the Koue River (Chevalier Herbarium,
Paris; National Herbarium, Washington, D. C.).

Chevalier has described Ciiropsis mirabilis in detail, but unfortu-
nately no fruits are known. The leaves are 3- to 5-foliate, with broadly
oval or oblong leaflets 90 to 190 by 40 to 100 mm. The petioles are
usually elongate elliptical, 60 to 70 by 20 to 30 mm., rather acute at
both ends, rarely broadly rounded at the tip. The segments of the
rachis are 80 to 70 by 12 to 28 mm., usually narrowly elliptical, rarely
broadly rounded at tip. The spines are single, 10 to 28 mm. long,
sometimes wanting. The flowers occur in dense many-flowered clusters
in the axils of the leaves. The pedicels are well-developed, 5 to 6 mm.
long. The buds are linear elliptical, 12 to 14 by 3 mm., the flowers
when open are 18 to 24 mm. in diameter, usually 4-merous, but some¬
times 5-merous. The pistil is 12 to 14 mm. long, the style 10 to n mm
long, very slender, and not appreciably broadened at the base.

Feb. 16, 1914

CitropstSy an African Genus Allied to Citrus

433

Citropsis mirabUis differs from all its congeners in having large flowers
with slender styles not much broadened at the base and, in consequence,
rather clearly delimited from the tip of the ovary. (See fig. 4.) It
somewhat resembles C. Preussii in the size and shape of the leaves.

imperfectly known species
Citropsis articulata (Willd.), n. comb.

Citrus articulata Willd., 1826, in Spreng., Syst. Veg., v. 3, p. 334-

The following material has been consulted: Gold Coast. — Isert (No. ?; Willde-
now Herbarium No. 14357)* June or July, 1786, near Kommang, Akwapim 1 (Dahlem
Herbarium). (?) Togo. — Baumann (No. 552), 1894-5, on the Koli River near Kame
(Dahlem Herbarium).

The specimen in the Willdenow Herbarium at Dahlem, of which a
photograph was kindly sent to the writers by Prof. Urban, now shows a
single twig, 21 cm. long and to 4 K mm. in diameter, with 10 or 11
intemodes which are mostly 2 to 2^ cm. long, slightly angular, with
prominent leaf scars. Only two single spines are preserved, one 8 by 1
mm., the other 14 by 1 % mm. Two petioles are on the sheet: One, still
attached, obovate in outline, 52 by 32 mm. tapering gradually into the
sharp base 4 to 5 mm. long; the other broadly rounded at tip, 60 by 37 mm.
with prominent veins, running nearly at right angles to the midrib, the mar¬
gin very shallowly undulate crenate. It is evident that the Isert specimen
at Dahlem was more complete when Sprengel published Willdenow’s
description, as the leaves are said to be oblong and the peduncle many-
flowered. Probably only a single terminal leaflet was originally present.
The many-flowered peduncle seems also to have fallen off since Willde-
now's time, as none can now be seen on the photograph.

To this species has been doubtfully referred a specimen from Togo-
land collected by E. Baumann (No. 552), on May 16, 1895, on the Koli
River near Kame, probably not very remote from the locality where
Isert's type was collected. The Baumann specimen has 3- to 5-foliate
leaves, with petioles varying somewhat in size and shape, 35 by 13,
60 by 25, 32 by 50, 40 by 18, or 30 by 20 mm. Curiously enough, the
terminal portion of the twig, including the last five or six internodes,
has lost its leaves except one obovate petiole. It has three single spines
and in general resembles in a striking manner Isert’s specimen, upon
which Willdenow based his species. Another curious coincidence is in
the presence of the terminal leaflet of an originally trifoliate leaf from
which the two lateral leaflets have fallen. It was probably from such
an apparently unifoliate leaf originally present on Isert’s specimen that

1 Isert found this plant in the mountains some so to 75 km. north of Accra and says of it: 41 Je vis une
nouvelle esp^ce de citroniers, avec des feuilles articulees.” (Isert, P. E. Voyages en Guinee et dans les
lies Caraibes en Amdrique, p. 255-256. Paris, 1793- A reprint of the original edition, Reise nach
Guinea . . . 1788, appears in Allgemeine Geschichte der neuesten Reisen und Entdeckungen, v. 1.)

434

Journal of Agricultural Research

Vol. I, No. 5

Willdenow described the leaves as oblong. This Baumann specimen
shows an axillary inflorescence comprising some 6 to 8 flowers with
slender ovaries (io to n mm.) and very slender styles somewhat like
Ciiropsis mirabilis. The leaves of the Baumann specimen have more
broadly winged petioles than the C. mirabilis , and doubtless because of
this it was referred in the Dahlem collection to C. Preussii , from which it
differs in the distinctly shorter, more slender style, the narrow smaller
leaflets, and the broadly rounded tips of the winged petioles and seg¬
ments of the rachis. The flowers in the Baumann specimen are more
densely clustered and shorter pediceled than in C. mirabilis .

It is to be hoped that more complete material collected by Isert may
be found in the Copenhagen Herbarium which will permit the affinities
of this species, the first of the group to be discovered, to be determined
with exactitude.

Besides the foregoing, there remain two more African species of
Limonia which undoubtedly belong to Citropsis, but which can not as
yet be satisfactorily placed because of insufficient material. These are
Limonia Poggei , var. laiialata De Wild., doubtless distinct from L. Poggeiy
and Limonia Demeusei De Wild. Both have been described and beauti¬
fully figured.1

In addition to the material cited, specimens are to be found in the
various European herbaria and in the National Herbarium at Washing¬
ton, D. C., which it has been impossible to place, owing to the lack of
flowers or fruits. This additional material represents collections, prin¬
cipally from Congo, by Auguste Chevalier, 13m. and M. Laurent, Demeuse,
L. Gentil, and others.

POSSIBLE USES OF THE AFRICAN CHERRY ORANGES

The bringing to light of a new genus belonging to the true-orange group
opens up a new field for the plant breeder, especially as some of the
species are said to bear delicious fruits in abundance.

The unusually large compound leaves — often with five leaflets, each
one of them larger than any ordinary orange leaf — give several of the
species of Citropsis a distinct advantage over any other member of the
true orange group. Large leaves are an outward and visible sign of an
active assimilating system, and it must not be forgotten that over three-
fourths of the dry substance of a plant is made up of starch, sugar, oil,
flavoring matter, and other substances manufactured in the leaves, and a
species with large leaves is equipped with the first essential for rapid
growth and for developing sweet fruits of high flavor.

1 Wildeman, Lmile de. Etudes sur la Flore du Bas- et du Moyen-Congo. In Ann. Mus. du Congo, Bot.
s. 5, v. i, p. 159-160, pi. si, 53* 1904.

Bimonia Poggei, var. latialata. In Card. Chron., s. 3, v. $3, no. 1380, p. 378, fig. 159, June 7, 1913.

Feb. 16, 1914

Citropsis, an African Genus Allied to Citrus

435

GRAFTING OF CITROPSIS

Experiments conducted under the directions of the authors in the
greenhouses of the Department of Agriculture at Washington, D. C., show
that Citropsis Schweinfurthii can be grafted readily and that it will grow
rapidly and vigorously on sweet orange, sour orange, grapefruit, and
lemon stocks. It can also be grafted on the tabog ( Chaetospermum gluti-
nosa) and the wood-apple (Feronia elephantum) , two stocks on which
species of Citrus graft readily. However, it does not grow as vigorously
on these stocks as on Citrus. The very rapid growth of Citropsis when
grafted on Citrus (see PI. XUX) is an added and striking proof of the close
affinity of these two genera. Additional experiments in budding and
grafting on other genera related to Citrus are now under way.

In view of the considerable botanical differences between Citrus and
Citropsis, it is probable that the latter will show immunity to diseases and
adaptations to soil and climatic conditions not. possessed by the stocks
upon which citrous fruits are commonly grafted. Experiments conducted
by the authors have already indicated that Citropsis Schweinfurthii is well
adapted to poor, sandy soils (“high pine lands” ) in Florida. Every new
stock well adapted to Citrus gives the grower and the pathologist a new
tool in the work of perfecting the culture of citrous fruits and in prevent¬
ing the ravages of diseases by using stocks which are immune. The
scarcity of material of the African cherry oranges has hitherto prevented
any extensive experiments in the use of this new stock, but grapefruit
and oranges have both been budded successfully on Citropsis stocks in
the greenhouse at Washington and out of doors in Florida.

HYBRIDIZATION OF CITROPSIS

The fact that there are a number of closely allied yet distinct species
of Citropsis native to the forests of tropical Africa is an advantage to the
plant breeder in furnishing material for the improvement of the African
cherry oranges by hybridization. Whether the waxy-fruited species will
yield edible hybrids when crossed with the juicy-fruited species can only
be told by experiment.

So far, the scarcity of flowers of the African cherry oranges has pre¬
vented any decisive test as to whether they can be crossed with species
of Citrus or not. This much can be said, that flowers of the common
lime, Citrus aur antifolia (Christm.) Swing., pollinated with Citropsis
Schweinfurthii set fruit and produced seed. Only a few seed were se¬
cured and none of them gave rise to a hybrid, but this is not uncommon
in Citrus. The fact that the pollen of Citropsis was able to cause the
development of seeds is a very hopeful sign that hybrids will be secured
from pollinations in the course of the breeding experiments now being
carried on by using the pollen of Citropsis on as many species of Citrus
as possible.

436

Journal of Agricultural Research

Vol. I, No. s

That hybrids of the common citrous fruits with the African cherry
oranges would be promising table fruits is rendered probable by the fact
that both Citrus and Citropsis have species which in a wild state yield
fruits beautiful to the eye, fragrant, and delicious to the taste.

Because of their beautiful foliage, their very fragrant, large white
flowers, much resembling those of the orange or lime, and their abundant,
though small, fruits, borne in tufts like cherries, the African cherry
oranges are of unusual promise for ornamentals and for hedge plants in
subtropical regions.

The fact that the true relationships of so large and so striking a group
of plants, ranging clear across equatorial Africa, could remain misunder¬
stood by botanists for so long a time, is another proof of the rich harvest
of new material which awaits the attention of the plant breeder as soon
as a critical taxonomic study of the wild relatives of our principal cul¬
tivated plants makes it available for his use.

Plate XUX. Citropsis Schweinfurthii grafted on grapefruit stock ( Citrus decumana ),
showing vigorous growth made in 2% years. Plant grown in green¬
house, Department of Agriculture, Washington, D. C., from seed
from Budongo Forest, Uganda, Africa. (C. P. B. No. 2902.) One-
sixth natural size.

Plate XLL

PRELIMINARY AND MINOR PAPERS

WINTER SPRAYING WITH SOLUTIONS OF NITRATE

OF SODA1

By W. S. Ballard, Pathologist , Fruit-Disease Investigations , Bureau of Plant Indus¬
try , and W. H. Volck, County Horticultural Commissioner of Santa Cruz County ,

California .

INTRODUCTION

Recently several investigators 2 have reported results in shortening
the rest period of a number of woody plants by immersing the dormant
shoots in weak nutrient solutions or by injecting solutions of alcohol,
ether, and various acids into the twigs. These experiments have been
conducted in the laboratory with short cuttings of the plants. The
effect of such treatment has been to force the dormant buds out several
days ahead of the normal opening period.

During the last two years the writers have obtained similar and addi¬
tional results on a much larger scale by spraying dormant fruit trees
with strong solutions of certain commercial fertilizers, especially nitrate
of soda. Since these experiments have been conducted on the entire
trees in the orchard, it has been possible to observe the effects throughout
the whole season. The investigations have not yet been carried far
enough to permit drawing any conclusions regarding the physiologic
action of such spraying, but because of its practical value these prelimi¬
nary results seem deserving of attention at this time.

EXPERIMENTS IN 1912

In the course of the investigations of the writers on the control of
apple powdery mildew in the Pajaro Valley, Cal., it became evident
that the general vigor of the tree and the thriftiness of the foliage growth
had much to do with the success of the summer spraying treatment
for the control of the mildew, and after a number of experiments in
applying plant-food materials to the foliage in the form of summer
sprays, and after seeing that certain crude-oil emulsions used as dormant
sprays had a marked effect in stimulating an increased vigor of the trees
the following spring, it was decided to try the effect of a strong solution
of nitrate of soda as a winter or dormant spray. Caustic potash (potash
lye) was also added for the purpose of giving the spray an insecticide
value. The mixture was prepared according to the following formula:

Nitrate of soda . 50 pounds.

Caustic potash . 7 pounds.

Water . 50 gallons.

The experiment was conducted in a Yellow Bellflower apple orchard
owned by Mr. O. D. Stoesser, of Watsonville, Cal. This orchard is

1 These investigations were conducted in cooperation between the Office of Fruit-Disease Investi¬
gations of the Bureau of Plant Industry and the Office of the County Horticultural Commissioner of Santa
Cruz County, located at Watsonville, Cal. The writers5 names appear above in alphabetical order.

2 See references to literature, p. 444.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(437)

Vol. I, No. s
Feb. 16, 1914
G — 14

438

Journal of Agricultural Research

Vol. I, No. 5

situated about 5 miles from the ocean shore and is in a district that is more
subject to ocean fogs and trade winds than is the main portion of the
Pajaro Valley. It is a common characteristic of the numerous orchards
of Yellow Bellflower apples of this particular district that they bloom
abundantly, but set only a partial crop. The trees are on a deep sedimen¬
tary soil and grow well.

Seven 12-year-old trees were sprayed on February 2, 1912. The appli¬
cation was very thoroughly made, so that all of the small twigs were
drenched. About 7 gallons of spray solution were applied to each tree.
Adjoining this row on one side was a check row of seven trees which received
no winter spraying, and on the other side were several rows of seven trees
each which received various applications of crude-oil emulsions and soaps.
For the purpose of gaining some idea of the effect of nitrate of soda
used as a fertilizer, 50 pounds were applied as a surface dressing to one
vigorous tree selected from the row adjoining the nitrate-sprayed row.
This fertilizer was later plowed in and washed down by the rains.

EFFECTS ON BLOSSOMING AND ON THE FOLIAGE

Notes taken at the time the trees were coming out in the spring show
the following results:

April 7,1912. Trees in the row sprayed with nitrate of soda and lye are well in bloom ,
while those in the check row adjoining and in the remainder of the unsprayed orchard
are showing only an occasional flower fully opened.

April 14, 1912. The relative advancement of the row sprayed with a solution of
nitrate of soda and lye and the check plat is the same as noted on April 7 . The nitrate-
sprayed trees are nearly in full bloom, whereas comparatively few blossoms have
opened on the check plat.

When the check row had reached full bloom, the row sprayed with a solution of
nitrate of soda and lye was practically out of bloom.

Thus, the nitrate spraying advanced the blossoming time about two
weeks ahead of the normal period. It is characteristic of the Yellow
Bellflower variety of apples in the Pajaro Valley that the foliage buds
come out early, so that by the time the full-bloom period is reached the
trees are showing a considerable amount of young foliage. The nitrate
spraying produced a change in this respect. While the flower buds were
greatly stimulated in coming out, the foliage buds were not so much
affected, and the result was that when the trees sprayed with a solution of
nitrate of soda and lye were in full bloom and two weeks in advance of
the check trees in that regard, their foliage condition was relatively nearer
that of the check. Plate L shows the comparative stages of the nitrate-
sprayed and the check trees at that time. A decided contrast will be seen
in the relative advancement of the bloom on the tree sprayed with nitrate
of soda (PI. L, fig. 1) as compared with the check tree (PI. L, fig. 2). This
contrast is shown more in detail in Plate LI, in which figure 1 shows a
branch from a nitrate-sprayed tree, while figure 2 shows one from a check
tree. Both branches were collected on the same day. An examination
of the figures in Plate L will show that the advancement of the foliage on
the nitrate-sprayed tree is comparatively less marked than that of the
bloom. This same condition is shown in detail in Plate LI, in which it
will be seen that there is relatively little difference in the advancement of
the foliage of the sprayed and unsprayed branches. Later in the spring,
however, the effect on foliage growth became more pronounced, and the
sprayed trees assumed a more vigorous, green appearance than the check
trees. The single tree that received the 50 pounds of nitrate of soda
applied to the soil showed no greater vigor than the check trees.

Feb. 16, 1914

Winter Spraying with Nitrates

439

Both the row sprayed with nitrate of soda and the check row received
summer sprayings directed toward the control of apple powdery mildew
and of codling moth and various other insect pests. While the treatment
of the two rows was not the same, there was no essential difference in
the results — that is, the crop loss from codling moth and other insect
pests did not exceed 1 per cent on either plat and there was no damage
to the fruit from summer spraying. It is therefore evident that the
difference which showed up in the crop production of the two rows must
be attributed to the winter nitrate spraying.

CROP RESULTS

The check row of seven trees, which received no winter spraying but
which was properly protected by summer sprayings, produced 8 loose boxes
of fruit at picking time. On the other hand, the adjoining row, sprayed
in February with the solution of nitrate of soda plus lye, produced a
total of a little over 40 boxes. Thus, the winter nitrate spraying in¬
creased the crop production to fully five times that of the unsprayed row.
Similar adjacent plats, which were winter-sprayed with various crude-oil
emulsions and soap sprays, produced crops varying from 5 to 9 boxes
per plat. The single tree which received the 50 pounds of nitrate of
soda applied as a fertilizer gave no increased production, whereas none
of the trees in the nitrate-sprayed row failed to respond.

Regarding the single, heavily fertilized tree, it might be stated that in
addition to its showing no increase in production, the tree bloomed no
earlier than normal, there was no improvement in the growth and no
change in its general appearance throughout the growing season of 1912,
and in the spring of 1913 it came out normally and not differently from
the other trees in the same row, being one of the trees in a check plat.
The tree is still in normal condition and shows no noticeable effect from
the heavy fertilizing. The orchard is not irrigated, and the rainfall has
been much less than normal during the last two years.

Attention might again be called to the conditions under which these
results were obtained — namely, thrifty-growing trees in a deep residual
soil and having the characteristic of blooming abundantly each year
but setting only a shy crop. Even the 40 boxes produced by the nitrate
spraying does not represent the full crop that such trees should bear,
but the fourfold increase much more than paid for the cost of spraying,
and the possibility remains of still further increasing that production by
similar treatment in following years.

EXPERIMENTS IN 1913

The one small experiment on seven trees in 1912 did not furnish suf¬
ficient grounds for drawing any general conclusions as to the applicability
of winter nitrate spraying, but the striking results obtained opened a
wide field of inquiry. For instance, potash lye was added to the solu¬
tion of nitrate of soda in the experiment of 1912, so the questions arise
as to whether the lye was necessary and whether an acid medium would
increase or decrease the effect of the nitrate of soda; also, would a
weaker nitrate solution prove as effective and would other nitrogen-bear¬
ing fertilizer materials, such as lime nitrate, lime cyanamid, and sulphate
of ammonia, give similar results? Following along this line it would
be interesting to know what effect, if any, the other fertilizer elements,

440

Journal of Agricultural Research

Vol. I, No. s

potash and phosphoric acid, might have when applied as sprays, and
finally, what results might be obtained from a similar application of other
substances not ordinarily considered as having any particular fertilizer
value.

Experiments intended to answer these and a number of other more
or less important questions were started in February, 1913, in the same
orchard in which the previous year's work was done. Eleven 13-year-
old trees were used in each plat. A frost occurred at the time the fruit
was setting which ruined the crop and made it impossible to obtain
results in crop production. Data were obtained, however, on the effect
of the various sprays on the blossoming of the trees in the spring, and
the notes taken may be summarized as follows :

The plats sprayed with nitrate of soda at the rate of 1 pound to the
gallon came into bloom earlier than the check trees, just as they had done
in 1912. This effect was more marked in the cases in which lye was
added to the nitrate solution than when the plain water solution was
used — that is, the addition of lye in the proportion of 16 pounds of caus¬
tic soda in 100 gallons of spray solution increased the action of the nitrate
of soda in bringing the trees out earlier. Caustic soda appeared to be
just as effective as caustic potash. Nitrate of soda used at the rate of
half a pound to the gallon, either with or without the addition of lye, was
not nearly so effective as a solution of 1 pound to the gallon. A solution
of one-fourth of a pound to the gallon, with lye added, had practically no
effect. Nitrate of soda, at the rate of 1 pound to the gallon, to which
oxalic acid was added in the proportion of 50 pounds to 125 gallons of
solution, produced results similar to nitrate of soda plus lye, so far as the
effect of hastening the blooming period is concerned. Lime nitrate, 130
pounds in 100 gallons of water, and lime cyanamid, 92 pounds in 100 gal¬
lons of water, stimulated an earlier blooming of the trees, and subsequent
experiments will probably put these substances in a class with nitrate of
soda. Normal Yellow Bellflower apple blossoms have considerable pink
color, and it was interesting to note that when the trees sprayed with the
lime cyanamid came into bloom the flowers were nearly white. The
effects from sulphate of ammonia were not nearly so marked as those
from nitrate of soda. These various nitrogen-bearing fertilizer sub¬
stances were used in such strengths as to carry relatively the same quan¬
tities of nitrogen per gallon. Sulphate of potash had some effect in stimu¬
lating an early blooming, but double superphosphate did not. Of a num¬
ber of other substances tried, common salt used at the rate of 68 pounds
to 100 gallons of water produced a distinct effect.

It will be borne in mind that the above remarks apply simply to the
effects of the various sprays in causing an earlier blooming of the trees,
but since this early blooming was a striking characteristic of the nitrate-
sprayed trees of 1912, which showed a fourfold increase in production, it
seems permissible to conclude that this effect on the fruit buds is some
criterion of what might have been expected in the way of crop increase
had not the fruit been lost by frost.

The row of seven trees used in the nitrate experiment of 1912 was left
unsprayed this last season for the purpose of determining whether the
nitrate effect would continue to the second year. It was noticed that the
fruit buds on these trees were particularly large and plump, and some¬
what unexpectedly at blossoming time these trees came into bloom several
days ahead of the check rows. The bloom came out very uniformly all
over the trees, whereas ordinarily it is considerably delayed on the wind-

Feb. i <5, 1914

Winter Spraying with Nitrates

441

ward side. Also, the individual blossoms were conspicuously larger than
those of any other plat, and, so far as could be judged at the time the
frost occurred, a good crop was setting all over the trees. Thus, it ap¬
pears that this effect of the nitrate of soda had continued over to the
second year.

At present, all things considered, the best results have been obtained
by using a mixture made up as follows :

Nitrate of soda . 200 pounds.

Caustic soda . 25 pounds.

Water . 200 gallons.

In preparing this solution the required quantity of water was placed
in the spray tank and the agitator started. When the water was in motion,
the required weight of nitrate of soda was added gradually. Any large
lumps were first broken up into pieces about the size of hen's eggs. The
caustic soda was then added, and in about 15 minutes from the time the
preparation was begun the mixture was ready for applying.

The trees were very thoroughly sprayed on all sides, so that all of the
small twigs were drenched. The best results so far obtained have come
from the spraying applied about the 1st of February. Of course,
weather conditions must be taken into consideration. A rain immedi¬
ately following the application will wash much of the material off of the
trees, and it is probable that at least a week of clear weather should
follow the spraying, in order to insure good results.

In all of this work on spraying a solution of nitrate of soda on the trees
a considerable quantity fell to the ground, and the question will be raised
as to whether the various effects observed have not been simply the re¬
sult of the fertilizer action of the nitrate on the soil. About 7 gallons of
the solution were used in spraying each tree, and if the whole of this had
gone on the ground it would have amounted to about 7 pounds of ni¬
trate of soda per tree. The single tree in 1912 that had the 50 pounds of
nitrate applied to the soil therefore received over seven times the total
quantity applied to any single sprayed tree. As has been previously
stated, this single, excessively fertilized tree bloomed no earlier than
normal, produced no increased crop, and showed no improvement in
general vigor and appearance; whereas, none of the trees in the sprayed
plat failed to respond in all of these particulars. Of course, this single tree
test in the application of nitrate to the soil is too small an experi¬
ment to permit concluding positively that the effects that we have
reported from the spraying experiments are of an entirely different nature
and belong in a different category from those produced by the ordinary
soil application of nitrate. A careful consideration of the results of
ordinary orchard practice in fertilizing seems to make it plain that there
is no similarity between them and the results from spraying. For
instance, in the usual practice of applying nitrate of soda as a fertilizer to
apple orchards in the region of Watsonville, Cal., a winter or early spring
application does not force the bloom out 10 days or 2 weeks ahead of the
normal opening period and has had no measurable effect in increasing
the set of fruit that same year. The fact that the addition of caustic
soda or oxalic acid to the nitrate spray augments these various effects
further emphasizes the difference between the results from spraying and
the ordinary results from the application of fertilizer. Caustic-soda
solution alone applied as a spray has no effect on the time of blooming or
the crop production.

442

Journal of Agricultural Research

Vol. I, No. 5

EXPERIMENTS OF GROWERS IN 1913
YELLOW BELLFLOWER APPLES

During the past season a number of growers made more or less exten¬
sive tests of the spraying with nitrate of soda. An aggregate of several
hundred acres of Yellow Bellflower apples was sprayed with nitrate of
soda plus caustic soda, but practically all of this acreage was in the
same district in which the writers' experiments were conducted, so the
crop was lost by frost. It was noticeable during the past summer,
however, that the foliage in such orchards as received very thorough
winter nitrate sprayings had a better appearance than in years past,
due apparently to the effect of the nitrate. One orchard, that of
MacDonald & Sons, is located in a district that practically escaped frost
damage, and the results obtained indicated a marked crop increase in
consequence of the spraying. The entire orchard, with the exception of
a few trees, was sprayed with various combinations of nitrate of soda
and lye, and, while no exact data on the production of the unsprayed
trees as compared with the rest of the orchard was obtained, the amount
of fruit on the trees indicated that the spraying had produced a marked
increase. This conclusion was more reliably substantiated by comparing
the total orchard production this year with that of previous years.

SWEET CHERRIES

Mr. A. W. Taite, of Watsonville, sprayed portions of two blocks of
Napoleon (Royal Ann) cherries with nitrate of soda, i pound to the
gallon, to which caustic soda was added at the rate of 25 pounds to 200
gallons. Unsprayed rows adjoining the sprayed ones were left in each
block. In one case the sprayed trees were distinctly advanced over the
check trees in coming into bloom. In both cases there was an increase
in the foliage growth and a consequent improvement in the appearance
of the trees. No effect on crop production could be noticed, though it
is possible that treatment in successive years may bring such results.

PEARS

For our observations on pears the writers are indebted chiefly to Mr.
George Reed, of San Jose, who carried out extensive tests in the orchards
of the J. Z. & G. H. Anderson Fruit Co. The spraying was done about
the 1st of February and the following notes are taken largely from
Mr. Reed's observations :

Clairgeau. — Four rows of about 40 trees each were sprayed with commercial
lime-sulphur solution (33 0 Baume) diluted 1 to 9. Adjoining these were four rows
sprayed with lime-sulphur solution diluted 1 to 9 and to which was added nitrate of
soda at the rate of 1 pound to the gallon of the diluted spray. The rows sprayed
with the combined solution of nitrate of soda and lime-sulphur came into bloom
about a week ahead of those that received the lime-sulphur solution alone . The devel¬
opment of the fruit on these nitrate-lime-sulphur solution rows continued to show
an advancement of about a week throughout half the growing season, and at picking
time the fruit was greener and hung on better than that of the plain lime-sulphur-
solution rows. Both plats bore a full crop, so there was no opportunity for observing
any effect on production. The Clairgeau variety blooms early, and the further
advancement due to nitrate spraying might result in frost injury in some localities.
The fruit ordinarily has a habit of dropping off during the latter part of the growing
season. This difficulty, however, was largely eliminated on the nitrate-sprayed rows.

Feb. 16, 1914

Winter Spraying with Nitrates

443

Comice. — The major portion of the block was sprayed with a plain water solution
of nitrate of soda at the rate of 1 pound to the gallon. A small portion was sprayed
with commercial lime-sulphur solution, diluted 1 to 9, with nitrate of soda added
at the rate of 1 pound to the gallon of diluted spray. Through a misunderstanding
the men doing the spraying left no check rows in this block, so that crop data couln
not be obtained. However, Mr. Reed’s exact knowledge of the previous productiod
of this block as a whole indicates that the marked increased production this last
season was more than probably due to the nitrate spraying. The Comice is a rela¬
tively shy bearer, and a valuable pear commercially, so that any increased production
that could be obtained by nitrate spraying would be much appreciated by the
grower. One portion of the block that regularly produces less than the remainder
gave a good crop this year, and it appeared that the addition of the lime-sulphur
solution augmented the effect of the nitrate of soda just as the addition of lye has
done in the experiments of the writers.

GlouT Morceau. — A block of Glout Morceau pears was sprayed with the combi¬
nation of lime-sulphur solution, diluted 1 to 9, plus nitrate of soda 1 pound to the
gallon of diluted spray. This block had never produced a full crop, and while no
unsprayed checks were left, the increased production this year would appear to be
due to the nitrate spraying.

"Winter Nelis. — A block of Winter Nelis pears was sprayed with a solution of
nitrate of soda 1 pound to the gallon of water. No lime-sulphur solution was added
in this case. No check rows were left, and a frost destroyed a large percentage of
the fruit after it had set. However, at that time the trees were carrying the largest
crop they had ever produced, and again it would appear that the nitrate spraying
had had a beneficial effect. The trees came into bloom about 10 days ahead of
normal opening period.

DISCUSSION OF RESULTS AND SUMMARY

It is not the writers' intention to convey the impression that dormant
spraying with nitrate solutions will solve the problem of shy bearing of
fruit trees nor offer a more advisable method of applying nitrogen
fertilizer. The purpose of this paper is simply to present the results
as they now stand.

It is evident that, at least under certain conditions, some varieties of
apples and pears that are more or less self-sterile may have their crop
production materially increased by dormant spraying with solutions of
nitrate of soda plus lye. The combination of a solution of nitrate of soda
and lime-sulphur is apparently capable of bringing similar results.

Actual quantitative data on increased production from spraying with a
solution of nitrate of soda are available from only one source, that of
the first experiment on Yellow Bellflower apples in 1912. No production
records were obtainable from the various tests made by growers during
the season of 1913, but the one test on Yellow Bellflower apples and
several others on pears indicate that such an increase had undoubtedly
been brought about. It is considered that the growers' knowledge of
the crops of the previous years as compared with that of this year fur¬
nishes a basis for conclusions that are at least corroborative.

That nitrate spraying of dormant trees will bring about an earlier
blooming of certain varieties of fruit is a satisfactorily established fact,
which has been demonstrated on Yellow Bellflower apples at Watson¬
ville, Cal., and on various varieties of pears at San Jose, San Juan,
and Suisun, Cal., during the past season. How generally this statement
will apply to other varieties of apples and pears and in other localities
remains to be determined. Results on stone fruits have not been as
striking as those on pears and apples, but it is possible that stronger
solutions, earlier spraying, or a repetition of the spraying in successive
years may bring about such results.

444

Journal of Agricultural Research

Vol. I, No. s

The greater danger of injury from frost that might result from forcing
trees into bloom earlier than normal would have to be taken into con¬
sideration in making practical use of nitrate spraying in winter.

Aside from the effect on crop production, there has also been a very
noticeable improvement in the color, abundance, and vigor of the foliage,
and it seems possible that nitrate spraying of dormant trees may be a
valuable supplement to the ordinary fertilizer practices in obtaining
quick results in orchards suffering from lack of nitrogen.

The writers will make no attempt at present to explain the peculiar
effect of nitrate of soda in increasing the production of more or less
self-sterile varieties of fruits, or in improving foliage growth. The
similarity between the writers' results in forcing dormant buds by
winter nitrate spraying and the results obtained by other investigators
by treating cuttings with various weak solutions has been mentioned.
In experiments of the writers, however, a more or less lasting effect on
the vigor of the foliage and also some valuable results in increasing crop
production have been obtained. It furthermore appears that the effects
obtained by spraying with a solution of nitrate of soda may continue
over to the second year, as shown by the original plat of 1912, which was
left unsprayed in the winter of 1913.

The effects of the nitrate spraying seem to be proportional to the
strength of the solution employed and the thoroughness with which
it is applied. The addition of caustic soda materially increases this
action.

LITERATURE

The following is a short list of some of the more recent literature on
forcing the buds of dormant cuttings of woody plants.

Jesenko, Franz.

Einige neue Verfahren die Ruheperiode der Holzgewachse abzukurzen. Ber.
Deut. Bot. Gesell., Bd. 29, Heft 5, p. 273-284, pi. 12, 1911; Bd. 30, Heft 2, p. 81-93,
pi. 3, 1912.

liber das Austreiben im Sommer entblatterter Baume und Straucher. Ber.
Deut. Bot. Gesell., Bd. 30, Heft 4, p. 226-232, pi. 9. 1912.

Lakon, Georg.

Die Beeinflussimg der Winterruhe der Holzgewachse durch die Nahrsalze.
Ein neues Friihtreibe verfahren. Ztschr. f. Bot., Jahrg. 4, Heft 8, p. 561-582.
1912.

Weber, F.

liber die Abkurzung der Ruheperiode der Holzgewachse durch Verletzung
der Knospen, beziehungsweise Injektion derselben mit Wasser. Sitzber. K.
Akad. Wiss. [Vienna], Math. Naturw. Kl., Bd. 120, Abt. 1, Heft 3, p. 179-194,
pi. 1. 1911.

Plate Iy. Fig. i. — Yellow Bellflower apple tree in full bloom on April 16, 1912,
showing effect of spraying with a solution of nitrate of soda plus caustic
potash on February 2 previous.

Fig. 2. — Unsprayed check tree for comparison with figure 1.

The illustrations are from photographs taken on the same day.

Plate L

Plate LI. Fig. i. — A branch from a Yellow Bellflower tree in full bloom on April io,
1913, showing the effect of spraying with a solution of nitrate of soda
plus caustic soda on February 3 previous.

Fig. 2. — A branch from an unsprayed check tree for comparison with
figure 1.

The illustrations are from photographs taken on the same day.

JOURNAL OF AM1TOAL RESEARCH

DEPARTMENT OF AGRICULTURE

Voe. I Washington, D. Cm March 25, 1914 No. 6

TYLOSES: THEIR OCCURRENCE AND PRACTICAL
SIGNIFICANCE IN SOME AMERICAN WOODS

By Eloise Gerry,

Micro scopisi, Forest-Products Laboratory , Forest Service
GENERAL DESCRIPTION OF TYLOSES

The large open pores or vessels conspicuous in hardwoods frequently
become closed by growths called tyloses.1 These growths render the
wood practically impermeable to air and liquids. On the split surfaces
of a wood such as white oak or pignut hickory the tyloses appear in the
vessel channels as glistening cellular growths resembling masses of soap
bubbles. (PI. LII, fig. 1.) These masses are protrusions from the living
parenchyma cells of the wood itself into adjacent vessel or tracheid cavi¬
ties. They enter at the thin places or pits in the wall of the wood ele¬
ments (see PI. LII, figs. 2 and 3), and expand to a greater or less degree.
In the softwoods (PI. LVI, fig. 1) tyloses are relatively small, but in the
hardwoods they frequently form bladderlike sacs of considerable size
(PI. LII, figs. 2 and 3, and PI. LIII, figs. 1,2, and 3), often developing
simultaneously in many of the parenchyma cells surrounding the tube¬
like vessel cavities. (PI. LII, fig. 3.) Under such circumstances, if
growth is vigorous, the tylosal sacs, after pushing into the vessel cavity,
grow together, completely filling it. In this way the ability of the
vessel to conduct air or liquid is effectually checked. (PI. LIII, figs. 1
and 2.) Sometimes, however, the tylosal growths do not entirely fill the
vessel, and only a clogging action results.

The purpose of this study was to determine the occurrence of tyloses in
the most important commercial species of native woods and their signifi¬
cance in relation to the adaptability of these woods to certain practical uses.

Observations were made not only of the presence or absence of tyloses
in a species, but also of the extent and degree of development and the
regions (sapwood or heartwood) where the growths are found.

1 The$e growths received, in 184s, the name “ThyUe” (tyloses) from a German botanist who signed as
“Ungenannte,” or ''unknown,” the paper discussing them. This writer is, however, believed by Boehm
and Winkler to have been Frl. Hermine von Reichenbach. The name " Thylle ” is derived from the Greek
word 6Mcc, meaning a purse or sack. The occurrence of tyloses was, however, noted as early as 1675 by
Malphigi, in the drawing of a cross section of chestnut wood. They are also given the descriptive name
"Fiillzellen,” or filling cells, by the Germans.

(44s)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Vol. I, No. 6
Mar. 25, 1914

446

Journal of Agricultural Research

Vol. I, No. 6

Only a brief discussion is given of the causes leading to the formation
of tyloses or of their function in the living plant, since studies for this
purpose have already been made by other investigators.

MORPHOLOGICAL RELATIONS OF TYLOSES IN WOODY TISSUE
ORIGIN AND DEVELOPMENT

A tylose can not be considered as a distinct cell, for as a rule a cell is
defined as a body consisting of cell substance, cell wall, and cell nucleus.
With very rare exceptions (Molisch) 1 a tylose, as found in woody tissue,
is not completely surrounded by a wall and has no nucleus. It is only a
portion or prolongation of a wood or medullary-ray parenchyma cell.
(PI. EH, figs. 2 and 3 ; PI. LVII, fig. 2.) Frequently more than one tylose
is formed from one parenchyma cell, but only one active nucleus — that of
the parenchyma cell — is present, though this may be found in one of the
tyloses. (PI. EH, fig. 3.) A parenchyma cell which has given' rise to
two tyloses is shown in Plate LH, figure 2.

The growing or arching out of tyloses has been found to follow a re¬
duction in internal pressure or cessation in sap conduction in the large
vessels. When this occurs, the living parenchyma cells, which possess a
considerable growth potential, expand and press into the adjacent empty
vessel cavities. In pitted vessels this expansion is localized in the thin
unlignified membranes of the one-sided bordered pits which are present
on the dividing walls between vessels or tracheids, and parenchyma
(De Bary; Green; Haberlandt; Hanausek; Molisch; Rees; Russow; Sachs;
Strasburger; and Winckler). These membranes contain plasma and
therefore possess the power of growth. The internal pressure of the
turgid parenchyma cells, when exerted against these relatively thin spots
or pits, causes the pit membranes to stretch and grow by intussusception 2
(Green; Molisch). The protrusions increase gradually in size and finally
develop into the characteristic bladder-shaped sacs known as tyloses. An
open passage through the space previously occupied by the unstretched
closing membrane of the pit is formed in this way between the tylose and
the parenchyma cell. (PI. EH, fig. 2.) The contents of the tylose are
therefore the same as those of the parenchyma cell.

NORMAL AND ABNORMAL TYLOSE FORMATION

It has been shown beyond doubt that the wounding of trees through
cuts or bruises or at the points where branches are broken off tends to
stimulate tylose formation, and throughout the study this mode of
tylose formation has been constantly borne in mind. Generally, how¬
ever, tyloses are not due to wounding. They are a characteristic feature
of the normal uninjured wood of many families of trees. Nevertheless,

1 Bibliographic citations in parentheses refer to "Literature cited, ” pp. 468-469.

2 "Intussusception" means in botany, according to Nageli, the growth of cell walls by the irregular inter¬
position of new solid particles between those already in existence.

Mar. 25, 1&14

Tyloses in American Woods

447

the wood produced by felling the tree may have an important bearing on
the presence of tyloses in the outer rings of a log, where the parenchyma
cells are still living and capable of growth. It is possible to find in these
rings young or old, or large and small, tyloses together in the same vessel.
(PI. LIV, i?3<) Although exceptions have been noted, the idea that a
considerable number of the outer rings are entirely free from tyloses
has, however, been very generally accepted (Strasburger).1 The data
obtained from the present study show that there is a very considerable
formation of tyloses in the outer rings of the sapwood. The question
then arose as to whether these sapwood tyloses were of normal origin or
whether they were due to some wound stimulus, such as the felling of
the tree. It was finally concluded that they were normally formed
tyloses, because their development throughout the vessels was very
uniform instead of being sporadic or irregular, as in the case of tyloses
associated with wounds (PI. LIV, Ri and R2), and because an exami¬
nation of branches from living trees of Rhus, the sumach, Catalpa, and
Robinia, the black locust, made immediately after cutting, confirmed the
other observations of the relatively early formation of tyloses in many
species. In material which was not received for examination until
several weeks after it was cut, thin, irregularly distributed tyloses were
often found in the outer vessels, though the latter must have been func¬
tioning in sap condition at the time the tree was felled.

It is noteworthy also that in this study tyloses were found to reach the
most remarkable development in ring-porous woods, such as oak, hickory,
black locust, or osage orange. (PI. LIU, figs. 1 and 3, and PI. LVI,
fig. 2.) In woods where tyloses are few and scattered there is consider¬
able variation from specimen to specimen in the actual number of tyloses
present. This tendency is clearly shown in the woods of the diffuse
porous group. (Table II.) It is also noticeable that in the two or three
rings surrounding the pith in a diffuse porous wood tyloses are often
much more abundant than elsewhere in either the heartwood or sapwood.

EFFECT OF THE DISTRIBUTION OF PARENCHYMA TISSUE

Since tylose formation depends upon the presence of parenchyma cells
either in the form of wood parenchyma or medullary rays in close prox¬
imity to vessels or tracheids, the variation in position, abundance, and
vitality of these cells affords at least a partial explanation of the irregular
development of tyloses in different species of wood. Parenchyma tissue
is considerably developed in the following families and their respective
genera.2 This study has shown that in these families are a large number
of native woods exhibiting ^tyloses.

1 Tyloses are . . . instrumental in closing the water courses of the heartwood. . . . These are intrusive
growths from living cells which penetrate the cavities of the adjoining tracheal elements during the transi¬
tion of sapwood into heartwood.

2Solereder, Hans. Systematic Anatomy of the Dicotyledons . . . v. 2, p. 1143. Oxford, 1908. Certain
other woods with abundant parenchyma frequently produce gummy substances rather than tyloses.

448

Journal of Agricultural Research

Vol. I, No. 6

Family. Genera.

Cupuliferae or Fagaceae . Castanea, Fagus, Quercus.

Juglandaceae . Hicoria Juglans.

Papilionaceae . Robinia.

Magnoliaceae . Liriodendron, Magnolia.

Moracese . . Morns, Toxylon.

The arrangement of wood parenchyma cells in the annual ring has
been divided 1 into three different types, as follows:

1. Terminal parenchyma, which is situated at the periphery of the annual growth
ring, on the outer face of the summer wood.

2. Metatracheal or diffuse parenchyma, which is scattered among the other ele¬
ments in the ring, usually forming tengential bands.

3. Paratracheal or vasicentric parenchyma, or parenchyma cells, aggregated around
the vessels.

Table I. — Native woods grouped according to the degree of tylose development and the
most marked distribution of wood parenchyma in ring.2 3

Abundant Tyloses.3

Species.

Type of
parenchyma.

Species.

Type of
parenchyma.

Catalpa speciosa .

Paratracheal.

Do.

1 Do.

Do.

Hicoria orata .

Paratracheal.

Do.

Chilopsis linearis .

Juglans cinerea

Morns rubra .

* nigra

Rhus hirta .

Quercus alba

Robinia pseudacacia .

garryana .

Toxylon pomiferum .

lyrata .

Hicoria alba .

lohata

aquatica .

macrnearpa

glabra .

michauxii . .

lacinosa .

minor .

minima .

platanoides

myristicaeformis .

densiflora

odorata .

marilandica

Many Tyloses.

Castanea dentata .

Fraxinus lanceolata . .

Paratracheal.

Celtis occidentalis .

. j Paratracheal.

profunda

Eucalyptus globulus .

. | Do.

quadran gulata

Do.

Fagus atropunicea .

Sassafras sassafras .

Fraxinus americana .

Scattered Tyloses.

Aesculus vetandra .

Scanty para¬

Platanus occidentalis

Metatracheal

tracheal.

Populus grandidentata .

Terminal.

Liquidambar styraciflua .

Metatracheal.

tremuloides

Liriodendron tulipifera .

Terminal.

trichocarpa

Do.

Magnolia acuminata .

Do.

Ulmus alata .

Paratracheal.

fraseri .

Do.

americana

glauca .

Do.

pubescens

Do.

1 Jeffrey, E. C. A Natural Classification of Woods.

Holden, Ruth. Some features in the anatomy of the Sapindales. In Bot. Gaz., v. 53, no. 1, p. 50-58,
pi. 2-3. 1912.

2 The data here given concerning the distribution of parenchyma were obtained from: (1) Solereder,
Hans, op. cit.; (2) Jeffrey, E. C., op. cit.; and (3) from original observations made during the study.

3 By “abundant” is meant a very large number. “Many” is aised to signify a considerable number
but less than 1 ‘ abundant.”

Mar. 25, 1914

Tyloses in American Woods

449

These three types of arrangement and the degree of their development
bear a definite relation to the development of tyloses, since they indicate
whether the parenchyma cells are near enough to the vessel cavities to
send their prolongations into them. In addition to the wood paren¬
chyma, the position and number of the medullary rays adjacent to the
vessels must be taken into account. A grouping of the species of wood
with the twofold object of indicating the distribution of tyloses and the
arrangement of the wood parenchyma clearly brings out some of the
reasons why tyloses are so much more abundant in certain woods than
in others. Wherever the paratracheal or vasicentric type of parenchyma
is well developed, the tendency for marked tylose formation, or else for
gum production, is very noticeable. From Table I it is further evident
that when tyloses are strongly developed either paratracheal or abundant
metatracheal parenchyma is always found.

SHAPE, THICKNESS OF WALL, AND CONTENTS OF TYLOSES

The shape of the tylosal projections varies widely. They are some¬
times spherical, or again they appear as elongated vesicles. (PI. LII,
fig. 3; and PI. LIII, figs, i, 2, and 3.) Often when the walls are very
thin they appear much collapsed and wrinkled as, for instance, in ash
or the wound tyloses in cow oak. (PI. LIV, Ri.) The extent to which
the tylose wall increases in thickness varies also. The wall may be an
extremely thin delicate membrane as found in ash or osage orange (PI.
LV, fig. 2) or it may be of medium thickness as in oak. (PI. LIII, figs.
1 and 2.)

The contents of the tyloses are in general the same as those of the paren¬
chyma cells producing them. Starch is common, and resin, calcium
crystals, and gums have also been observed.

When normal parenchyma cells do not give rise to tyloses, the so-called
“gums” (Prael)1 are often produced, as in mesquite, maple, or cherry.
This gum usually collects in the vessels (PI. LIII, fig. 4) and parenchyma
cells. In the vessels it sometimes assumes the form of globules or drop¬
lets which may easily be mistaken for tyloses. In order to determine
whether gum or tyloses are present, a section of the wood may be treated
with some gum solvent, such as absolute alcohol or caustic soda. When
the wood is dry, the gum droplets are often characteristically cracked
and split. Their general appearance is illustrated in Plate LIII, figure 4.

MATERIAL AND METHODS USED IN THE STUDY

The material used for this study of tyloses was a collection of logs of
commercial size from native-grown trees. As a basis for the study of
tyloses this material was unique, since most of the work of other inves¬
tigators has been done not on wood from the bole of the tree, but on

1 “ Schutzgummi.

450

Journal of Agricultural Research

Vol. I, No. 6

branches, twigs, roots, leaves, vines, herbaceous plants like the squash,
or on such of the lower forms as ferns,1 and did not cover to any extent
the American species.

The method of examining the wood was as follows: The ends of the
logs which form the collection of commercial American woods (PI. LIX,
fig. i) of the Forest-Products Laboratory were examined with a hand
lens. Blocks cut from these were also studied microscopically. Small
strips extending from the bark through the trees to the pith, including
the sapwood, the so-called transition region, and the heartwood,2 were
cut from the logs. Microtome sections about i inch by one-half inch
in area and 5 to 20 micromillimeters in thickness were cut from the three
planes, transverse, radial, and tangential, taken from each of these dif¬
ferent regions and were studied under the compound microscope. The
observations for hardwoods are given in Table II. Stains were often
employed to differentiate the tissues, and macerations were made with
potassium hydroxide or chromic acid for special studies of the relations
between the tylose and the parenchyma cell producing it. Fresh mate¬
rial from seedlings and branches was also examined, in order to deter¬
mine whether the sapwood tyloses were of normal or abnormal origin.

The Forest-Products Laboratory collection of woods begun in 1910 is
not yet complete, and in many cases only one log of a species was available
for study. Nevertheless, the majority of the commercially important
Species are included in the laboratory collection, and in addition to the
study of these it was possible to make further observations on authentic
material of a number of other important species. Moreover, whenever
two or more specimens of the same species were examined, results were

1 This list of the plant genera 'where tyloses have been found in wood, roots, leaves, or other portions is

given by Kiister. It includes Molisch’s observations on the Vienna wood collection and other material

as well as those of other authors, whose names are given in parentheses after the genera they investigated.

Abies (Raatz).

Coccoloba.

Taurus.

Portulacca.

Achyranthes.

Coleus.

Ligustrum.

Prunus (Wieler).

Aesculus (Maiile, Tison).

Convolvulus (Dutailly).

Toranthus.

Pterocarya.

Alnus (Tison).

Comus (Maiile).

Toxapteryguim.

Quercus.

Ampelopsis.

Corypha.

Machura.

Rhus.

Aralia.

Cucumis.

Mansoa.

Ricinus.

Aristolochia (Tison).

Cucurbita.

Maranta.

Robinia.

Artocarpus.

Cuspidaria.

Micania.

Rosa (Maiile).

Arundo.

Dahlia.

Morns.

Rubia.

Asarum.

Diospyros.

Musa.

Rumex (Dutailly).

Banisteria.

Elaeagnus.

Ochroma.

Salix.

Begonia.

Euphorbia.

Olea.

Sambucus.

Betula.

Fagus.

Ostrya.

Santalum.

Bigonia.

Ficus.

Passiflora.

Schinus.

Boehmeria.^

Fraxinus.

Paulownia.

Sideroxylum.

Broussonetia.

Gleditsia (Tison).

Perilla.

Solanum.

Byronia.

Hammamelis (Tison).

Pharbitis.

Sparmannia.

Canna.

Hedera.

Philodendron.

Strelitzia.

Carica.

Hedychuim.

Phyllanthus.

Styinatoph yllum.

Carya.

Cassis.

Heliconia.

Picea (Raatz).

Taraxacum.

Humulus (Tubeuf).

Pinus (Raatz).

Thunbergia.

Castanea.

Inula.

Piratinera.

Ulmus.

Catalpa.

Jatropha.

Pistacia.

Urtica.

Celtis.

Juglans.

Plantago.

Vitis.

Chiliantus.

Cladrastis (Tison).

Koelreuteria.

Latania.

Platanus.

Populus.

Xanthoxylon (Tison).

2 The cross section of a mature tree may be divided into at least two regions: The outer or last-formed
rings, variable in number, which are termed the “sapwood” or “alburnum,” and the inner rings around
the pith or center of the tree, which in dry material are sometimes indistinguishable in appearance from
sapwood, but which are more often definitely marked by a difference in color and are then termed the
“heartwood” or “duramen.” (PI. UX, fig. i.)

Mar. 25, 1914

Tyloses in American Woods

45i

found to check reasonably well, as shown in Table II. The greatest
variation occurs in the species in which tyloses are very rare or else
scatteringly developed and, therefore, where their practical importance
is relatively slight.

OCCURRENCE OF TYLOSES IN NATIVE HARDWOODS

Table II gives the results of observations made on the distribution and
region of first development of tyloses in 143 specimens of hardwoods
grown in the United States. The very marked development of tyloses
in certain species has been noted in Table I.

Special attention was given to the early development of tyloses. The
results show their presence in the sapwood of all the species in which
they occur in the heartwood. The hickories, for instance, give some
interesting data concerning the occurrence of tyloses in sapwood. It
has been maintained that if tyloses ever occurred in sapwood they would
be found only in very narrow sapwood — that is, where the transition from
sap to heartwood begins at the end of the first or second year after the ring
is formed, as, for instance, in some of the oaks. In the hickories, however,
tyloses are always present in the sapwood, and are generally developed
even in the outermost rings as abundantly as in the heartwood. Plate
Till, figure 3, shows a cross section of the sapwood of pignut hickory
(Hicoria glabra) , including the fourth to the seventh rings in from the bark.
This particular tree had 31 rings of sap, or uncolored wood, and tyloses
were well developed in the very outermost rings. (PI. LIX, fig. 1.)

Tyloses are normally lacking in the red-oak group, although there are
many exceptions. An illustration of vessels not filled by tyloses is given
by those in the middle of Plate LIV, R2, and by some of those in Plate
LV, figure 1 . In some cases tyloses occur in individual vessels in species
ordinarily free from them, as Spanish oak. (Table II.) In several instances
the few scattered tyloses present in both the sapwood and heartwood
have a rather abnormal appearance and are associated with areas of
fungous growth. (Table II, Scarlet oak.) In certain species of the red-
oak group, however, as blackjack oak (Quercus marilandica) , tyloses are
very generally developed in both the sapwood and heartwood.

In the white oaks, in contrast to the red-oak group, tyloses are generally
very abundant, even in the outermost rings. Some of the white oaks
where tyloses are slow in forming show striking examples of the growth
and development of the tylose in its early stages. This is illustrated in
Plate LII, figure 3, which is a reproduction of a photomicrograph of a
cross section of California white oak, or valley oak ( Quercus lobaia) , show¬
ing a piece of the sapwood next to the bark. Fragments of the bark may
be seen at the top of the illustration. The relatively small bladderlike
cells here shown increase in size until they grow together and fill the
vessels as shown at the bottom of this illustration and in Plates LIII,
figures 1, 2, and 3, and LV, figure 2.

452

Journal of Agricultural Research

Vol. I, No. 6

rare

Hamamelis virginiana L . I Witch-hazel

Mar. 25, 1914

Tyloses in American Woods

453

1 Both the Latin and common names used are those given by G. B. Sudworth in Bulletin 17, Division of Forestry, Department of Agriculture, 1898, and in a later unpublished
revision of the same.

* One of the Rosaceae, tyloses generally lacking in this family (Molisch).

Tabte II. — Occurrence of tyloses in the large vessels of the hardwoods — Continued.
The Ring Porous Woods.

454

Journal of Agricultural Research

Vol. I( No. 6

Honey locust . ' . * . ' None present . J None present (gum frequently found).

Gymnocladus dioicus (I,.) Koch . I Coffee tree

Mar. 25, 1914

Tyloses in American Woods

455

Table II. — Occurrence of tyloses in the large vessels of the hardwoods — Continued.
The Ring Porous Woods— Continued.

456

Journal of Agricultural Research

Vol. I, No. 6

Mar. 25, 1914

Tyloses in American Woods

457

1
0. a
8 8
SiS

VS V
S3 5

fj i
◄ «)

60666
ft Q P P fl

666666

pppppp

458

Journal of Agricultural Research

Vol. I, No. 6

TYLOSES IN SOFTWOODS

Coniferous or softwoods lack the large open pores or vessels which
characterize the hardwoods. They also either lack or show a scanty
development of wood parenchyma, the chief source of tylose formation
in the hardwoods. Since it is in relation to the closing of the vessels that
tyloses are of practical significance, the study of tylose distribution in
the conifers is of relatively small importance. However, since tyloses
or tyloselike cells are often present in the tracheids or in the resin canals
of certain normal coniferous woods, and since they have been found to
play some part in penetration of wood preservatives and in resin flow,
their occurrence in the softwoods was studied.

The occurrence of tyloses in coniferous woods has not received the
attention given to their occurrence in hardwoods. Often their presence
has been ignored, or they have been reported as entirely lacking.1 When
studied, moreover, investigations were usually confined to parts of the
plant other than the wood,2 though there are a few notable observations
on their occurrence in the wood itself (Boehm; Chrysler; Con wentz;
Krister; Mayr; Penhallow; Raatz).

TRUE TYLOSES IN CONIFERS

Tyloses in normal coniferous wood arise chiefly from the parenchyma¬
tous cells of the medullary rays. (PI. LVI, figs, i and 2.) As in the
hardwoods, it is by the growth of the membranes of the one-sided bor¬
dered pits that tyloses are formed, especially where the pits are of large
size, as in the white pines. In this case tyloses grow into the lumen of
the tracheid, just as in hardwoods they grow into the vessels or pores.
Tracheids, like vessels, function as sap conductors, but instead of having
in their end walls actual openings of considerable size they have only rela¬
tively thin regions or pits. These are more or less completely closed by
an irregularly thickened membrane, portions of which sometimes contain
very minute perforations (Bailey). Thus in these elements already
closed or nearly closed, tyloses have not the effect that they have in the
open vessels of the hardwoods. Moreover, tylose formation of this type
in conifers can only take place in a comparatively small percentage of the
tracheids — that is, in those adjacent to the medullary-ray parenchyma
cells produced as a result of wounds (Boehm; Raatz).3

TYLOSELIKE CELLS IN THE RESIN CANALS

Aside from true tyloses, there is often observed in certain species of
conifers a partial or complete closing of the resin canals, produced by
parenchyma cells, but not by growth of the membrane of the one-sided

1 Reported by Molisch after examining 700 species of plants of all sorts.

* They are said to be more abundant in the root than in the stem (Raatz). They also have been studied
in the leaf and in the cone axis.

8 Boehm and Raatz observed tyloses as a result of wounding in Abies pectinata, Pinus syhestris, Pinus
strobus, Pinus excelsa, Larix europea , and Thuja occidentals.

Mar. 35, 1914

Tyloses in American Woods

459

bordered pit. Such growths are termed “tyloselike,” since they produce
an effect very similar to that produced by the true tyloses of the hard¬
woods.

Resin canals or ducts are normally present in the following coniferous
genera: Inarch, or tamarack (Larix) , spruce (Picea), Douglas fir (Pseu-
dotsuga), and pine (Pinus). These canals when seen in cross section
often bear a superficial resemblance to the vessels or pores of the hard¬
woods. (Pl. LVII, fig. 1.) They are, however, different in both their
origin and function. Resin ducts are not cellular elements, but simply
intercellular spaces which result from the splitting apart of the common
walls of a group of parenchyma cells. A very early stage of this splitting
is shown in Plate LVI, figure 1. These parenchyma cells which surround
the canal opening are called “epithelial cells/’ They are the seat of
resin formation, and they cause the tyloselike closing of the resin canal.
Certain of them often remain thin walled and contain plasma. (PI. LVIII,
figs. 2 and 5.) After they split apart to form the canal, when they change
in shape and size, a further swelling and growth may take place which
closes the canal entirely or in part. (Pis. LVII, figs. 1 and 2, and LVIII,
figs. 2, 5, and 6.) The fact that it is the growth or expansion of the
whole cell, and not a portion of the wall of that cell, together with a
portion of the wall of the neighboring cell, as in the tylose-forming
membrane of the one-sided bordered pits of the hardwoods, clearly
indicates the difference between the true tyloses of the hardwoods and
the tyloselike cells in the resin canals of the conifers.

OCCURRENCE) OF TYLOSES AND TYLOSELIKE CELLS IN NATIVE

CONIFERS

Over 600 permanently mounted sections from coniferous woods in
the collection of the Forest-Products Laboratory were specially studied,
while more than three times this number were examined unmounted.

true tyloses

Ray or true tyloses were found in the normal wood of the conifers,
but were not abundant. Their shape and general appearance are well
illustrated in Plate LVI, figures 1 and 2. None of the long, saclike
vesicles which sometimes fill the whole tracheid lumen in the roots of
conifers were found. The greatest development of true tyloses was
found in the soft pines. In this group they were better developed in
spring wood than in summer wood and were more numerous in the sap-
wood than in the heartwood. Indeed, some of the pit membranes in the
heartwood were concave in shape, appearing to have collapsed inward
instead of protruding into the tracheid.

The size of the pits between the medullary ray cells and the tracheids
in conifers bears a definite relation to the formation of tyloses. As a
rule, the ray pits in the hard pines are small and tyloses are lacking,

460

Journal of Agricultural Research

Vol. I, No. 6

Norway pine (Pinus resinosa ), which is regarded as a hard or pitch pine,
offers an exception to this. Here we find numerous tyloses, but here
also we have large ray pits. The only soft pine examined which did not
contain tyloses was pinon pine (Pinus edulis ). This species is character¬
ized by small ray pits instead of the large ones common to this group.

Of the other conifers all of the species listed below have small ray pits.
No true tyloses were found in these species. (See Table III.)

Table; III. — Occurrence of true tyloses in native conifers .
Soft Pines.

Species.

Number
of speci¬
mens.

Sap wood.

Heart wood.

Limber pine (Pinus flexilis) .

Abundant. .

Sugar pine (Pinus lambertiana ) .

. . .do .

Niimermts.

Western white pine (Pinus monticola) .

. . .do .

Do.

White pine (Pinus strobus) .

Numerous. .

Do.

Pifion pine (Pinus edulis) .

None .

None.

Hard Pines.

Norway pine (Pinus resinosa) .

Jack pine (Pinus divaricata) .

Shortleaf pine (Pinus echinata) .

Spruce pine (Pinus glabra) .

Lodgepole pine (Pinus murrayana) ....

Longleaf pine (Pinus palustris) .

Western yellow pine (Pinus ponderosa)

Pitch pine (Pinus rigida) .

Loblolly pine (Pinus taeda) .

Scrub pine (Pinus virginiana) .

Table-mountain pine (Pinus pungens).

Numerous. .

Numerous,

None .

None.

. . . do .

Do.

. . .do .

Do.

. . .do .

Do.

...do .

Do.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

... do .

Do.

Other Conifers.

Tamarack (Larix laricina) .

Western larch (Larix occidental is) .

European larch (Larix larix) .

White spruce (Picea canadensis) .

Engelmann spruce (Picea engelmanni) .

Black spruce (Picea mariana) .

Red spruce (Picea rubens) .

Sitka spruce (Picea sitckensis) . .

Douglas fir (Pseudotsuga taxifolia) .

Balsam fir (Abies balsamea) .

White fir (Abies concolor) .

Lowland fir (Abies grandis) .

Alpine fir (Abies lasiocarpa) .

Red fir (Abies magnifica) .

Noble fir (Abies nobilis) .

Port Orford cedar (Chamaecyparis lawsonia)
Yellow cedar (Chamaecypdris nootkatensis) . ,
California juniper (Juniperus californica) . .

None .

None.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

...do .

Do.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

...do.......

Do.

...do .

Do.

. . .do .

Do.

. . .do .

Do.

. . .do .

Do.

Mar. 25, 1914

Tyloses in American Woods

461

TablH III. — Occurrence of true tyloses in native conifers — Continued.
Other Conifers — Continued.

Species.

Number
of speci¬
mens.

Sapwood.

Heartwood.

Western juniper ( Juniperus occidentalis) .

None .

None.

Red cedar (Juniperus virginiana) .

. . .do .

Bo.

Incense cedar (Libocedrus decurrens ) .

. . .do .

Do.

Redwood (Sequoia sempervirens ) .

. . .do .

Do.

Bigtree (Sequoia washingtoniana) .

...do .

Do.

Bald cypress (Taxodium distichum) .

. . .do .

Do.

Yew (Taxus brevifolid) .

. . .do .

Do.

Arborvitse (Thuja occidentalism .

. . .do .

Do.

Western red cedar (Thuja plicata) .

. . .do .

Do.

Eastern hemlock (Tsuga canadensis ) .

. . .do .

Do.

Western hemlock (Tsuga heterphylla) .

. . .do .

Do.

Black hemlock (Tsuga mertensiana) .

. . .do .

Do.

TYLOSELIKE CELLS

The tyloselike epithelial cells which surround the resin canals were
also carefully studied in Pinus, Larix, Picea, and Pseudotsuga. In these
woods both the horizontal and vertical resin canals often contained dis¬
tended cells which partly or sometimes completely filled the canal
openings. (PI. LVII, fig. 2; and PI. LVIII, figs. 2, 5, and 6.) This
closed condition of the vertical canals is particularly noticeable near the
medullary rays. (PI. LVI, fig. 1 ; and PI. LVII, fig. 2.) The distended
closing cells correspond to the plasma-containing cells described on page
446. (PI. LVIII, figs. 2 and 5.) A large number of the canals were,
however, entirely open.

In pines where many of the epithelial cells remain capable of growth,
three types of conditions may be found in the canals.

(1) The canals of the sapwood, especially of the outermost ring, may
not have yet opened — that is, the space which the canal will occupy may
still be filled by the parenchyma cells which later form the epithelium.
(PI. LVI, fig. 1.)

(2) Many canals may be partly open. (PI. LVII, fig. 1.) Frequently
the cells surrounding the opening are somewhat contracted and col¬
lapsed; or, again, individual cells containing plasma may become dis¬
tended, bow out into the open lumen of the canal, and thus assist in
partially closing it.

(3) Canals in the heartwood as well as in the outer rings of the sap-
wood may be completely closed.1 This may come about in two ways:
First, the groups of parenchyma cells observed in the sapwood may

287360— 14 - 2

1 Compare Thomson, R. B.

462

Journal of Agricultural Research

Vol. I, No. 6

never have split apart to form a canal opening. This was demonstrated
by the writer by means of serial sections following the course of a num¬
ber of horizontal resin canals from the bark into the heartwood. Second,
the canals once open may be closed completely by the growth of certain
of the epithelial cells, as before explained. This closing is not produced
by the equal action of all the cells which first split apart to form the
canal, but only by the later growth of certain of these which possessed
plasma and the growth potential for a longer period than their neighbors.
(PI. LVIII, fig. 5.) 1

PRACTICAL SIGNIFICANCE OF TYLOSES
TYLOSES AS A NATURAL “FILLER”

A good instance of the part played by tyloses in the structure of wood
is in the case of red oak and white oak. These two species have prac¬
tically the same structure, yet the red oak can not be used for tight
cooperage stock because the vessels are open tubes through which air or
liquid can escape. (PI. LIV, middle.) In white oak the vessels are
completely closed by tyloses, as shown in Plate LIU, figures 1 and 2,
or Plate LIV, R3.

In cabinetmaker's parlance, tyloses behave to some extent like a
natural “filler.” On a radial-cut surface the large vessels in the spring
wood of a red oak appear like hollow grooves, while those in the white
oaks are partly filled by the network of the tylosal cells which catch and
hold paint, for example. (PI. LII, fig. 1 ; and PI. LIU, fig- 2.)

TYLOSES A FACTOR IN DURABILITY

It is of interest to note the presence of tyloses (or sometimes of gums)
in the large vessels of those hardwoods which are particularly valued for
their durability. Many factors, such as the chemical composition of
the wood, its rate of growth, and hardness, are, of course, important in
determining durability, but the effect of tyloses should not be disregarded.
Moreover the vigorous growth of parenchyma, which in some cases
manifests itself by causing tylose formation and in others by producing
tannins, essential oils, etc., appears to be a fundamental characteristic
of naturally durable woods. White oak, in which tyloses are abundant,
is, for example, more durable than red oak, in which they are almost
wholly absent. The tylose walls present an added obstruction to the
advance of fungous hyphse and tend to make the vessels impenetrable
to air and water. They are especially effective in woods that have been
dried.

Although sapwood contains tyloses, it is usually less durable than
heartwood. The latter fact, however, holds true also for woods without
tyloses and can probably be explained by the condition of such materials

1 The illustrations reproduced in PI. LVIII of all conditions of open and closed horizontal resin canals
were taken from sapwood material.

Mar. 25, 1914

Tyloses in American Woods

463

in the sapwood as starches, which undergo a transformation when the
heart wood is formed.

The following tabulation of the “ Relative durability of hardwoods,”
compiled from the results of experiments, indicate that tyloses are a
factor in durability. The more durable species will be found, with a
few exceptions, to contain many or very abundantly developed tyloses.
(See Tables I and II.)

RELATIVE DURABILITY OP HARDWOODS 1

Durable .

Black locust.

Catalpa.

Osage orange.
Mulberry.

Chestnut.

Black walnut.

Live oak.

Sassafras.

White oak.

Post oak.

Black ash.

Honey locust.

Cherry.
Persimmon.
Slippery elm.
Bur oak.

Fairly durable .

Yellow poplar.

Red ash.

Red oak.

Scarlet oak.

Butternut.

Not durable.

Cottonwood.

White elm.

Red gum.

Hard maple.

White ash.

Black oak.

Red birch.

Beech.

Hickory.

Cucumber.

Black gum.

Watergum.

Basswood.

Buckeye.

Sycamore.

Gray birch.
Paper birch.
Aspen.
Willow.

The results of tests on 30,160 fence posts1 2 indicated the following
untreated hardwoods, in order of their durability, as the most suitable:
Osage orange, locust, mulberry, catalpa, certain oak (species not given),
and black walnut. The length of life in service varied from 10 to 50
years.

Some observations 3 on the life of untreated hardwood railroad ties
further confirm the relation between tyloses and durability. It must
be borne in mind, however, that for this type of service hardness has
been considered in judging durability. The list of woods, together
with their life in years under traffic, is as follows :

Species. Years of service.

Butternut . 4 Few.

Beech . . Do.

Black, red, or yellow oak . 4 to 5

Post oak . 6 to 8

Sassafras . 6 to 8

Chestnut oak . 9

Bur oak . 9

Species. Years of service.

Black walnut . 9

Chestnut . 5 to 10

Hickory . 7 to 10

Black locust . 7 to 10

White oak . 5 to 12

Mulberry . 5 Many.

Catalpa . Do.

1 This list is offered to show the comparative durability of some American timbers. It is not presumed
to obtain for all conditions.

* Crumley, J. J. The relative durability of post timbers. Ohio Agr. Expt. Sta. Bui. 219, p. 605-640,
10 pi. 1910.

8 Tratman, E. E. R. Report on the use of metal railroad ties and on preservative processes and metal
tie-plates for wooden ties. U. S. Dept. Agr., Div. For. Bui. 9, p. 216. 1894.

4 Life not given.

6 Little used.

464

Journal of Agricultural Research

Vol. I, No. 6

A few exceptions are noticeable. Chestnut oak, for' example, has
very few tyloses, but is hard and strong. Butternut has many tyloses,
but it is also much softer than the oaks. Hickory has many tyloses and
is here considered as durable a wood as black walnut. This is contrary
to observations of its durability by other investigators. The kind of
beech used is not specified, but if it was “white-heart” beech tyloses
were not present. The “ red-heart" beech, which contains tyloses, is
often reported as a very durable wood.

The following recent estimates are based on experience and actual
inspection by the Forest- Products Laboratory of woods in service (Table
IV):

Table IV. — Life of untreated wood placed subject to decay.

Untreated material.

Years.

Untreated material.

Years.

Tyloses abundant or many ; well

Tyloses lacking or scattered; few

developed.

or weakly developed — Contd.

Lumber:

Lumber — Continued .

Chestnut .

Maple .

White oak .

Birch .

Posts:

Poplar .

Locust .

2 c

Cottonwood .

Osage orange .

Tupelo.

Mulberry .

Basswood

Catalpa .

T A

* White-heart beech

Chestnut .

Red gtim ....

White oak .

Sycamore .

Ties:

Posts:

Black locust .

Red oak .

WThite oak .

Ash .

Chestnut .

Aspen ... .

Gum .

Tyloses lacking or scattered; few

Ties:

or weakly developed.

White-heart beech .

Birch .

Lumber:

Maple .

Elm .

Red oak

Ash .

Gum .

TYLOSES A FACTOR IN CREOSOTE PENETRATION
EXPERIMENTS WITH HARDWOODS

The study of the effect of structure on the penetration of artificial
preservatives, such as creosote, is a separate problem. Preliminary work
has shown some interesting results concerning the treatment of certain
tylose-filled hardwoods. A piece of air-dry black locust ( Robinia pseuda -
cacia ), 9 by by 1 inch, was subjected to a thorough treatment with

creosote in a treating cylinder. The piece contained sapwood and heart-
wood, the vessels of both of which were filled with tyloses. The stick
when split open after treatment showed no penetration except a faint
discoloration in the outer one-fourth inch of sap, which apparently did not
extend to the tyloses filling the vessels, but was located only in a few
scattered groups of fibers. The failure of the wood to absorb creosote

Mar. 25, 1914

Tyloses in American Woods

465

was not entirely due to the presence of tyloses, but the fact that the
creosote did not penetrate the tylose-filled vessels is significant.

In a piece of desert willow ( Chilopsis linearis) , 4 by 1% by 2 inches,
treated with carbolineum, no penetration was visible in the heartwood
except about one thirty-second of an inch near the surface. In the
sapwood, however, where, as shown in Table I, the large vessels of the
two outer growth rings are without tyloses, the dark discoloration of the
preservative was clearly visible following the lines of these open vessels.

Sapwood in general absorbs creosote much more easily than heartwood.
The supposed absence of tyloses in this region of the tree has previously
been regarded as one reason for this fact. As soon, therefore, as it was
satisfactorily determined that tyloses were unmistakably present in the
sapwood, special experiments were undertaken to discover what effect
they had on the absorption of the creosote. A piece of white oak was
given a commercial treatment at the same time and under the same con¬
ditions as the black locust. The sapwood absorbed the oil fully, but the
penetration stopped abruptly at the line of color demarkation between
the sapwood and heartwood. (PI. LIX, fig. 2, B.) To the eye the heart-
wood, except for a surface coating, was absolutely untreated. The ves¬
sels in both the sapwood and heartwood of this piece were filled with
strongly developed tyloses. Microscopic examination showed that the
tyloses in the vessels of the treated sapwood were entirely uncolored and
exactly like those in the vessels of the heart which was untreated through¬
out. The tyloses had then effectually kept the creosote out of the ves¬
sels, although there had been a full treatment of the wood fibers of the
sapwood. This shows that a considerable quantity of the preservative
was absorbed in spite of the fact that the presence of tyloses kept the
creosote out of the vessels. Hence, tyloses of themselves need not be
regarded as preventing the possibility of treating this species, at least in
the sapwood.

A piece of oven-dried hickory, 2 K by 2 by 14 inches, made up of both
heartwood and sapwood, was treated at the same time and under the
same conditions as the oak and locust, and showed a thoroughly good
penetration throughout. (PI. LIX, fig. 2 ,C.) Nevertheless, when the
wood was split, the tyloses, which were abundantly developed in the
vessels of both the sapwood and heartwood, were white and unstained by
the creosote, showing a marked contrast to the dark-brown fibers of the
surrounding treated wood. (PI. LII, fig. 1.)

The preliminary observations just described concerning the penetra¬
tion of creosote were based on results of treatments made on single speci¬
mens of the species studied and were regarded rather as valuable indica¬
tions than as conclusive evidence. To check them with other results,
the treatments with creosote were repeated on other specimens of the
woods previously used and more specimens of another species con¬
taining many tyloses. First, a piece of hickory taken from miscella¬
neous material was given a high-pressure treatment with creosote.

466

Journal of Agricultural Research

Vol. I, No. 6

A good absorption was obtained in both the sapwood and heartwood.
Nevertheless, the tyloses, which were everywhere well developed and un¬
damaged in the large vessels of both regions, remained colorless and
untreated. In addition, two other blocks of hickory from material col¬
lected with special care were also given pressure treatments in the cylin¬
der. These specimens were from pignut hickory, Hicoria glabra , and
mockernut hickory, Hicoria alba . Both specimens contained sapwood
and heartwood, with tyloses strongly developed in the large vessels.
Again, the wood was thoroughly treated with creosote in both the sap-
wood and the heartwood, and once more the tyloses could be observed
on a split surface to be quite uncolored and visible even to the naked
eye through their marked contrast with the blackish brown of the treated
wood. (PI. LII, fig. i.)

Thus, results from four specimens of hickory from different sources
clearly showed that in spite of the presence of tyloses a high absorption
of creosote may be obtained in the wood substance outside of the vessels
and the tyloses filling them.

The other species used in these experiments was the so-called red-heart
beech, a form of Fagus atropunicea. This had white tylose-free sapwood,
but a reddish heartwood with many tyloses. It was treated in the cylin¬
der at the same time as some of the hickories. The sapwood was thor¬
oughly penetrated, but the heartwood remained untreated except for a
surface coating and a very slight infiltration near the ends.

Lastly, a second piece of white oak was treated, as a check on the
piece treated previously. After the creosote treatment, which was
given at the same time as that of the hickories and beech, the sapwood
was found to be penetrated, and, as before, the heartwood was unpene¬
trated. Careful examination showed, however, that the discoloration
of the creosote extended down the large vessels of the sapwood and into
the tyloses which they contained. This apparent contradiction of
previous observations was explained when the material was examined
under the microscope. The tyloses were found to be full of fungous
mycelium and riddled with holes produced by the hyphse in passing
through the tylose walls. Under these circumstances, even when abun¬
dant tyloses are present, it is clear that some penetration may be secured
in the vessels.

The marked difference to be observed in the penetrance of creosote in
treatments of red oak and white oak is, however, chiefly the result of the
presence or absence of tyloses. The unobstructed vessels of red oak
give such open channels and offer so much additional surface for absorp¬
tion through their walls that the penetrability of the other elements
lying between the vessels is of relatively little importance. In white
oak, on the other hand, it is only the elements of structure other than
the large vessels that are available for penetration. The type of pene¬
trance obtained in red oak is shown in Plate LIX, figure 2, A. The dark
streaks mark the course of the creosote, which passed almost entirely

Mar. 25, 1914

467

Tyloses in American Woods

through the open vessels. The practical effect of this is evident in the
results obtained in penetrance treatments. It is possible to force creo¬
sote for long distances through red oak just as it would be possible to
force it through similar distances in small open pipe lines. In com¬
parison with this, the distance the oil will pass through white oak is very
short, since it has to penetrate through many cell walls, and the resistance
of the material must be overcome by high pressures.

Thus, although tyloses have a distinct effect, they are not the only
factor in the penetrance of wood. The characteristics of the other
elements in the annual ring must be considered. However, in the cases
examined, wherever the large vessels contained abundantly developed
tyloses or filling cells, the vessels and the tyloses, but not necessarily
the rest of the woody tissues, were impenetrable to creosote.

OBSERVATIONS ON CONIFERS

The presence of resin canals and their condition — that is, whether they
are open or partly or entirely closed by cells — considered in conjunction
with the general permeability of the tracheids, is a factor of practical
significance in the selection of wood for creosoting. (Pis. LVI and LVII.)
The number of the resin canals is very small in comparison with the
number of tracheids. However, if the canals are unobstructed, pene¬
trance is easily obtained for considerable distances through their cavities.
In a wood whose tracheids are penetrated with difficulty, the creosote
does not spread to any great extent from the canals into the tracheids,
even when the former are full. Nevertheless, the presence of creosote
or other toxic liquid in the resin-canal regions, which are among the first
affected by fungous infection, is of considerable assistance in prolonging
the life of the wood. Many of the resin canals, especially the vertical
canals in both the sap wood and the heartwood of the pines, are not com¬
pletely closed (PI. TVII, fig. 1, and PI. TVIII, figs. 1 and 4) and can
for this reason be penetrated. The effect of the presence or absence
of tylose-like cells in the resin canals, while a minor factor, is significant
in connection with the treatment of poles, ties, and paving blocks.

EFFECT OF TYEOSES ON THE WATER-EOGGING OF WOOD

In order to test the effect of tyloses on the water-logging of wood, some
roughly comparable air-dry blocks of several species were placed in a tank
of water and the length of time required to water-log each block suffi¬
ciently to sink it was noted. The blocks were grouped with reference to
their specific gravity (dry)1 and their actual weight. The woods in which
tyloses were few or wholly lacking invariably sank before those contain¬
ing abundant tyloses. Chestnut oak sank before white oak and bur oak,
persimmon before osage orange, flowering dogwood before hickory, yel¬
low poplar and aspen before catalpa, and blue beech and honey locust

1 Sargent, C. S. Report on the Forests of North America ... 612 p., maps. Washington, 1884.
(U. S. froth Census Reports, v. 9].)

468

Journal of Agricultural Research

Vol. I, No. 6

before black locust. The dogwood and persimmon sank in about 18
hours, while the catalpa floated for 20 days, and one piece of black locust
with a large percentage of heartwood remained floating for 46 days.

SUMMARY

The 143 specimens of hardwoods examined included 45 genera (94
species), of which 24 contained tyloses. The 60 specimens of conifers
examined included 13 genera (45 species), of which 1 contained tyloses.
Of the 139 species examined, 56, belonging to 25 genera, contaired
tyloses.

Tyloses were found in the sapwood of all species in which they occurred
in the heartwood.

Well-developed tyloses were found in the outermost rings near the
bark of 30 species of hardwoods.

True tyloses occur in the wood tracheids of certain pines, principally
of the white-pine group.

Epithelial cells sometimes effect a partial or even complete tyloselike
closing of the resin canals in Pinus, Larix, Picea, and Pseudotsuga.

A considerable proportion of the vertical canals, even in the heart-
wood of the pines, are fully or partly open.

Tyloses act like a natural filler in the hardwoods.

The woods in which tyloses are abundant as a rule are durable.

Tyloses, because they are very impermeable to air, water, and creosote,
reduce the penetrance of the woods in which they are strongly developed.
The presence of tyloses in the vessels of a hardwood, however, does not
prevent the penetrance of creosote into the other wood elements.

LITERATURE CITED

Bailey, I. W.

1913. Preservative treatment of wood. In Forestry Quart., v. 11, no. 1, p.
5-20, 2 pi.

Bary, Anton de.

1884. Comparative Anatomy of the Vegetative Organs of the Phanerogams
and Ferns. Translated by F. O. Bower and D. H. Scott ... p. 170.
Oxford.

Boehm, Josee.

1867. Ueber Function und Genesis der Zellen in den Gefassen des Holzes. In
Sitzungsber. K. Akad. Wiss. [Vienna], Math. Naturw. Cl., Abt. 2,
Bd. 55, p. 851-866, 2 pi.

1877. Ueber den aufsteigenden Saftstrom und den Abschluss lebender Zellen
gegen aussere Einwirkungen. In Bot. Ztg., Jahrg. 35, No. 7, p. 112-113.
1879. Ueber die Function der vegetabilischen Gefasse. In Bot. Ztg., Jahrg.
37, No. 15, p. 225-239; No. 16, p. 241-258.

Chrysler, M. A.

1908. Tyloses in tracheids of conifers. In New Phytol., v. 7, no. 8, p. 198-204,
ph 5-

Haberlandt, G. F. J.

1884. Physiologische Pflanzenanatomie. p. 217. Leipzig.

1887. Ueber die Beziehungen zwischen Function und Lage des Zellkemes bei
den Pflanzen. p. 71-74. Jena.

Mar. 25, 1914

Tyloses in American Woods

469

Hanausek, T. F.

1907. Microscopy of Technical Products. Translated by A. L. Winton and
Kate G. Barber, p. 200. New York.

Kcister, Ernst.

1903. Pathologische Pflanzenanatomie. 312 p., illus. Jena.

Mayr, Heinrich.

1883. Uber die Vertheilung des Harzes in unseren wichtigsten Nadelholzbau-

men. In Flora, Jahrg. 66 (n. R. Jahrg. 41), No. 14, p. 223.

1884. Enstehung und Vertheilung der Secretions-Organe der Fichte und

Larche, In Bot. Centbl., Bd. 20, No. 8, p. 246-253; No. 9, p. 278—
283; No. 10, p. 308-310, pl. 1-3.

1893. Das Harz der deutschen Nadelwaldbaume. In Ztschr. Forst u. Jagdw.,
Bd. 25, p. 313-324, 389-417, 565-593, 654-670, pl. 1-2. Reprinted as
Das Harz des Nadelholzer ... 1894.

Molisch, Hans.

1888. Zur Kenntniss der Thy lien, nebst Beobachtungen iiber Wundheilung
in der Pflanze. In Sitzungsber. K. Akad. Wiss. [Vienna], Math.
Naturw. Cl., Abt. 1, Bd. 97, Heft 6, p. 264-298, 2 pl.

Penhallow, D. P.

1907. Manual of the North American Gymnosperms ... 374 p., illus. Boston.
PraEl, Edmund.

1888. Vergleichende Untersuchungen fiber Schutz- und Kern-Holz der Laub-
baume. In Jahrb. Wiss. Bot., Bd. 19, p. 1-81, pl. 1.

Raatz, Wilhelm.

1892. Ueber Thyllenbildungen in den Tracheiden der Coniferenholzer. In Ber.
Deut. Bot. Gesell., Bd. 10, p. 183-192.

Reess, Max

1868. Zur Kritik der Bohm ’schen Ansicht fiber die Entwickelimgsgeschichte
und Function der Thyllen. In Bot. Ztg., Jahrg. 26, No. 1, p. 1-11,
pl. 1.

1896. Lehrbuch der Botanik. p. 88. Stuttgart.

Russow, Edmund.

1872. Vergleichende Untersuchungen ... der Leitbiindel-Kryptogamen, mit
Berficksichtigtmg der Histologie der Phanerogamen ... 207 p., 11 pl.
St. P6tersbourg. (M6m. Acad. Imp. Sci. St.-P6tersb., s. 7, t. 19, no. r.)
1883. Zur Kenntniss des Holzes, insonderheit des Coniferenholzes. In Bot.
Centbl., Bd. 13, No. 4, p. 134-144; No. 5, p. 166-173, pl- 1-5.

Sachs, Julius.

1887. Lectures on the Physiology of Plants. Translated by H. M. Ward. p. 581.
Oxford.

Strasburger, Eduard.

1891. Ueber den Bau und die Verrichtungen der Leitungsbahnen in den Pflanzen.

p. 191. Jena. (His Histologische Beitrage, Heft 3.)

1902. Das botanische Practicum ... Aufl. 4, p. 249. Jena.

■ - , Schenck, Heinrich, Noll, Fritz, and Karsten, George.

1908. Text- Book of Botany, ed. 3, rev. with German ed. 8, 746 p., illus. Lon¬

don.

Thomson, R. B.

1913. On the comparative anatomy and affinities of the Araucarineae. In Phil.
Trans. Roy. Soc. London, s. B, v. 204, p. 1-50, pl. 1-7.

Winckler, Hans.

1905. Ueber einen neuen Thyllentypus nebst Bemerkungen fiber die Ursachen
der Thyllenbildung. In Ann. Jard. Bot. Buitenzirg, v. 20 (s. 2, v. 5),
pt. 1, p. 19-37.

PLATE LII

Fig. 2. — Split radial face of a creosoted hickory block, showing tyloses (T) in a large
vessel. Magnified 12 diameters. Tyloses tmcolored; remaining wood substance
black with creosote.

Fig. 2. — Tangential section of Aesculus ociandra , yellow buckeye X 680, showing
two tyloses (T) which have grown out of one medullary-ray parenchyma cell (MR).
Shows open connection between the tyloses and parenchyma cell.

Fig. 3. — Cross section of valley oak, a white oak, showing young tyloses ( T ) next
the bark ( B ) in vessels (F).

(470)

» « V V/ -
;*;;;•!« ,

| ;i H

;:■ : ■

1 t ' mm

hi . | ; K

, . t\ m

[J^|i

1 _ |

American

Plate Llll

Agricultural Research

ol. |, No. 6

PLATE LUI

Fig. i. — Cross section of a white oak, showing fully developed tyloses ( T ) in the
large vessels (F).

Fig. 2. — Radial-longitudinal view, quarter-sawed surface, of the white oak shown
in figure i, showing complete closing of the vessel (V), which makes this wood
valuable in light cooperage, etc.

Fig. 3. — Cross section of sap wood of pignut hickory, showing fully developed
tyloses (T).

Fig. 4. — Radial view of mesquite, showing “gum” droplets ( G ) and formations
often stimulating tyloses.

PLATE LIV

Cross section of cow oak, a white oak, showing normal and abnormal tyloses. From
top to bottom are bark (B) and three annual growth rings (Ri, i?2, #3).

Fig. 1. — Wound tyloses ( WT ) induced by the felling of the tree and the sudden
cessation of sap flow.

Fig. 2. — No tyloses ( V)\ empty vessels. Normal tyloses not yet developed.

Fig. 3. — Young ( YT) and well-developed normal tyloses ( T ).

Plate LIV

Tyloses

>f Agricultural Research

PLATE LV

Fig. i. — Cross section of a diffuse porous wood, yellow poplar or tulip, showing
scattered tyloses X 50. Tt tylose-filled vessels; V, empty vessels.

Fig. 2.— Cross section of a ring porous wood, osage orange, with vasicentric paren¬
chyma, showing abundantly developed tyloses (T) X 50.

PLATE LVI

Fig. i. — Cross section of western white pine, showing ray tyloses (T), closed ver¬
tical resin canal ( VRC ) in young sapwood, and nuclei ( N ) visible in epithelial cells
of canal which is beginning to split open at S.

Fig. 2. — Tangential section of Norway pine, showing ray tyloses (T).

«»*»»»*’ »v |!

Tyloses

jricultural Research

mm _ _ _ _

• m m mm m —T—. — _ ~

A* J*V*VAVm • •

I* .••*>• •••»«*• ••••*»••••(

I- «*' - m **+'*■ «*

man

’ ■» •XSSBHSBP * f ? v v I

:!ir»iisf mmssSSl

l&sai&fe} ^&xSSi

-'?v5r^u

V*Vii»"<iii « ■ • •»*iii/*,.,1'SSSSS

iVt «««■■•'

PLATE LVII

Fig. i. — Cross-section view of shortleaf pine, showing open and partly closed ver¬
tical resin canals ( VRC ). These are typical of many canals in pine heartwood.
Shows thin- walled epithelial cells ( E ).

Fig. 2. — Heartwood of Sitka spruce, showing closed vertical canal (VRC).

PLATE LVIII

Open and closed horizontal canals in sap wood.

Fig. i. — Open canal in tamarack ( TE ) thick- walled epithelium.

Fig. 2. — Partly closed canal with distended epithelial cells ( DE ) in Douglas fir.

Fig. 3. — Young canal which has never opened in western white pine. Cells with
protoplasm and nuclei. Vertical canal ( VivC) in same condition on right; this is
longitudinal view of same canal as is shown in cross section, Plate LVI, figure 1.

Fig. 4. — Open canal in red spruce surrounded by thick-walled epithelium (TE).

Fig. 5. — Partly closed canal in red spruce. TE, thick-walled, and DE, thin-
walled distended epithelial cells.

Fig. 6. — Closed canal in Engelmann spruce. From old sapwood. The epithelial
cell has completely closed the canal and its wall has become thickened.

Woods

Plate LIX

PLATE LIX

Fig. x. — Log from collection of woods in the Forest-Products Laboratory — a speci¬
men of the material used in this study; 5, sapwood; H, heartwood.

Fig. 2. — Specimens of woods showing creosote penetrance in sap and heartwood
as affected by tyloses. The three specimens each contain both sapwood and heart-
wood. Specimen A. — Red oak. Has no tyloses; creosote passed chiefly down the
large vessels; note black streaks. Wood substance between vessels little treated;
note white streaks. Specimen B. — White oak. Has abundant tyloses in sap and
heartwood. Creosote penetrated the sapwood only. Thorough absorption obtained
in the sapwood substance between the impenetrable, tylose-filled vessels. Speci¬
men C. — Pignut hickory. Has abundant tyloses in sap and heartwood. Creosote
penetrated both. Good absorption throughout in the wood substance between the
tylose-filled vessels. Compare Plate LI I, figure i, an enlarged view of a portion of
this block.

28736°— 14 - 3

THE CAMBIUM MINER IN RIVER BIRCH

By Charles T. Greene,

Entomological Assistant, Forest-Insect Investigations,

Bureau of Entomology

The species of the family Agromyzidae generally mine in the leaves and
stems of various plants, while some mine in their roots. The species pre¬
sented in this paper, Agromyza pruinosa Coq.,1 is quite out of the ordinary
in that it mines in the cambium of the living tree, the mine leaving a scar
known as a “ pith-ray fleck. ” 2 These flecks in the various kinds of wood
have been known for many years to be the result of the work of insects,
and extensive investigations have been carried on in Europe as well as in
this country in order to determine the species causing the damage. In¬
vestigations in Europe have proved that at least the pith-ray fleck in
birch may be accredited to Agromyza carbonaria ,3 which is closely related
to the American species. The pith-ray flecks in birch in America have
been studied carefully, and it has been decided that Agromyza pruinosa is
at least one of the insects that produce flecks and is possibly the only one.
A gromyza pruinosa taken from river birch has just been reared to maturity.
This is the first record in America of the production of flecks in birch by a
definitely known species. (PI. LX, fig. 2.)

SEASONAL HISTORY

During July and the early part of August, 1912, the work of this
dipterous larva was very common in river birch at the Chain Bridge, in
the District of Columbia, every tree that was examined containing new
work; but in 1913, in the same locality, only a few trees disclosed new
work. A dipterous larva and similar work were found frequently in
red maple ( Acer rubrum ), but not so commonly as in birch. In 1913
Mr. T. E. Snyder found in wild cherry (Prunus sp.) on the Virginia shore
of the Potomac River at the Chain Bridge two larvae which are identical
with the larvae of Agromyza pruinosa in the birch, except that they are
only two- thirds as long, although to all appearances full grown. The
work of this species in wild cherry is identical with that in red maple
and black birch, but the mines are correspondingly smaller.

1 Thanks are due to Mr. J. R. Malloch for assistance in determining the species.

* Brown, H. P. Pith-ray flecks in wood. U. S. Dept. Agr., Forest Serv., Circ. 215, 15 p., 6 pi. May 7,
1913.

3 Nielsen, J. C. Zoologische Studien iiber die Markflecke. Zool. Jahrb., Abt. System., Geogr. u. Biol.
Tiere, Bd. 23, Heft 6, p. 725-738, pi. 30. 1906.

(471)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Vol. I, No. 6
Mar. 25, 1914
K-s

472

Journal of Agricultural Research

Vol. I , No. 6

CHARACTER OF TREES ATTACKED

The trees attacked are apparently healthy, and infested ones can not
be detected by their outward appearance. The only way in which to
detect the larva is to remove the bark and expose the cambium, where
at a glance you can generally recognize the new galleries from the old
ones, since new larval mines are only faintly darker than the living
cambium; in fact, they are sometimes of a delicate pink color, whereas
all the old work is generally dark brown. In Vilas and Oneida Counties,
Wis., the trees in the vicinity of Tomahawk and Trout Lakes were care¬
fully examined by Mr. S. A. Rohwer last fall (1913), and no evidence of
the cambium miner was found in white birch (Betula populifolia) , red
oak (| Quercus rubra), red maple (Acer rubrum), or sugar maple (Acer
saccharum).

Pith-ray flecks were found in red oak (< Quercus rubra) at Charter Oak,
Pa., by Mr. T. E. Snyder and in mountain holly (Ilex monticola) at
Endeavor, Pa., by Mr. F. C. Craighead, but the particular insect or
insects causing them are not yet known.

LIFE HISTORY OF THE SPECIES
METHODS OF REARING

Numerous experiments were conducted while rearing this species. All
the breeding jars were placed in a pasteboard box, which was put in an
ordinary soap box lined and covered with about five thicknesses of news¬
paper. This box was kept outside during the winter in an inclosed shed.
The frost penetrated all the protective coverings, but not so thoroughly
as though the boxes had been completely exposed. Jars containing
earth and sand gave the best results in these rearing experiments. From
April 15 to May 12, 1913, six adults emerged. On May 1 a single adult
which was reared from the larva emerged, a hymenopterous parasite
emerging from another pupa case on May 13.

THE EGG

The writer unfortunately did not succeed in securing the egg of this
species, but it is apparently deposited in the fork of two branches which
are about 5 to 8 years old and near the top of the tree. From the shape
of the ovipositor (PI. LXI, fig. 4) the egg is more than likely deposited on
the outside of the bark, as the mine, which has been traced from a twig to
the base of the tree, a distance of 40 feet, starts from this point like a hair
line and, increasing in width as it goes down the trunk, reaches a width
of one-eighth of an inch at the base.

THE larva1

The larva (PI. LXI, fig. 1) is white, opaque, and cylindrical, averaging
from 20 to 25 mm. in length and 1 mm. in diameter. One larva, collected

1 The larva of this species was discovered by Mr. H. P. Brown and was first shown to the writer by Mr.
T. E. Snyder.

Mar. 25, 1914

Cambium Miner in River Birch

473

on June 19, 1913, was 30 mm. in length and 1 mm. in diameter. The
hooklet is shiny black and chitinized, the exposed portion being more
highly chitinized than the rest. The hooklet complete (cephalopharyn-
geal skeleton) dissected out is shown in Plate LXI, figure 1, a. Back of
the large hooklet are two smaller toothlike processes, one on each side,
the position of these being shown at b. The anterior spiracles at c and
the posterior pair at d are a very pale yellow, and their position is shown
in outline. At the caudal end of the larva are two padlike surfaces, very
faintly raised from the surface of the body, reaching nearly around the
circumference of the body and covered with numerous brown, hooklike
hairs or bristles. Several stages of the larvae were observed, and the
only noticeable difference was in their size.

If theJarva reaches the base of the tree before the time to pupate, it
will turn and mine up the cambium for some distance; on one occasion
the larva retreated for 6 feet, then returned, thus encircling the root, and
followed it for 2 feet from the trunk. The exit hole is sometimes made on
the side of the root, but generally it is on the underside, and the larva
pupates immediately on emergence. The pupae were found from one-
half to one inch from the exit hole. A portion of river birch ( Beiula
nigra) with the bark removed is shown in Plate LX, figure 1, to illustrate
the larval mines, while figure 2 is part of a cross section showing the
“pith-ray flecks” from above.

The only larva that was reared by the writer, and in fact the only one
that reached maturity, was placed in a large vial July 30, 1912, with a
piece of freshly cut river-birch bark, the inner surface of which was
covered freely with fresh sap. A piece of gauze was placed over the
opening of the vial. On August 6, 1912, at 8.30 a. m., the larva com¬
menced pupation, first becoming rigid and then changing to deep yellow
at both ends, while the central portion remained the natural white color.
It was 25 mm. in length and 1 mm. in diameter, but by noon it had
decreased to about 10 mm. in length and increased to 2 mm. in diameter.
Both ends had changed to dark brown and were perfectly formed, as in
the pupa, and the middle was a light yellowish. At 5 p. m. the pupa
was perfectly formed and dark brown all over, its dimensions now being
5 mm. in length and 2 mm. in diameter. The larva pupated under the
thin folds of the outer bark, as there was nothing else in the vial.

THE PUPA1

The pupa (PI. LXI, fig. 2) is of the usual cylindrical type and dark
reddish brown in color, averaging from 4 to 5 mm. in length by 2 mm.
in diameter, and is formed by the shrinking of the larval skin. The
anterior spiracles are slightly more prominent than the posterior pair.

The pupa of the species was discovered and first shown to the writer by Mr. T. E. Snyder.

474

Journal of Agricultural Research

Vol. I, No. 6

THE ADULT

The adult (PI. TXI, figs. 3 and 4) of Agromyza pruinosa Coq.,1 six
specimens of which were reared by the writer in the spring of 1913,
is closely related to Agromyza carbonaria Zett. of Europe. Agromyza
pruinosa remains in the pupal stage in the ground during the winter
and emerges from the pupa case in one of two ways : Either the end of
the pupal case is pushed off completely, or emergence is accomplished
by tearing the end of the pupal case into shreds. Of the six specimens
just referred to five were males and one a female. This species of
Agromyza is represented in the United States National Museum col¬
lection by Coquillett's type, a single male specimen (Catalogue No.
6659, U. S. National Museum). The writer's specimens agree perfectly
with the type, except that they are very slightly larger. '

The general appearance of the adult female corresponds to that of the
male, with the exception that it is slightly more robust. The ovi¬
positor is slightly over one-half of a millimeter in length, chitinized, and
somewhat shiny on the sides and edges of the dorsal surface. It is
slightly flattened and a little broader at the apex than at the base. On
the dorsal surface is a granular space, rounded toward the base of the
ovipositor.

The total length of the female is 4 mm., and of the male about 3 mm.
The abdomen of the female is shown in figure 4 of Plate LXI.

In an adult that had just emerged from the pupal case, the eyes were
brownish and the frons and face a pale yellow or orange color. The
thorax was pale gray, the legs yellowish, and the wings opaque white,
clearing to hyaline in about two hours. The abdomen was of a dull
orange color, with a faint gray line along the edge of each segment.
The whole insect assumed its natural color in two and a half hours.

A HYMENOPTEROUS PARASITE

On May 13, 1913, a hymenopterous parasite, Sympha agromyzae
Rohwer 2 (PI. LXI, fig. 5), issued from a pupa case of Agromyza pruinosa
Coq. This parasite is nearly as large as its host. Apparently it deposits
its egg within the egg of the host. The apparently normal dipterous larva
mines down the tree trunk and enters the ground; the pupa is perfectly
formed, outwardly exhibiting no signs of parasitism, but about the time
the host should emerge the parasite issues instead. At maturity the end
of the pupal case is pushed open by the parasite in the same manner as
the host would do it.

]Coquillett, D. W. New acalyptrate Diptera from North America. Jour. N. Y. Ent. Soc., v. 10, No.
4, p. 177-191. Dec., 1902. “ Agromyza pruinosa , sp. nov.,” p. 1S9.

2 " Sympha agromyzae , n, sp. Female. Length 3 mm. Notauli well defined; prescutum with afove-
olate furrow; face sparsely punctured; propodeum with a transverse carina; hind tarsi pale. Type Cat.
No. 16474 U. S. Nat. Mus." (S. A. Rohwer). A detailed description will appear later in the Entomo¬
logical News.

PLATE LX

Fig. i. — River birch with bark removed, showing larval mines of Agromyza
pruinosa .

Fig. 2. — Section through wood of river birch, showing “ pith-ray flecks” produced
by the work of Agromyza pruinosa .

Photographed by H. B. Kirk.

PLATE LXI

Fig. i. — Agromyza pruinosa: Larva and details.

Fig. 2. — Agromyza pruinosa: Pupa.

Fig. 3. — Agromyza pruinosa: Adult male.

Fig. 4. — Agromyza pruinosa: Abdomen of adult female, showing ovipositor.
Fig. 5. — Sympha agromyzae: Adult.

A STUDY OF SOME IMPERFECT FUNGI ISOLATED FROM
WHEAT, OAT, AND BARLEY PLANTS

By Edward C. Johnson,

Formerly Pathologist in Charge of Cereal-Disease Investigations ,

Bureau of Plant Industry

INTRODUCTION

Of the imperfect fungi, many are parasitic on cereals wherever climatic
conditions favor their development. They occur as scab on the heads,
as leaf spots, and as infections in the culms and roots. Usually one or
more species are present in the roots and culms of stunted plants, more
particularly where some one cereal crop has been grown year after year
on the same land. A study of the fungi occurring on wheat, oats, and
barley, with particular reference to their pathogenicity, is therefore of
much economic importance.

Such a study was begun in the cereal-disease laboratory of the Office
of Grain Investigations of the Department of Agriculture in 1910. Species
of imperfect fungi were isolated from wheat, oats, and barley obtained
from various parts of the country. Helminthosporiums, Altemarias,
Cladosporiums, and Fusariums were obtained. They were secured from
leaf spots or from the lower nodes, root crowns, or roots of more or less
stunted plants. In many cases they were obtained pure from fresh sporu-
lating material on leaves and stems. In other cases they were obtained
from the nodes, root crowns, and roots by sterilizing these parts exter¬
nally in a 1 to 1,000 solution of mercuric chlorid, washing them in several
changes of sterile water, and incubating them in moist chambers. After
incubation for three to five days at a temperature of 720 to 770 F., sporu-
lating myceliums were usually obtained. Plate cultures were then made
and the fungi present isolated in pure cultures and propagated. On corn-
meal agar, corn meal, and potato cylinders most of them grew and sporu-
lated profusely.

Pure cultures were obtained and grown and the identity determined
as Fusarium culmorum W. G. Sm., Helminthosporium gramineum Rabh.,
Cladosporium gramineum Cda., and a species of Altemaria. The deter¬
mination of Fusarium culmorum was made by Dr. H. W. Wollenweber,
of the Bureau of Plant Industry; the other determinations were made
by the writer. Helminthosporium gramineum was isolated from the
lower parts of the culms of stunted wheat plants growing on land con¬
tinuously cropped to wheat at the Minnesota Agricultural Experi¬
ment Station, and from wheat leaves and barley leaves at the same

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(47s)

Vol. I, No. 6
Mar. 25, 19 14
O-iS

476

Journal of Agricultural Research

Vol. I, No. 6

place. Cladosporium gramineum was obtained from the leaves of oats
at the same station. The Alteraaria species occurring with Helmintho-
sporium or independently were isolated from wheat culms in the same
manner as Helminfhosporium gramineum. Fusarium culmorum was iso¬
lated from wilted oat plants obtained from a io-acre field on the farm of
Mr. Peter Hanson, Sandy, Utah, on May io, 1910. On this farm about
10 acres of oats had been practically destroyed by disease a few weeks
after the seed was planted. The plants sent to the cereal-disease labora¬
tory for examination and diagnosis were sterilized by immersing them
in a 1 to 1 ,000 mercuric-chlorid solution for 10 minutes, followed by wash¬
ing in sterile water. They were then placed in a moist chamber at a
temperature of about 75 0 F. for several days and were soon covered with
a luxuriant fungous growth. This proved to be a pure culture of Fusa¬
rium culmorum . It was plated and grown on potato cylinders and corn
meal and sporulated abundantly.

After securing these fungi in pure cultures and inducing profuse
sporulation, tests were made as to their pathogenicity on the leaves,
seeds, and seedlings of wheat, oats, barley, and rye.

INOCULATION OF LEAVES OF WHEAT, OATS, BARLEY, AND RYE WITH
SPECIES OF IMPERFECT FUNGI
*

Seedling plants of wheat (Haynes Bluestem, Minn. No. 169), oats
(Early Gothland, Minn. No. 26), barley (Manchuria, Minn. No. 105), and
rye (winter) were grown in the greenhouse at Washington, D. C., in 6-
inch pots under temperature and moisture conditions as nearly normal
as possible. When the seedlings were 2 to 3 inches high, inoculations
were made about an inch from the leaf tip, with spores transferred from
pure cultures in test* tubes by means of a flattened inoculating needle,
care being taken that little or none of the nutrient medium was trans¬
ferred to the leaves. If any of the medium accompanied the spores,
control plants were similarly treated with the same medium minus the
spores. Care was taken not to injure the leaves in any way. The
inoculated plants were placed under bell jars standing in pans of sand
and water, thus permitting the moisture transpired to condense on the
leaves, making an ideal condition for spore germination. They were
allowed to remain under the bell jars for 48 hours and were then removed
and placed in the greenhouse at a temperature ranging from 550 to 65° F.
Table I shows the results of these inoculations.

Mar. 25, 1914

Imperfect Fungi from Wheat, Oats, and Barley

477

Table I. — Results of inoculating seedling leaves of wheat , barley , oats, and rye with
imperfect fungi obtained from cereals.

Test

No.

Species.

Origin.

Inocu¬
lated
on —

Date of
inocula¬
tion.

Length

incu¬

bation.

Num¬
ber of
inocu¬
la¬
tions.

Infection.

Control.

Num¬

ber.

Per

cent-

age.

Total

num¬

ber.

Num¬
ber in¬
fected.

1911.

Days.

Helminthospo-

Wheat node 1 ,

Wheat. .

Oct. 31

100

rium grami-

neum.

. do .

. . .do. . . .

Nov. 10

. do . .

. do .

Barley. .

Oct. 31

. do .

. . ,do .

Nov. 10

. do .

Oats. . ..

Oct. 31

. do .

. . .do .

Nov. 10

. do .

Rye .

. . .do. ... .

1912.

. do .

Barley leaf l...

Wheat. .

Jan. 24

. do .

Barley. .

. . .do .

. do .

Oats. . . .

. ..do .

100

. do .

Rye .

. . .do .

100

Cladosporium

Oat leaf 1 .

Wheat. .

. . .do .

gramineum.

. do .

. . .do .

. do .

Barley . .

. . .do .

. do .

. . .do .

. do .

Oats. . . .

. . .do .

. do .

. . .do .

. do .

Rye .

. . .do .

. do .

... do ....

. . .do .

Fusarium cul-

Oat seedling s.

Wheat. .

Mar. 19

morum.

. do .

Barley. .

. . .do .

.... .do .

. do .

Oats. . . .

. . .do .

. do .

Rye .

. . .do .

1 From University Farm, St. Paul, Minn. * From farm of Mr. Peter Hanson, Sandy, Utah.

Table I shows that the strains of Helminthosporinm gramineum from
both wheat and barley infected the leaves of wheat, barley, oats, and rye.
On wheat, barley, and rye the leaf spots at the point of inoculation became
distinct in a little less than three days after inoculation. These spots,
which had a dead central area surrounded by a brown margin, slowly
increased in size until their diameter was almost equal to the width of
the leaf. A tendency to striation of the leaf area contiguous to the
spots was noticed. No striking difference could be detected in the effect
of the fungus from wheat or barley, the strain from wheat attacking
barley and rye fully as severely as the strain from barley, and the strain
from barley attacking wheat and rye fully as severely as the strain from
wheat. On oats the two strains showed a slight difference in virulence,
the fungus isolated from the barley apparently showing greater vigor in
its attack than the fungus from wheat. In fact, three days after inocu¬
lation with the fungus from barley, oat leaves were so severely affected
that in many cases they were cut in two, the tip portion often breaking
off and falling to the ground. The two strains behaved so similarly,
however, that physiologically they undoubtedly may be regarded as iden¬
tical. Morphologically, no difference was detected.

478

Journal of Agricultural Research

Vol. I, No. 6

Table I also shows that Cladosporium gramineum and Fusarium cul-
morum did not form leaf spots, even though the number of inoculated
leaves was fairly large. This was rather unexpected in the case of
Cladosporium, as it was obtained in pure culture by plating direct from
a fresh mass of spores from a badly infected oat leaf in the field. Con¬
tinuous culture on artificial media apparently either reduced its virulence,
the temperature and moisture conditions in the greenhouse not being
such as were conducive to infection by this fungus, or infection took
place normally only after aphid injury or other wound. That Fusarium
culmorum did not produce leaf spot was to be expected, as it usually
does not occur in this manner and was not isolated from a leaf but from
a wilted plant.

INOCULATION OF SEED OF WHEAT, OATS, BARLEY, AND RYE WITH
SPECIES OF IMPERFECT FUNGI

Seed of wheat, oats, barley, and rye was inoculated with spores of the
same strains of imperfect fungi used in the seedling-leaf inoculation tests.
The fungi were grown in pure cultures in the same manner as those used
for the leaf inoculation work. When sporulating profusely, sterile water
was poured into the test tubes, the spore masses were loosened by the
use of platinum needles, and the contents were well shaken. The water
containing the spores was then poured off and diluted with sterile water
until a drop placed under the microscope was found to contain from 5
to 25 or more spores. Seed of wheat, barley, oats, and rye was sterilized
by immersion for one hour in a formalin solution consisting of 2.5 parts
of 40 per cent formaldehyde to 1 ,000 parts of water and was immediately
dried and inoculated with spores by soaking it in the water containing
them. The seed was then planted in 6-inch pots filled with a sandy loam
soil rich in humus and placed in the greenhouse at temperatures ranging
from 550 to 65° F. The soil used had been sterilized previously in a
steam sterilizer at a pressure of 15 pounds for two hours, the tempera¬
ture being approximately 265° F. Control seed which had been steril¬
ized but not inoculated was planted for comparison in every case. The
results from such inoculation and plantings in the greenhouse are shown
in Table II.

Mar. 25, 1914

Imperfect Fungi from Wheat , Oats , and Barley

479

Table II. — Results of inoculating seed of wheat , barley, and oats with imperfect fungi
. isolated from grain plants.

Inoculated seed.

Control seed.

Test

No.

Species.

Origin.

Inocu¬

lated

Date of
planting.

Num¬

ber

plant¬

ed.

Germi¬

nated.

Num¬

ber

plant¬

ed.

Germi¬

nated.

Num¬

ber.

Per¬

cent¬

age.

Num¬

ber.

Per¬

cent¬

age.

Helminth osporium

Wheat culm ....

Wheat.

1911.
Nov. 21

150

22. 6

75-5

gramineum.
do .

. do .

. . .do _

Dec. 2

34-8

112

83.0

. do .

. . .do....

. . .do .

23. 0

65-3

6-1

. do .

1912.
Jan. 19

105

60. 9

105

87.6

do .

. do .

1911.
Dec. 2

xi2

26. 7

112

83.0

do .

. do .

Barley.
Oats. . .

. . .do .

105

82.8

105

79*o

. do .

78.5

81.4

. do .

Barley leaf .

Wheat.

Nov. 19

105

S9-o

105

87.6

do .

. do .

Barley.

. . .do .

105

81. 9

105

79-0

. do .

Oats. . .

74. 2

81.4

Fusarium culmorum. .

Oat seedling. . . .

Wheat.

1912.
Mar. 1

22. 9

83-3

do .

. do .

Barley.

. . .do .

65. s

88.5

do .

. do .

Oats. . .

120

1* 7

85.0

ATternftrift Rp .

Wheat culm .

Wheat.

1911.
Nov. 21

*5°

108

61. 2

81. 1

Dec. 2

112

82.0

112

83.0

Wheat seedling.

. . .do..,.

1912.
Mar. s

98.7

93-7

1 8

.do .

. do .

Barley .

...do .

86.2

86.1

. do .

Oats. . .

. . .do .

85. s

87. 7

Cladosporium grami-

Oat leaf .

Wheat.

Feb. 28

112

105

93*7

92.8

neum.

. do .

Barley.
Oats. . .

. . .do .

87. 2

83-3

. do .

. . .do .

112

104

92. 8

85-7

Table II shows that the strains of Helminthosporium gramineum
isolated from wheat and barley were decidedly pathogenic to germi¬
nating wheat, only 22 to 60 per cent of the inoculated wheat in five
trials producing plants, while 65 to 87 per cent of the controls not inocu¬
lated produced sound plants. These results are shown further in Plate
LXII, figure 1. Barley and oats were not affected to any appreciable
degree so far as germination and sprouting were concerned, the inocu¬
lated seed producing as large a percentage of plants as the clean seed.
Those wheat plants which developed from inoculated seed were stunted
and not nearly so vigorous as those produced from clean seed. At the
end of six weeks the difference in height of plants from inoculated and
clean seed was very marked. The plants from seed inoculated with
H . gramineum from wheat were 5.5 inches high to the tip of the second
leaf and those from seed inoculated with H . gramineum from barley
4.88 inches high to the tip of the second leaf, while control plants grown
from clean seed averaged 6.45 inches high to the tip of the second leaf.

A similar difference was noticeable in barley plants grown from inocu¬
lated and clean seed, although the difference was not quite as marked

480

Journal of Agricultural Research

Vol. If No. 6

as in the wheat plants. This is shown in Plate LXII, figure 2. Barley
plants from seed inoculated with Helminthosporium gramineum from
barley were 5.82 inches high at the end of six weeks, those from seed
inoculated with H . gramineum from wheat were 6.34 inches high, and
those from clean seed, 6.46 inches high. The measurements are the
averages of 50 plants in each case. There was no measurable difference
in the height of oat plants grown from inoculated and from clean seed.
H. gramineum was easily reisolated in every trial both from stunted
wheat and stunted barley plants by external sterilization in mercuric-
chlorid solution and incubation at room temperature.

Fusarium culmorum was even more virulent than Helminthosporium
gramineum , particularly on oats. Inoculated wheat seed produced only
22.9 per cent of sound plants, barley seed 65.5 per cent, and oat seed
only 1.7 per cent, while the controls produced 83.3, 88.5, and 85 per
cent of sound plants, respectively. The results of the inoculations are
further strikingly shown in PI. LXII, figures 3, 4, and 5. The 10-acre oat
field where this fungus was secured had been practically destroyed by
some disease, and these results show that F. culmorum undoubtedly was
the causal organism.

The two strains of Alternaria sp., one isolated from wheat culms from
University Farm, St. Paul, Minn., the other from wilted wheat seedlings
from Vermont, had no pathogenic effect on wheat, oats, or barley, the
differences in percentage of germination from inoculated seed and control
seed being so slight as to be negligible. Cladosporium gramineum also
had very little if any effect on the seedlings, the percentage of germina¬
tion from inoculated seed being only slightly smaller than from control
seed.

To determine further how the Helminthosporium gramineum attacked
the seed and seedlings, a large number of seeds and seedling plants grown
from inoculated seed were dug and examined a few days after germina¬
tion. It was found that many of the seeds had been attacked by the
fungus so rapidly that they had not had an opportunity to germinate.
Many others had germinated, apparently became infected immediately,
and were killed before they were an inch high. Plants which survived
were severely affected, as shown by the brown discoloration at the base
of the culms, a condition not noticed in any of the controls. This dis¬
coloration usually occurred in the basal leaf sheath. When the plants
had grown for several weeks, it was also very noticeable in the root crown.
The discoloration was not as marked in barley grown from inoculated
seed as the discoloration in wheat and was entirely absent in oats.

Numerous seeds and seedlings inoculated with Fusarium culmorum were
also examined. Many seeds were found to have been killed before the
process of germination had proceeded sufficiently far for any roots to
form and before the plumule emerged from the ground. Eight days after

Mar. 25, 1914

Imperfect Fungi from Wheat , Oats , and Barley

481

planting, the whole seed often was permeated by the fungus, the contents
of the seed coats having a pink coloration. The plants which survived were
discolored at the base in a manner similar to those of plants from seed
inoculated with Helminthosporium gramineum. Where discolorations oc¬
curred, it was the first leaf sheath which was affected, while the central
stem or culm was normal in appearance and color. The vigor of the
plants from inoculated seed was markedly reduced, and they were shorter
than the normal plants during the six weeks in which they were grown.
This was true also of wheat and barley grown from seed inoculated with
this fungus.

COMPARATIVE ROOT DEVELOPMENT OF WHEAT PLANTS GROWN FROM

SEED INOCULATED WITH HELMINTHOSPORIUM GRAMINEUM AND

FROM CLEAN SEED

To determine the comparative development of the root systems of sur¬
viving plants from seed inoculated with Helminthosporium gramineum
and from clean seed, two pots of wheat containing five plants each, one
grown from inoculated and the other from clean seed, were removed
to the laboratory and the soil carefully washed away from the root sys¬
tems. The roots were spread out by floating them in water and then
drawing off the water. The difference in development of the root
systems of the two sets of plants was very marked. The roots of plants
from inoculated seed were discolored near the root crown. They were
also much shorter and much less vigorous than roots of plants from clean
seed; this is strikingly shown in Plate LXIII. Numerous other plants
were examined, and it was found that in practically every case where
inoculated seed had produced plants which survived, the root systems
were less vigorous than in plants grown from clean seed.

SOIL INFECTION WITH HELMINTHOSPORIUM

To determine whether or not soil in which seed inoculated with Hel¬
minthosporium gramineum had been planted would remain sufficiently
infected for any length of time to injure later plantings, inoculated seed
was planted in pots in the greenhouse at Washington, D. C., on November
21, 1 91 1 , and the resulting plants were grown for five weeks and then cut
off. Control pots were similarly planted with clean seed and the plants
removed after five weeks. These pots were again sown on January 13,
1912, with wheat which had been previously sterilized in a 2.5 to 1,000
formalin solution. Of 1 50 seeds planted in the soil in which wheat plants
had been grown from seed inoculated with H . gramineum , 104, or 69.3
per cent, germinated and produced plants, while of 90 seeds planted in
control pots 76, or 84.4 per cent, germinated. This indicates that the
soil remained infected during the two months in which the experiment
was in progress. How long soil remains infected in this way is one of
the important problems in plant pathology.

287360— 14 - 4

482

Journal of Agricultural Research

Vol. I, No. 6

FIELD EXPERIMENTS WITH SEED INOCULATED WITH IMPERFECT

FUNGI

In order to test whether the imperfect fungi which were found patho¬
genic in the greenhouse on seeds and seedlings would act similarly under
field conditions, field experiments were undertaken at University Farm,
St. Paul, Minn., in the spring of 1912.1 The two strains of Helmintho-
sporium gramineum and the one strain of Fusarium cidmorum which had
been found pathogenic in the greenhouse were tested in connection with
wheat, barley, and oat seed. The same varieties of grains which were
used in the experiments at Washington, D. C., were used in the field
experiments. The seed was treated in a formalin solution of 3 parts of
40 per cent formaldehyde to 1 ,000 parts of water for one hour and after¬
wards was inoculated exactly as in the greenhouse work already described.
Immediately after inoculation, the seed was planted in the field in rows
1 rod in length and 10 inches apart, with controls every alternate two
rows. The seeds were counted. After the grain had sprouted and the
plants were from 3 to 6 inches high, careful counts were made to deter¬
mine the percentage of germination and observation made of the vigor
of the plants during the first few weeks of growth. The results are given
in Table III.

Table III. — Results of inoculating seed of wheat } barley , and oats with imperfect fungi
isolated from grain plants , and of planting them in the field at University Farm,
St. Paul , Minn.

Test

No.

Species.

Origin.

Inocu¬
lated on —

Date of
planting.

Inoculated seed.

Control seed.

Num¬

ber

planted.

Germi¬

nated.

Num¬

ber

planted.

Germi¬

nated.

Num¬

ber.

Per

cent-

age.

Num¬

ber.

Per

cent-

age.

19x2.

Helminthospo r i -

Wheat culm.. .

Wheat. .

Apr. 27

160

12$

78.1

159

i5S

97-5

11m gramineum.

. do .

Barley . .

. . .do .

160

71.9

160

130

81.2

. do .

Oats ....

. . .do .

160

I45

90. 6

160

143

88. 7

. do .

Barley leaf. . . .

Wheat. .

160

55-6

160

122

76. 2

. do .

. do. .

Barley. .

. . .do .

160

109

61. 8

160

13 1

81. 9

. do .

Oats ....

. . .do .

160

121

75. 6

160

142

88.7

Fusarium culmo-

Oat seedling . .

Wheat. .

. . .do .

160

61. 2

160

130

81. 2

rum.

. .do .

. do .

Barley. .

. . .do .

160

XI4

71. 2

160

130

81. 2

. do .

Oats. . . .

. . .do .

160

60.0

160

140

87.5

The results given in Table III substantiate the results of the experi¬
ments in the greenhouse. Helminthosporium gramineum from wheat
when applied to the seed reduced the percentage of germination of both
wheat and barley, but not to the same extent as in the greenhouse tests.
Oats were not appreciably affected. The material used for inoculation

In these experiments the writer was assisted by Messrs. Alden A. Potter and John H. Parker.

Mar. aS, 1914 Imperfect Fungi from Wheat , Oats , and Barley

483

was not in a profusely sporulating stage and therefore not in as active a
condition as the material which was used in the inoculations in the green¬
house. The seed which was inoculated also was still slightly damp after
the treatment in the formalin solution and this trace of formalin might
have reduced the effectiveness of the spores to some extent. The strain
of H. gramineum from barley was more virulent than the strain from
wheat, the percentage of germination being less where this strain was
used for inoculation than where the strain from wheat was used. After
inoculating with this strain, even the germination of the oats was con¬
siderably affected. The material used for inoculation, however, was in
better condition than the material of the strain from wheat, as the fungus
was sporulating abundantly when used. The plants of both wheat and
oats which survived were less vigorous than the plants from clean seed,
being slightly smaller than the plants in the control rows.

Fusarium cultnorum also wasi virulent, particularly on oats, and its
effect on wheat and barley was marked. The wheat plants which sur¬
vived after inoculation with this fungus were smaller than those in the
control rows, the difference being measurable. Several of the plants
were dying when counted. In the case of barley the difference in the
plants from inoculated seed and control seed was not marked, while in
the case of oats many plants from the inoculated seed were very weak
when counted, the difference in vigor between them and plants from
clean seed being very noticeable. There was a sufficient difference in
stand between rows from inoculated and from clean seed in the case of
wheat, oats, and barley to be noticeable even without counting the plants.

That the reduction in germination and injury to seedlings was less
marked in the field experiments than in the greenhouse experiments may
be due to several causes. The temperatures in the field were consider¬
ably lower than under greenhouse conditions, and the fungi may have
been less active for that reason. Again, the grain which had been treated
with a formalin solution was not absolutely dry when inoculated and the
trace of formalin present may have reduced the vitality of the spores.
One other fact, however, which may have had a marked influence is that
in the field the fungi used for inoculation would have to compete with
other fungi and bacteria in the soil and many of the spores may have been
injured before they could germinate and infect the grain. That such
competition between fungi and bacteria in the soil may not be uncommon
was indicated in a preliminary experiment in the greenhouse where wheat
inoculated with Helminthosporium gramineum was planted in sterilized
and unsterilized soil. It was found that the wheat planted in the ster¬
ilized soil was more severely injured by the fungus than the wheat planted
in unsterilized soil, the percentage of germination being less in the ster¬
ilized soil than in the soil not sterilized. In a second experiment of this

484

Journal of Agricultural Research

Vol. I, No. 6

nature the results were not as marked as in the first, although there was
a difference in germination of 3.8 per cent between inoculated wheat
planted in sterilized soil and inoculated wheat planted in soil which
had not been sterilized.

A SYNOPSIS OF WORK RELATIVE TO HELMINTHOSPORIUMS AND
FUSARIUMS ON CEREALS

The most comprehensive study of Helminthosporiums on grains is
that of Ravn (20) 1 who isolated three species from barley and oats and by
cultural and inoculation experiments, as well as a study of the morphology,
definitely established their identity. Eidam (12) was the first to under¬
take inoculation experiments with species of Helminthosporiums. He
inoculated barley with a strain of Helminthosporium secured from oats,
but without positive results. Ritzema Bos (21) describes some of the
diseases of barley in Holland and ascribes them to H . gramineum .
Frank (13) describes a disease of barley which appears on the lower leaves
of young plants and spreads gradually upward and believes it to be due
to an infection of H . gramineum . Ritzema Bos (22) describes a dis¬
ease on oats slightly different from a leaf spot in barley and believes it
to be caused by H . gramineum. Pammel (18) describes a characteristic
barley disease appearing in the United States and believes H . gramineum
to be the causal organism. Many other investigators, both in Europe
and this country, have studied the Helminthosporiums on grains with
more or less definite results, and the literature on the subject is extensive.
Practically all these studies, however, have been based on examinations
of diseased plants and, with the exception of the work of Eidam, already
quoted, have not been based on cultural and inoculation work. Hecke
(14) secured a pure culture of H . gramineum from barley plants. He
inoculated seedling barley plants both with mycelium and sclerotia and
secured positive results in the formation of brown spots on the leaves.
Ravn (20) cleared up the question of identity of three species of the
Helminthosporiums attacking barley and oats. In extensive cultural
and inoculation studies he obtained pure cultures. One of these he
secured from stunted barley plants and established that it was the cause
of deep-seated infection in the tissues of leaf, stem, and roots, while
another species affected only the leaves, but was not systemic. The
first he attributes to H . gramineum , the second to H . teres Sac. A
similar disease on oats is attributed to H . avenae Br. and Cav. These
three fungi were studied in pure cultures on beer wort and other culture
media and found to differ in cultural characteristics, H . gramineum ,
after 14 to 20 days’ growth on beer wort, producing a snow-white, uni¬
formly smooth mycelium; H. teres , a much less abundant mycelium,

bibliographic citations in parentheses refer to "Literature cited," pp. 487-489.

Mar. as, 1914

Imperfect Fungi from Wheat , Oats , and Barley

485

which gathers more or less in masses; and //. avenae, a mycelium more
nearly resembling H . gramineum , but less smooth and with more of a
tendency to mass together. The developmental history and morphology
of the mycelium and conidia in culture was very similar for the three
species, but when the conidia were measured in large numbers those of
H. teres were slightly longer than those of H. gramineum and those of H.
avenae slightly larger than those of H. teres .

In a series of inoculation experiments Helminthosporium teres from
barley transferred to barley, but not to oats, rye, or wheat; H . grami¬
neum to barley, but not to oats; and H . avenae to oats, very slightly to
barley, and not to rye.

Until Ravn made these intensive studies of the three Helminthospo-
riums they had been confused in the literature as to identity. The strain
of H . gramineum discussed in this paper corresponds in cultural and
morphological characteristics to the descriptions by Ravn.

Pammel, King, and Bakke (19) report a number of species of Helmin¬
thosporium on cereals in Iowa, among them H . gramineum . They cite
inoculation tests to show that infection occurred when barley seedlings
were inoculated with spores of this fungus and when the soil in which
seedlings grew was inoculated. Beckwith (5) reports the isolation of
undetermined species of Helminthosporium from old wheat soils, roots,
and stems of wheat in North Dakota, but no inoculation experiments
are mentioned. A comprehensive bibliography of the literature on
Helminthosporiums up to 1900 is given by Ravn (20).

The literature relating to Fusariums on grains is also very extensive.
Chester (10) reports that F. culmorum is the cause of the disease known
as scab of wheat and shows that many shrunken wheat kernels con¬
tain a fungous mycelium. Detmers (11) shows that the disease known
as wheat scab in Europe and caused by F . culmorum has become preva¬
lent in America. Selby (30) ascribes wheat scab in Ohio to the fungus
F. roseum Link, and believes the conidial form of Gibberella saubinetti to
be its conidial stage. Some field inoculations with Fusarium attempted
by him were unsuccessful.

The first investigator to show with any degree of certainty that Fusa¬
rium infection can be carried with the seed is Rostrup (25, 26, 27, 28).
Ritzema Bos (24), Westerdijk (35), Volkart (34), Appel (1,2), and Selby
and Manns (31) came to similar conclusions. Sorauer (32, 33) was the
first to prove that infection could be carried with the seed. He main¬
tains, however, that infection in this manner is of small consequence
as compared with infection through the soil.

Selby and Manns (31), in their studies on the form Gibberella, con¬
clude that this fungus attacks rye, oats, barley, and spelt. Inoculations
on wheat with pure cultures of Gibberella saubinetti (Mont.) Sacc. from
perithecia on wheat reduced germination to the extent of 17.1 and 32.4
per cent, respectively. Similar results on both wheat and oats were
obtained by them with Fusarium roseum from wheat and clover.

486

Journal of Agricultural Research

Vol. I, No. 6

Appel (3) believed that infection with Fusarium nivale Ces. is due
principally to soil infection, while Hiltner and Ihssen (15) believe that
seed infection is of more importance.

Muth (17) carried on pure culture inoculation experiments on rye with
Fusarium roseum . In these, 55 per cent of the inoculated seed sprouted
while only 63 per cent of the controls sprouted. A large number of
plants from inoculated seed, however, showed the results of infection
through a yellowish or yellowish brown discoloration of the roots.

Beckwith (4) reports numerous isolations of Fusarium species and
other imperfect fungi from stems and roots of wheat grown on soil
continuously cropped to wheat and from the soil itself.

Mortensen (16) demonstrated that rye seed heavily infected with
Fusarium nivale Ces. produced diseased plants. He states that not only
F. nivale but other Fusariums produce root diseases in cereal plants.

Bolley (6), from extensive field studies on wheat from land continu¬
ously cropped to wheat, has come to the conclusion that “through the
practice of continuous wheating, soils in many cases have become in¬
fected with from one to three or four definite parasitic fungi which
attack in the same manner as the flax-sick fungi attack and destroy the
flax crop on flax lands and, therefore, such wheat lands may be said to
be ‘wheat sick.’” These views .are further elaborated by him from
extensive field studies and observations (7, 8). Bolley (9) also reports
on the isolation of a considerable number of imperfect fungi from the
nodes and intemodes of wheat plants grown on experimental plats at
the North Dakota Agricultural Experiment Station. Among them unde¬
termined species of Helminthosporium and Fusarium occurred in abun¬
dance. No inoculation experiments are reported.

Schaffnit (29) in a comprehensive work on “ Schneeschimmel ” gives
a discussion of the fungus Fusarium nivale with relation to its occur¬
rence, morphology, cultural characteristics, physiology, and preventive
measures. He shows that this disease is due both to soil infection and
seed infection, the former being more common. Incidental to his work
on F. nivale Schaffnit (29) performed some inoculation experiments
with F. rubiginosum Appel and Woll. on etiolated rye seedlings in
damp atmosphere with positive results. The number of inoculations
is not stated. F. rubiginosum has recently been demonstrated by
Dr. H. W. Wollenweber to be identical with F. culmorum . A compre¬
hensive bibliography of literature dealing with Fusariums on cereals is
given by Mortensen (16).

CONCLUSIONS

The experiments described in this paper and the literature cited
show that some of the imperfect fungi occurring on small grains and
inducing leaf spots or systemic infections are pathogenic when, under
favorable conditions, they come in contact with seeds and seedlings,

Mar. 25, 1914

Imperfect Fungi from Wheat, Oats, and Barley

487

while other forms apparently are nonparasitic. Helminthosporium
gramineum and Fusarium culmorum were found to be parasitic, while
Cladosporium gramineum and an undetermined species of Altenaria were
not parasitic under the conditions here described. That only certain
species are pathogenic is to be expected. Their identity as well as that
of the large number of forms apparently saprophytic on cereals is more
or less confused in the literature but should be determined, and the
extent to which these fungi affect cereals should be ascertanied by
laboratory and greenhouse studies. These need to be reinforced by pure
culture inoculations of seeds, seedlings, plants in various stages of
growth, and soil under field conditions before the exact relation of such
fungi to cereal cropping can be definitely established.

LITERATURE CITED

1. Appel, Otto.

1907. Fusarien als Erreger einer Fusskrankheit des Getreides.
Anst. Land- u. Forstw., Heft 4, p. 32-33.

In Mitt. K. Biol.

2. -

1908. tiber die Schadigung von Getreide dnrch Fusarien. In Mitt. K. Biol.
Anst. Land- u. Forstw., Heft 6, p. 10-11.

3- -

1909. Einige Kranklieiten und Schadigungen des Wintergetreides. In Illus.
Landw. Ztg., Jahrg. 29, No. 70, p. 665-666.

4. Beckwith, T. D. *

1910. Mycological studies upon wheat and wheat soils to determine possible
causes in deterioration in yield. In Science, n. s., v. 31, no. 803, p. 798.

5- -

1911. Root and culm infection of wheat by soil fungi in North Dakota. In
Phytopathology, v. 1, no. 6, p. 169-176.

6. BollEy, H. L.

1909. Deterioration in wheat yields due to root rots and blight producing dis¬
eases. N. Dak. Agr. Expt. Sta. Press Bui. 33, 4 p.

1911a. Interpretations of results noted in experiments upon cereal cropping
methods after soil sterilization. In Science, n. s., v. 33, no. 841, p. 229-232.

8. -

1911b. The work of imperfect fungi in cereal crop deterioration. Abstract. In
Science, n. s., v. 33, no. 842, p. 259-260.

9- -

1912. [Report on the work of the] Department of Botany and Plant Pathology.
N. Dak. Agr. Expt. Sta. 22d Ann. Rpt., 1911/12, p. 23-60.

10. Chester, F. D.

1891. The scab of the wheat. Del. Agr. Expt. Sta. 3d Ann. Rpt., 1890, p. 89-90,
fig. 14-15.

11. Detmers, Freda.

1892. Scab of wheat. In Ohio Agr. Expt. Sta. Bui. 44, p. 147-149, fig. 4-5.

12. Eidam, E.

1891 . Das Vorkommen der Fleckenkrankheit auf Gersten- und auf Haferblattem.
In Der Landwirt, Bd. 27, p. 509. Original not seen.

13. Frank, A. B.

1897. Kampfbuch gegen die Schadlinge unserer Feldfruchte. 308 p., 46 fig.,
20 pi. Berlin.

488

Journal of Agricultural Research

Vol. I, No. 6

14. HECKE, L.

1898. Die Braunfleckigkeit oder Blattbraune der Gerste. In Wiener Landw.
Ztg., Bd. 48, p. 435.

15. Hiltner, Lorenz, and Ihssen, G.

1911. Uber das schlechte Auflaufen und die Auswinterung des Getreides infolge
Befalls des Saatgutes durch Fusarium. In Landw. Jahrb. Bayern, Jahrg. 1,
No. 1, p. 20-60, 8 fig.; No. 4, p. 315-362, 2 fig.

16. Mortensen, M. L.

1911. Om Sygdomme hos Kornarteme , foraarsagede ved Fusarium- Angreb
(Fusarioser). In Tidsskr. Landbr. Planteavl, Bd. 18, p. 250.

17. Muth, Franz.

1908. Uber die Infektion von Samereien im Keimbett. Ein Beitrag zur Samen-
untersuchung und Samenziichtung. In Jahresber. Ver. Angew. Bot., Jahrg. 5,
1907, p. 49-82.

18. Pammel, L. H.

1892. New fungous diseases of Iowa. In Jour. Mycol., v. 7, no. 2, p. 96-97.

19. - , King, Charlotte M., and Bakke, A. L.

1910. Two barley blights, with comparison of species of Helminthosporium upon
cereals. Iowa Agr. Expt. Sta. Bui. n6, p. 179-190, 4 pi.

20. Ravn, F. K.

1900. Nogle Helminthosporium-Arter og de af dem Fremkaldte Sygdomme hos
bygog Havre. 220 p., illus., 2 pi. K^benhavn.

21. Ritzema Bos, J.

1898. De bladvlekziekte der gerst, veroorzaakt door Helminthosporium grami-
neum Rabhst. In Landbouwk. Tijdschr., p. 42. Original not seen.

22. -

1900. Phytopathologisch laboratorium Willie Commelin Scholten. Verslag
over de inlicKtingen gegeven in 1899. In Landbouwk. Tijdschr., p. 126.
Original not seen.

23- -

1904-5. Geringe kiemkracht van in 1903 gewonnen zaad. In Tijdschr. Plan-
tenziekten, jaarg. 10, afl. 5/6, p. 152-165, 1904; jaarg. n, afl. 4/5, p. 124-137,
1905.

24* -

1905. Phytopathologisch laboratorium Willie Commelin Scholten. Verslag
over onderzoekingen gedaan in — en over inlichtingen gegeven van wege
hovengenoemd laboratorium in het jarr 1904. In Tijdschr. Plantenziekten,
jaarg. 11, afl. 1/2, p. 24-25.

25. Rostrup, E.

1893. Oversigt over de i 1892 hos Markens Avlsplanter optraadte Sygdomme.
In Tidsskr. Landokonom., Rsekke 5, Bd. 12, p. 633-664. Original not seen.

26. -

1895. Oversigt over Sygdommenes Optraeden hos Landbrugets Avlsplanter i
Aarets 1893. In Tidsskr. Landbr. Planteavl, Bd. 1, p. 140.

27. -

1903. Oversigt over Landbrugsplantemes Sygdomme i 1902. In Tidsskr.
Landbr. Planteavl, Bd. 10, p. 364.

28. -

1904. Oversigt over Landbrugsplantemes Sygdomme i 1903. In Tidsskr.
Landbr. Planteavl, Bd. 11, p. 402.

29. Schaffnit, E.

1913. Der Schneeschimmel nnd die fibrigen durch Fusarium nivale Ces. hervor-
gerufenen Krankheitserscheinungen des Getreides. In Landw. Jahrb., Bd. 43,
Heft 4, pi. 1-4.

Mar. 25, 1914

Imperfect Fungi from Wheat, Oats, and Barley

489

30. Selby, A. D.

1898. Some diseases of wheat and oats. Ohio Agr. Expt. Sta. Bui. 97, p. 40-42,
fig- 4-

31. Selby, A. D., Manns, T. F.

1909. Studies in diseases of cereals and grasses. Ohio Agr. Expt. Sta. Bui. 203,
p. 212-224.

32. Sorauer, Paul.

1901. Der Schneeschimmel. In Ztschr. Pflanzenkrank., Bd. u, Heft 4/5, p.
217-228.

33* - ..

1903. Uber Frostbeschadigungen am Getreide und damit in Verbindung ste-
hende Pilzkrankheiten. In Landw. Jahrb., Bd. 32, Heft 1, p. 1-68, 1 fig.,
pi. 1-4.

34. Volkart, Albert.

1908. Pflanzenschutz. In Landw, Jahrb. Schweiz, Jahrg. 22, p. 32-33.

35. Westerdijk, Johanna.

1909. Fusarium in de tar we. In Phytopath. Eab. “Willie Commelin Schol-
ten,” Jaarverslag. 1907/08, p. 3-4.

PLATE LXII

Fig. i. — Wheat seedlings from seed inoculated with spores of Helmintkosporium
gramineum and from seed externally sterilized ; photographed three weeks after plant¬
ing. The three rows of pots on the left were sown with seed inoculated with spores
of H. gramineum from barley, the two rows in the center with sterilized seed, and
the three rows on the right with spores of H. gramineum from wheat.

Fig. 2. — Barley seedlings from seed inoculated with Helminihosporium gramineum
and from sterilized seed; photographed three weeks after planting. The three rows
of pots on the left from seed inoculated with spores of H. gramineum from barley,
the next three rows from seed externally sterilized, and the row on the right from
seed inoculated with H. gramineum from wheat.

Fig. 3. — Wheat seedlings from seed inoculated with spores of Fusarium culmorum
from oat seedlings (two rows of pots on right) and from seed externally sterilized (two
rows of pots on left). Photographed two weeks after planting.

Fig. 4. — Barley seedlings from seed inoculated with spores of Fusarium culmorum
from oat seedlings (two rows of pots on right) and from seed externally sterilized
(three rows on left). Photographed two weeks after planting.

Fig. 5. — Oat seedlings from seed inoculated with spores of Fusarium culmorum from
oat seedlings (two rows on right) and from seed externally sterilized (three rows on
left). Photographed two weeks after planting.

(490)

iperfect Fungi

Plate LXIII

PLATE LXIII

Root systems of wheat seedlings grown in 6-inch pots from seed externally sterilized
(left) and from seed inoculated with Helminthosporium gramineum from wheat
(right). Photographed six weeks after planting.

THE ORIGIN OF SOME OF THE STREPTOCOCCI FOUND

IN MILK

By L. A. Rogers and Arnold O. Dahlbbrg,

Dairy Division , Bureau of Animal Industry

INTRODUCTION

In the higher plants and animals we are accustomed to associating
species with a more or less definite habitat. Certain animals are found
only in certain localities. One species of trees may be found only on
a particular type of soil. A still narrower limit of distribution is found
in some of the parasitic fungi which grow only on closely related host
plants. Zoologists or botanists find the types on which they base their
descriptions in the natural habitat of the organism. This relation has
not always existed in the published descriptions of bacteria. The
association of a natural group with a particular habitat has been more
or less incidental, except perhaps with the pathogenic bacteria, and
even with some of these it is not impossible that the pathological con¬
ditions under which they are found may not be the true habitat of the
species. The colon group, while it is frequently found in water and milk,
has its natural habitat in the intestinal tract of warm-blooded animals.
Winslow found that certain chromogenic cocci were associated with
the skin of animals.1 Some of the English bacteriologists have pointed
out that the streptococci from horse manure, for instance, have a set of
physiological reactions which differentiates them from those from saliva
or pathological conditions.3 It is only through a knowledge of the habitat
and the study of sufficient cultures to establish a type that true bacterial
species can be determined. If we were to write a description of the
German people we would go to Germany, not to an American city
where German immigrants live.

Countless descriptions have been written of bacteria isolated from
milk until we have come to consider certain types as peculiar to this
medium. The bacteria found in milk, however, are a heterogeneous
collection, and the true types of milk bacteria are to be sought in the
sources from which milk is contaminated. Esten has suggested that
the streptococci or lactic-acid bacteria of milk come originally from
the mouth of the cow.8 The feces of the animal must, unfortunately,

1 Winslow, C. E. A., and Winslow, Anne R. Systematic relationships of the Coccacese. ed. i, 300 p.,
iUus. New York, 1908.

* Andrewes, F. W. Report on the micro-organisms present in sewer air and in the air of drains. 36th
Ann. Rpt. Local Govt. Bd. [Gt. Brit.], 1906-07, Suppl. Rpt. Med. Off., p. 183-204. 1908.

8 Esten, W. M. Bacterium lactis acidi and its sources. Conn, Storrs Agr. Expt. Sta. Bui. 59, 27 p.,
5 fig. 1909.

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Vol. I, No. 6
Mar. 2 s, 1914

492

Journal of Agricultural Research

Vol. I, No. 6

be considered as a possible source of bacteria in milk, among which
would undoubtedly be found members of the lactic group. Kinyoun
and Dieter believe that the presence in milk of cocci which form chains
in lactose bile at 3 70 C. is presumptive evidence that the milk is con¬
taminated with feces.1 It is the more common practice, however, to
consider this type as the indication of the presence in the herd pro¬
ducing the milk of one or more cows with infected udders.

The mouth is known to contain streptococci, and the habit of cows of
licking their flanks and udders provides a more or less direct connection
between the mouth and the milk pail. Each of these sources may be
considered as the normal habitat of bacteria. Under these conditions
they persist for indefinite generations, adapting themselves to their
environment until it is reasonable to suppose the characters acquired
become sufficiently fixed to have at least varietal significance.

The study of streptococci originating within such circumscribed limits
is of interest in addition to its taxonomic importance, in the light it may
cast on the origin of some of the bacteria in milk and the significance
from the hygienic standpoint of the presence of certain types.

In this paper are recorded the results of a study of streptococci repre¬
senting three of the possible sources from which this group may find its
way into milk. The morphology of this collection was studied with
the hope that this would give some basis for a division into varieties.
The ability of these cultures to utilize a number of carbohydrates and
alcohols was determined. On the basis of these fermentations several
groups are established, each of which is made up of a large number of
identical cultures constituting the type about which are grouped similar
cultures, but which varied from it in one or two reactions. The prob¬
able relation of one of these groups to well-known species is pointed out.

THE CULTURES STUDIED

A collection of streptococci were obtained from milk, from bovine
feces, from the mouths of cows, and from the udders of cows. With the
exception of those from milk an effort was made to make the cultures
as representative as possible. The procedure of isolating the milk cuh
tures followed that usually employed in the laboratories of boards of
health. Small portions of the milk were added to lactose-bile tubes
which were incubated at 370 C. Tubes showing streptococci in distinct
chains on microscopical examination were plated on lactose agar and the
chain-forming cocci subcultured. In this way 42 cultures were isolated
from 25 samples of milk and cream collected at Washington or at the
creamery at Troy, Pa. No two samples came from the same farm. A
few cultures were obtained through the courtesy of Dr. Kinyoun and
Mr. Dieter from lactose-bile tubes in the laboratory of the health depart¬
ment of the District of Columbia. These cultures, therefore, did not

1 Kinyoun, J. J., and Dieter, I,. V. A bacteriological study of the milk supply of Washington, D. C,
Jour. Amer. Pub. Health Assoc., v. 2, no. 4, p. 262-274. 1912.

Mar. 25, 1914

Streptococci in Milk

493

represent the normal streptococci of milk but rather those which would
usually be distinguished as indicating contamination from infected udders
or fecal sources.

Fifty-one cultures were isolated from 19 samples of milk obtained
by milking directly into sterile test tubes. The cows from which these
samples were obtained represented all gradations of infected udder
from occasional evidence of garget to acute mammitis. Part of these
were in the Dairy Division herd at Beltsville, Md., and the remainder in
the herd on the Naval Academy farm at Annapolis, Md. One hundred
and fourteen cultures came from 56 samples of cow manure obtained t
with the exception of a few from Troy, Pa., at the Dairy Division farm
and at the dairy of the Government Hospital for the Insane at Wash¬
ington. Thirty-nine cultures were made from the mouths of animals
at the Dairy Division farms. With the exception of one culture obtained
from the mouth of a mule, all of these cultures were of bovine origin.
In Table II the origin of the culture is indicated by M for milk, U for
udder, F for feces, and B for mouth. The sample from which the culture
was secured is indicated by a number following the letter. For instance,
“F15” represents sample of feces No. 15. This will enable the reader
to determine the origin of each culture and the number of cultures from
each sample.

MORPHOLOGY OF THE CULTURES

While it is generally recognized that there is little morphological basis
for subdivisions of the streptococci, reference is frequently made to
certain types of cells. Stowell, Hilliard, and Schlesinger,1 in selecting
streptococci from milk for comparison with those isolated from the
human throat, rejected diplococci and the oval-chained form which they
designate as the Streptococcus lacticus of Kruse or the Bacillus lactis
acidi group, respectively. In selecting our cultures no attention was
paid to morphology beyond determining that it was a coccus apparently
dividing in one plane, with the exception of those from milk, which were
not accepted if they did not form chains of at least 8 or 10 cells. The
morphology of nearly all cultures was determined by examination of
specimens stained with gentian violet. Camera-lucida drawings were
made using a Eeitz 3 mm. objective and No. 18 ocular, a combination
which gave a magnification of 2,400 diameters at the ocular, or 4,800
diameters on the drawing board. Sufficient light to give a clear image
was obtained by using a special arc light with a copper-sulphate ray
filter.

Preliminary studies showed that the medium on which the culture was
grown had an appreciable influence on both the size and the form of the
cell. This is shown in figure 1 , which is reproduced from camera-lucida
drawings of typical cultures grown on various media. Milk gave quite

1 Stowell, B. C., Hilliard, C. M., and Schlesinger, M. J., A statistical study of the streptococci from milk
and from the human throat. Jour. Infect. Diseases, v. 12, no. 2, p. 144-164. 1913.

494

Journal of Agricultural Research

Vol. I, No. 6

uniformly smaller cells and less tendency to chain formation than broth
or agar. The cells at a are from culture lo on milk, b on broth, and c on
agar, all incubated 48 hours at 3 70 C. The difference between the dis¬
tinctly rod-shaped cell found on agar and the small round cell obtained
from milk is marked. That differences in size of cells are not due entirely
to differences in the medium is shown by the chain at h. This com¬
bination of small and large cells in a single chain is not unusual in broth,
a medium in which there is a marked tendency to form enlarged and
abnormal cells. In some cultures the transition from normal cells to
those of monstrous size and form was so rapid that it was difficult to

obtain preparations

<$>

0 0
e

showing what could be
considered normal cells.
The most satisfactory
preparations were ob¬
tained in incubating
broth cultures until a
distinct cloudiness was
obtained, centrifuging
the culture, siphoning
off the broth, and wash¬
ing the sediment with
sterile water. After cen¬
trifuging again the wa¬
ter was siphoned off, and
apreparationmadefrom
the sediment. This gave
a clear field suitable for
examination under a
high-power microscope.

Various types of cells which were found in this collection are shown in
figure 2. It will be observed that much of the variation in these types is
in size only or in chain formation. The slender-pointed cells shown at F
were peculiar to the cultures obtained from the mouth of animals, but the
cultures from this source were not confined to this type. In Table II the
letter under the heading “Morphology” refers to figure 2, although it is
obvious that in many cases the assignment to a particular type can be
only an approximation. The variation of the morphology is so great and
so easily affected by the environment that it was not considered in the
final arrangement of groups. It should be stated, however, that among
the udder cultures the tendency to chain formation was much more
marked and more constant than among all other cultures.

METHODS OF DIFFERENTIATION

When morphological distinctions are lacking, we are forced to use the
physiology of the organism as a basis of classification. No single system

Fig. i.— Cells of streptococci, showing variation in size and morph¬
ology. a, culture lo on milk; 6, culture lo on broth; c, culture lo on
agar; d, culture li on lactose bile; e, culture li on broth; /, culture qm
on milk; g and Mt, culture gin on broth.

Mar. 25, 1914

495

Streptococci in Milk

of characters can be adopted for all classes of bacteria. The significant
characters will be found for each group only by a study of its normal
activities and the utilization of those functions which show the nature,
limitations, and relationship of the group. The striking characteristic
of the streptococci is their ability to form acids from carbohydrates and
related substances, and this peculiarity has been very generally utilized
for purposes of classification. The voluminous literature bearing pro
and con on the constancy and the value of these tests has been reviewed
fully in various papers and need not be taken up here. It may be safely
asserted, however, that the fermentative ability is as constant and as
significant for purposes of classification as the characters adopted by
those who reject the fermentation tests as too variable. For instance,
Davis, who rejects the sugar fermentations as untrustworthy, divides the

C08 o ^ D

C D UE F

Fig. 2. — 'Types of cells of streptococci.

streptococci into five groups on the basis of hemolysis, green colonies on
blood agar, capsule, solubility in bile, inulin fermentation, experimental
arthritis, and experimental endocarditis.1

For purposes of classification, we have used the liquefaction of gelatin
and the fermentation of dextrose, saccharose, lactose, raflinose, starch,
inulin, mannite, and glycerin. Adonite and dulcite were tested, but as
they were fermented by only one or two of these cultures they were of
no value. The liquefaction of gelatin was determined by inoculating
the surface of the gelatin tube with a few drops of a broth culture and
measuring the liquefaction after 30 days at 20° C.

The fermentation of the test substances was determined in a medium
made as follows :

Per cent.

Beef extract . o. 4

Peptone . 1. o

Dibasic potassium phosphate . 5

Test substance . 2.0

1 Davis, D. J. Interrelations in the streptococcus group with special reference to anaphylactic reactions.
Jour. Infect, Diseases, v. 12, no. 3, p. 386. 1913.

287360— 14 - s

496

Journal of Agricultural Research

Vol. I, No. 6

The cultures were incubated for seven days at 30° and titrated cold
against twentieth-normal sodium hydrate with phenolphthalein as an indi¬
cator. From the results so obtained is subtracted the titration of a blank,
and the result is expressed as the percentage of normal acid. Some objec¬
tion may be raised against the use of 30° C. as an incubation temperature
rather than the more common one of 370. The lower temperature was
adopted because practically all streptococci will grow at this temperature,
while a few grow at 370 slowly or not at all.

The fermentation produced by the streptococci is in almost all cases
so marked that there is very rarely any question about the presence or

absence of the fermentation. Of all the substances we have used gly¬
cerin forms an exception to this rule. The fermentation proceeds
slowly and in seven days may be slightly above or slightly below 1 per
cent normal acid, the point selected as marking the line between fer¬
mentation and no fermentation. This is illustrated by Table I, which
shows the progressive rate of fermentation by typical cultures. Three
cultures fermenting dextrose are included to show the usual course of the
fermentation in the more easily fermented sugars. Each titration was
made from a separate tube. A study of this table shows that the 12
cultures may be divided into three quite distinct types on the basis of
the rate of fermentation of glycerin. This is shown more clearly in
figure 3, in which the average titrations for each of the three types are
plotted. Two of these cultures fermented the glycerin with comparative

Max. 25, 1914

Streptococci in Milk

497

rapidity and after three days there was no question that the cultures
were able to utilize glycerin. Those represented by the curve Z>, on
the other hand, produce only a very slight increase in acidity, which even
at the end of 14 days is only slightly above 1 per cent normal.

Between these two is a third group in which there is a slow but dis¬
tinct increase in acidity. At seven days the acidity is above 1 per cent
normal. While an error may be introduced in some cases by drawing
the line between fermentation and no fermentation of glycerin at 1 per
cent normal, it is believed that in these results this error will be slight.
These results illustrate the value of the exact results obtained by titra¬
tion which we have always used in preference to the simpler and more
rapid way of determining the change of reaction with litmus in solution
in the broth or with litmus papers.

We also consider it a decided advantage to allow sufficient time for
the completion of the fermentation, thus securing an end point rather
than some intermediate and varying determination. Seven days are
not sufficient for the completion of the glycerin fermentation, but it is
undoubtedly ample for other test substances which we have used.

Table I. — Progressive fermentation of dextrose and glycerin .
Dextrose.

ec .

0. 15

0. 19

0.31

0. 34

0.44

0. 55

o- 58

O. 91

0. 53

0. 96

0. 78

1. 21

• 25

• 23

■ 80

1. 67

I. 31

3. 65

2. 73

4. 96

4. 85

4. 96

5. 43

eh .

. 22

.46

•35

i- 57

I- 54

2- 40

2. 38

4. 61

3*74

4.68

• IS

• 14

•37

*96

.86

.28

• 63

1. 83

.92

•33

i- 13

et .

* 05

• 59

.82

• 95

I. 09

1. 30

X* 49

2. l6

1. 73

2. 18

2. 30

2. 63

ev .

•is

.09

•34

• 17

.00

.40

•45

.66

1. 03

.96

• 70

1. 18

hw . . .

. 20

. 11

.67

•95

I. 07

1-35

I. 67

1.38

2. 08

2. 71

a- 45

3- 15

nr . T , f .TT .

• 15

. 09

• 14

. 00

• 25

• °3

• 02

• 31

• 43

. 76

. 21

jiy „ , T , . .

. 20

. 29

• 04

. 00

* °5

. 00

. 01

. OO

. 21

.00

. II

. 09

• 02

• 07

. 09

• 05

. 02

. 11

.48

.46

• 03

om .

•is

.00

.09

. II

.09

.06

.00

. 21

• 13

•19

.00

Pi .

• 05

.09

. 22

• 32

.24

•36

•31

• 71

•53

• 76

.66

•78

THE CORRELATION OF PHYSIOLOGICAL CHARACTERS

The complete results of the tests made on this collection of cultures are
presented in Table II. The reduction of neutral red and the curdling
of milk are given in the table, but are not used in the correlations.
Adonite and dulcite are necessarily excluded, since these cultures almost
without exception fail to ferment these two substances.

498

Journal of Agricultural Research

Vol. I, No. 6

Table II. — Physiological characters of all cultures.

Culture No.

Origin.

Morphology.

Chains.

Liquefied gelatin.

Neutral red re¬

duction.

Milk curd.

Percentage of normal acid.

Milk.

Broth.

Dextrose.

Adonite.

Saccha¬

rose.

Lactose.

Raffinose.

Starch.

Inulin.

Mannite.

Glycerin.

Dulcite.

mm.

—

4- 1

0. 1

4-3

1*3

0.6

0. 1

—

i- 7

. 1

4. 0

4- 0

•4

2.4

—

2* 5

4. 1

4. 2

. 1

. 2

2. 9

. 1

—

2. 6

3- 6

3*8

. 2

1.8

•9

—

2. 7

4-6

. 1

. I

. 1

—

4. 4

4- 5

. 1

. 2

. 1

2* 3

—

5°

—

-f

4-3

. 1

4- 7

5- 1

. 2

•4

• 5

4.4

x* 7

. 2

—

4- r

. 1

4-8

4- 8

. 2

. 1

4.0

2. 1

. 1

—

4. 6

. 2

5-4

. 2

. I

• 7

—

6- 5

4. 1

5- x

. 1

. I

. 2

3*9

•9

—

6.6

•4

5*0

2.4

—

3- 0

• 3

5-o

. 1

2. 1

• 3

—

4. 9

3-9

3*6

i* 5

• 7

—

4- 7

3*o

3-6

. 1

. I

•9

—

2. 0

. 2

3-4

3- 2

. 2

■ 3

■4

. 2

—

T-

1. 8

• 4

. 2

—

6.6

• 4

5- 6

* 3

. 2

6. 2

2. 0

. 2

—

5* 6

5. 6

—

5- 6

5*3

4. 9

. 1

3* 5

Mio

—

7- 1

. 2

5- 7

• 3

. I

• 5

. 2

—

5*o

•4

5- 2

• 4

. 2

4* 3

x* 3

. 1

Mia

—

5. 1

K. 5

Ml2

—

5- 5

. 1

5- 6

4- 5

4-4

. 1

—

5-9

• 1

5-8

4- 7

. 2

4- 3

•4

M13

—

5-3

4.8

5- 5

•3

. I

4-3

I. 2

. 1

—

5-4

5* 1

5*3

. 1

. I

2. 7

. 2

• 1

Mio

—

7.0

. 1

5- 2

. I

• 5

Mio

—

6.6

. 1

5*6

. 1

. I

• 5

M14

— *

5. 4

4. 8

5* 2

Mis

—

5-3

5-4

4.6

3.8

. 1

M15

—

5. 3

5. 5

5* 2

. 2

. 1

1. 1

M16

6. 0

5- 6

M16

6. s

. 1

5. 4

. i

. 1

—

4.4

3*6

4.0

. 1

3*8

—

4-6

4- 7

4*8

•3

•4

. 1

x* 9

—

4. 2

4*9

. 1

. I

■ 1

—

4.8

. 1

4. 3

. 2

. 1

—

4- 5

. 1

4.4

4*8

. 1

. I

•3

4. 6

4- 3

—

4.4

. 1

4* 7

4.9

. 2

—

4- 5

4.8

4*8

—

4- 7

•3

4-6

4.6

. I

•4

•3

—

4- 5

4.6

5-o

. I

. X

—

4- 5

. 1

4-9

5-0

. I

• 2

4. 7

M17

—

6.4

5* 1

. I

2* 3

•3

M17

—

6.4

. 2

5- r

. I

3*8

•4

. X

M17

—

5- 5

. 1

5* 9

. I

. 2

. a

• 5

M18

—

5-8

5*6

5* 2

. I

3* 9

M18

—

6. 1

5* 5

5* 6

. I

3*4

•4

M19

—

5-9

. 2

5- 5

. I

. a

• 7

M19

—

6. 2

• 1

0. 1

5* 5

. I

. 2

0.6

M20

—

6.8

4* 7

. I

. 2

•4

Mai

. x

M21

—

7- 1

. 2

5*6

• 5

. X

M22

—

6. 1

•4

5*3

. I

2.8

• 3

M22

—

5-6

. 2

6- 3

. I

2. 7

•4

—

5-7

6. 0

5-5

5-3

. I

—

6.4

. 2

5- 5

5- 3

. 1

_ .

5- 8

5.8

MaJ

—

6. 0

• I

5- 5

5- 2

• 3

•3

4. 6

Ma3

—

6. 2

* 7

5- 1

< 7

5. 0

x- 5

. I

M24

—

5-3

■3

4. 8

5- 1

. 2

4-3

• 5

M24

—

5-o

. 1

4*7

5* 2

. 2

. X

4. 1

i* 5

—

5- 7

. 2

5*9

5- 5

5-3

4- 7

. I

—

5- 1

5- 4

5- 1

5- 0

4- 7

. 1

—

5-8

. 2

6.4

5-8

5* i

4*3

. 2

—

5*7

6.7

5- 6

5- 2

4*3

. 2

—

5*7

5*8

5- 6

5-4

0. 1

—

5* 7

6. 0

5-4

5-3

—

4*9

5- 5

5*4

0. X

—

4*3

4. 6

5* 2

5-9

O. I

. X

—

6. 1

. 2

5* 5

4.8

1.8

i* 7

. 2

—

7-7

. I

8. 9

5*4

3*2

0. a

1.4

■3

Mar. 25, 1914

Streptococci in Milk

499

Table II. — Physiological characters of all cultures — Continued.

Culture No.

Origin.

Morphology.

Chains.

Liquefied gelatin.

Neutral red re¬

duction.

Milk curd.

Percentage of normal acid.

Milk.

Broth.

| Dextrose.

Adonite.

Saccha¬

rose.

Starch.

Inulin.

Mannite.

Glycerin.

Duldte.

mm.

M25

—

5* 2

0. 2

4-3

4- 7

3- 2

0. 2

2. 9

4.0

0. 2

0. 1

M25

—

4.6

. 2

4-4

4- 7

3- 5

0. 2

3*2

3* 7

• 3

. X

Fio

—

— *

6.5

6. 1

5-0

4.8

4.6

•3

. X

Fio

—

4.8

0. 2

5*5

5-5

5- i

4*4

—

5-8

5*8

5* 1

5-o

. 2

Fl2

—

5-4

. 1

5*4

5-6

5-o

. X

. 2

Fi3

—

S-6

. 2

7-6

5. 7

4- 5

■3

1.8

. 2

Fl3

—

4.8

. 1

5-6

5-9

4*4

F13

5- 5

. 1

5* 7

5* 9

6. 0

4* 7

. 1

£5

—

5*5

5-3

4*9

4.8

6. 2

•4

—

6. 1

. 2

6. 2

5-4

. X

. 2

5*3

£14

—

5- 5

. 1

6. 2

5*8

5*x

4* 5

4. 1

. 2

FiS

—

6.9

6. 0

5. 2

4- 7

6. 2

Fl5

—

4-3

. 1

5-4

5- 1

5-o

E15

—

5-6

. 2

6- 3

5- 5

5*3

F16

—

5* 7

. 1

5-8

5* 5

5* 0

4. 4

4. 5

. 2

4. 2

5- x

. X

—

4-5

. X

4.6

4.9

. 2

. X

—

4. 7

4* 4

4. 6

. 1

. X

—

4.8

. X

4.9

4. 4

. x

. X

. I

. X

—

3< 4

. X

4. 8

4- 5

. 3

. 1

. 2

. 1

—

4.8

4* 5

3-6

. 2

•3

3*9

2.7

. 2

: —

4. 8

. X

3-8

. I

. 2

. X

3* 6

4. 5

. I

3-8

4. 0

3* 4

. X

E17

—

6.4

. 2

6. 2

4.9

•3

EI?

—

5- 5

. X

5-8

5* 3

5* 2

. 1

. I

E17

—

5-4

. 2

5-9

5-3

5* 1

4. 8

. 2

. I

F18

—

5-8

. X

4.9

5- 2

4-9

3*9

6.4

0. 1

. X

F18

—

4. 9

5* 1

4. 9

5* 5

. X

. 2

. X

Fi8

—

5* 9

4. 8

4. 9

5* x

6. 0

5*8

. X

F19

—

3* 5

5* 3

5- 4

6. 0

. 2

. 3

. I

F20

—

5-4

. 2

5-8

5- 1

. 1

. 2

F20

—

5-8

5.8

5-5

5*2

. I

F20

—

6.9

5*5

. 2

. X

. 1

. I

F21

—

5* 5

. I

4.9

5-3

4.8

4. 6

. 1

• I

F21

—

5-4

. I

5.8

5*o

5- 1

4-6

F21

—

6. 1

. X

5*o

5- 1

4.6

• 1

F22

—

5- 9

. I

6- 1

6.2

. 2

. 1

3*9

. I

F23

—

6. 1

. I

6-5

5- 2

4-8

. 1

£23

—

5-9

. I

5-9

5-o

5-3

F24

—

6.6

• 2

5-o

4-4

6.3

. 1

F 24

—

5-8

. I

5-9

5- 2

4-3

. X

4.0

F24

—

6.6

. I

6. 0

5-3

4- 7

. 2

4. 1

F25

—

5-9

. X

5-5

5*2

4*3

. 1

. 2

F25

—

4.9

5*2

4-3

4*8

. 1

4. 1

2*3

. I

E25

—

6-5

. 2

5-4

4.9

4*8

6. 2

. X

F26

—

5-8

. 2

5-9

5-4

5* 1

. I

F26

—

6. 1

. 2

6.4

5*4

4.6

6*5

. X

F26

—

5-8

• 2

6.1

5-5

5* i

. X

. I

F27

—

7-3

. 2

5*9

5- 7

. 2

2. 2

. X

F27

—

7- 5

. 2

7. 2

4.8

. 5

. 2

* 3

x*7

. 1

F28

—

5-8

3-8

5* 7

5*o

. 2

. 1

F28

—

5-9

6. 2

5*4

4*9

. X

. 1

. X

F29

—

5- 2

. I

5- 7

5-4

5- 1

3*8

5*9

. X

. I

F30

—

5- 5

. I

5- 0

4* 6

5*9

6. 2

• X

E31

—

5-9

5- 1

5-o

5* 7

5*6

. 1

E31

—

6. 0

. I

5-4

4. 7

F3i

—

5* 4

5- 5

5*3

. I

. 2

F32

—

5- 1

4-3

5. 0

5* 2

6*3

5*8

F33

—

4.9

. I

5-8

5-2

4.6

. 2

F34

—

5- 1

•5

4.8

3*5

•5

• 7

4*6

2*3

. I

F35

—

6. 2

6- 5

6.0

4*5

6. 1

6.4

4*5

• I

f3S

—

6.6

6.4

x* 7

4* 7

. 1

4*3

. I

F36

—

5-2

5-o

4.6

6. 1

6- 1

. X

F36

—

6.3

4.9

4- 7

4*6

. 2

. 1

. 2

F36

—

6- 5

5-o

5-0

4*8

. 1

F37

—

5* 7

5-o

4.6

4*4

6. 0

6.4

. 1

F37

_ _

—

-1-

. X

5* 2

5- 6

4* 3

6. 0

. 2

F37

—

6. 1

4-9

5* 1

4.9

5*9

6.3

• 1

. I

F38

—

6-3

5-2

4.9

4.8

6.4

. 1

F30

—

6.6

6. 3

5- 2

4.8

. 2

F39

—

6.8

6.2

5- 1

5*o

3*9

F39

—

6. 5

6.4

4* 7

4-4

. 2

F40

—

5-6

5- 2

4- 7

5*3

. 1

F40

—

5-8

5-3

5-6

6. 0

6. 2

' F40

—

1 +

5- 7

5*2

5-8

5*3

6. 1

. 2

3*6

5°°

Journal of Agricultural Research

Vol. I, No. 6

Tabi^E II. — Physiological characters of all cultures — Continued.

Morphology.

Chains

Percentage of normal acid.

1 Culture N(

■g V

O O
« *

■3

■s

S41

F4I

F42

543

S43

F44

E4S

F46

F38

I38

547

F47

6. 2
6.8
6.5
6. 2
6-5

. 1

. X

5-5

5.3

5-1

4- 3

5- 3

S' 1
S' 0
S' 4
4.6

5*2

5*0

3*4
5*3
5*3
5*1
5*5
4. 6
4*3
4*5
4.9
5*o

6. 1
6-3
6*3
• 2
6.0

0. 1

. I

2.9

6. 2

• X
4.4
0

• 2

♦ . 1

•4

. X

0. 1

. I

6.6

5-4

5- 9
6.4

6- 3
6.4
6. 0

. 2

5* 0
5-2
5*3
4.8
5*4
5*i
5* 1
4* 8

4. 0
4-8
5*o
4. G
46

5*o
4-6
5* 1

6. 2
7*o
4*6

6*5

6-5
. 1
6. 2

6. 2
0

• X

. 2

. X

. 2

• X
0

. X

. 1
.6
0

. 2

. X

. 2

F47

F48

F49

Fso

F50

FSi

F52

F$2

^53

553

JS3

554

F54

6. 4
5-6
5-7
6. 2
6.6
6.4
6. x
6. 0
5-9
5*7

. 2

4*7

5*9
5*9
5*3
5*i
5* 1
5®
5* 1
5*i
5*o

5*3
5*5
5.0
S- 3
4. 8
4.9
5-o
5*o

4- 7

4.8

5- o

4.9
4.8
S' 2

5*2
. X

5*5
5* x
0

5*o
5*o
S' 2
5*4

6*3

. X

- 1
6. 2
6. 2

6. 1
6.4

6.4

6. 2

6.2

5. 0
5*8
•3
. 2
6. 5
5*7
0

6. 2
0

6. 0
5*5

a*5
3*o
0
. X

. 2

•4

. X

. I

. X

5* 9

3* 0

. X

....

—

. 0

....

5*8

6.5

. 2

5* 2
3*9
5*o

4. 9

5*4

6. x
. 2

5*3

5*9
• X

. X

. 2

. X

•4

S' x

4* 7
4*9
5*4

. 1

Fss

5*6

5* x
5*9

5*3

. X

4* 5

. X

F56

5*8

5* 4

5* 5

. 2

F56

Uro

U10

U12

V13

U14

tlxs

U16

U17

‘ *

5* 1

S' x

• 2

. 2

. X

4* 7
4. 6

5* 2
4. 6

5* 5
3* 1
5*o
5* x
S' x
4*3

S' 1
3*8
* 4

. I

• 2

. X

. 2

. X

◦

5*i

7.2

7*4

4.4

5-8

. X

. 2

5* 2
4.8
4*7
5-x
4*4
4-7

• 2

. 2

•4

•5

3*o

. X

•3
. X

. 2

. I

4.9
5- 1
5*3
. 2

4*5

5*5
4. 8
. X

*3
. 2

. 1

•4
. 2

. X

. 1

•3
. 1
x. 8
x. 8
i*7
•3

. I

* X

. I

•3
. 2

. 1

. . .

6.4

5-8

3*8

5*9

5-4

5-9

5*2

6. 0

5- 7

. 2

. X

•4

4. 1

5*6

5*5

5* 1
4.4

4*7
4.4
4.0
. 1

4-5

4.4

4*7
x* 5

S' 1
4*8
4*7

4.8

4* 7
4*5
4*5

4. 6
5*2

4.9

4* 7
4*5
5*o
4*6

S- 1

5* 5
5*o

5*4

4* 8
5*4
5*5

S' 3
5*5

5*2

5*5

5*4

2. 0

5*x

. X

. 2

•3

5*6

. 1

I. X

. 2

4*8

. X

. I

5*5

2. 0

•5

•3

1*7

. X

•9

•4

. X

. 2

. X

. 2

§9

TP9

Bio

Bii

....

4- 3

5- 7

5- 4
5*8
5-4

6. i

5-9

5* 3
5*4
4*6
4.9

4- 5

4- 0

6. 2
5*8

4.4

1.4

. 2

6. 1

7* 1
6.0
5*9
5*9

4. 4

4*5

7* 1

4* 1

4. 0

3*9

. X

x* 7
3*8
4*3

. 2

. X

6. 2

4* x

. X

• X

3*o

3*3

3*o

3* 7

4* 0
3*8
3*9
3*8
4*6
3*9
3*8
3*9
4*3

4. 2
4*5
4*4

4. 6

. X

. 2

. I

• 5
•3
. 2

. 2

. X

0111

5* 7
5*4

6. 1
6.4

6. 0
6-3

6. 1
6.6

. X

. I

. z

•3

. X

. 2

5*o

•3
*4
•4
. 2

. I

. 2

* I

. X

. I

. X

. 2

Bii

B12

Bl2

E .

G .

6. 1
6.9
6.8
6.3

. I

. X

. 1

. X

4.4

4. 6

3*8

4*o

* I

•3

. 2

Bl3

G .

. X

5* 9

6. 1

5* ^
5*3

• X

0 1

. I

. 2
•3 1

. 2

Mar. 25, 1914

Streptococci in Milk

501

Tab i, 3 II. — Physiological characters of all cultures — Continued.

Culture No.

Origin.

Morphology.

Chains.

Liquefied gelatin.

Neutral red re¬

duction.

Milk curd.

Percentage of normal acid.

Milk.

Broth.

Dextrose.

Adonite.

Saccha¬

rose.

Starch.

Inulin.

Mannite.

Glycerin.

Dulcite.

mm.

B13

—

6.6

0. 1

5*8

5*6

0. 1

3* 7

0. 2

0. 1

U18

....

—

4-9

. 1

•4

4* S

. X

. 1

. 2

U18

5. 2

. 1

4. 3

. 1

U19

4. 9

4. 6

5. 3

• 2

. 2

U19

S* 0

5* 2

. X

. 2

. X

. 1

B14

—

4.8

. 2

6. 2

4- 5

. 1

3*9

. 2

B14

—

4*3

4* 1

4.6

3*5

. X

§IS

—

5*3

4*4

5*0

3*8

2.8

3* 7

. 2

B15

—

5*4

4* 5

5*o

3*9

2.9

3*6

B16

—

6. 0

. 2

1.8

4* 5

. 1

4*3

. X

. 2

B16

—

S-9

•3

1.8

4* 7

. 1

4* 7

. 2

. X

B17

—

4* 7

4. X

4* 7

. X

3* 5

. I

B17

—

4*6

4. 2

4*8

3*9

. 2

•3

B18

—

5*5

4. 2

5*o

3*9

2.8

3* 7

* 5

. 1

B18

—

5* 6

4. 2

4-8

4. O

2.8

3*8

•4

B19

—

4*8

4. 1

4.6

3*3

3*6

. 2

• 0

B19

—

4*8

4*3

4. 8

3* a

3*8

. X

■ X

B20

—

S* *

4.2

4. 8

. 2

3*6

1. 6

. 2

B20

—

5*4

. 1

4*3

4* 8

• 3

3*5

. 1

. 2

B21

—

6. 2

. 1

4.2

5* 1

. X

2.9

B21

—

6. 2

5*3

4.9

. 2

2. 9

In one particular our results do not agree with the conclusions reached
by Stowell, Hilliard, and Schlesinger 1 and by Howe and others in that
the “metabolic gradient” which they establish, in our opinion, can be
correct only for the particular group under consideration, since the
number of cultures utilizing any particular carbohydrates or similar
compound is dependent on the peculiarities of the cultures as well as
on the composition or the configuration of the test substance. While
in a general way our cultures follow the scheme outlined by Stowell,
Hilliard, and Schlesinger, this arrangement may be varied, as will be
pointed out later, by varying the source from which the cultures are
obtained. In one group of our collection a much larger percentage of
cultures give a fermentation with mannite than with raffinose; in others
the conditions are reversed. In no case did we obtain a higher per¬
centage of positive results with mannite than with inulin, although
both Winslow and Stowell, Hilliard, and Schlesinger put inulin above
mannite. Dulcite may be considered as one of the more difficultly
fermented alcohols, and yet in our work on the colon group we found
that dulcite was fermented most frequently, not by the more active
group but by the one which otherwise showed weak fermentative ability.
With adonite the conditions were reversed.

There is among all acid-forming bacteria and especially among the
streptococci considerable variation in the maximum amount of acid
produced. Winslow has shown that this may be a valuable aid in dis-

1 Stowell, E. C., Hilliard. C. M., and Schlesinger. M. J. A statistical study of the streptococci from
milk and from the throat. Jour. Infect. Diseases, v. 12, no. 2, p. 144-164. 1913*

502

Journal of Agricultural Research

Vol. I, No. 6

tinguishing cocci of different species.1 Stowell, Hilliard, and Schlesinger,
in the paper already quoted,2 have pointed out the marked difference
in this regard between streptococci from milk and those from the human
throat. In Table III is shown the distribution of cultures according to
their source and the quantity of acid formed in dextrose broth. This is also
shown graphically in figure 4. The mode for the culture from the mouth
falls over 6.5 per cent, while that for the udder organisms is over 5.0
per cent, and that for those from feces is 5.5 per cent. The mode for each
group is sharply defined, especially those for the udder and feces groups.
On the assumption that the cultures obtained from milk may have
come originally from any of the other sources, we would expect the
curve representing the milk cultures to spread over the space occupied

Fig. 4. — Frequency curves showing acid formation in dextrose broth.

by the other curves. This is true in a general way, but the curve for
the milk cultures has a mode falling between that for the udder and
the feces cultures. It should be remembered that the milk cultures
were not selected promiscuously but from bile tubes incubated at 3 70 C.

Table III. — Distribution of cultures according to the percentage of normal acid pro¬
duced in dextrose broth.

Source.

I Total number
of cultures.

4->

tA,

»d

Above 7.5.

Milk:

Number .

I42

/ 0

Per cent .

\ 0

2.38

II. 90

4- 76

26. 19

16. 67

19. os

II. 90

7- 14

Udder:

Number .

/ 0

Per cent .

l 0

3-02

1. 96

5* 88

1.96

9. 80

41. l8

11. 76

7-84

5-88

I.96

5.88

Feces:

Number .

J-II4

/ 0

Per cent .

l 0

◦

0. 88

2. 63

7- S9

11.40

38.59

20. 17

15-79

2.63

Mouth:

Number .

}39

/ 0

Per cent .

l 0

7-69

12. 82

:i7-95

20. 51

30. 77

IO. 26

1 Winslow. C. E. A., and Winslow, Anne R. Systematic relationships of the Coccaceae. ed. i. 300 p..
illus. New York. 1908.
a Stowell, Hilliard, and Schlesinger. Op. cit.

Mar. E5, 19x4

Streptococci in Milk

503

ACTION ON LITMUS MILK

Late in the course of the investigation it was noticed that there were
distinct differences in the action of different cultures on the litmus in
milk and that this difference was in some relation to the source of the
cultures. Some cultures decolorized the litmus promptly, leaving a
white curd, with the exception of a pink ring at the top, which slowly
extended downward. Other cultures produced a curd which remained
pink throughout for an indefinite period. This action was recorded for
the cultures then available, and the results are given in Table IV. It
will be noticed that while the ability to reduce litmus is characteristic
of the mouth cultures it is almost entirely lacking in the cultures from

Fig. 5. —Graphic representation of the characters of cultures of streptococci from milk and from bovine

feces.

the udder. The number of cultures in the two other groups in which
this character was recorded is too small to permit conclusions, but there
may be observed a tendency in the milk cultures to agree with those
from the udder.

TABhF IV .—Distribution of cultures according to action on litmus in milk.

Cultures recorded from —

Number

Cultures reducing
litmus.

Cultures
failing to
reduce
litmus.

cultures.

Number.

Percent.

Milk .

22. C2

Per cent .

76.57
62. 50
93. IO

17 T A

Feces .

O' 00

27. CO

Udder .

Of' O'**

6.89
82. 86

Mouth .

-y

5°4

Journal of Agricultural Research

Vol. I, No. 6

THE FERMENTATION OF TEST SUBSTANCES

In Table V the cultures are arranged on the basis of fermentation or
nonfermentation of eight test substances. In this table all reactions of
i per cent or over are counted ps positive and those falling below as
negative. The results given in this table are arranged in a form more
easily studied in figures 5 and 6. In these diagrams all positive results
are plotted to the left of a vertical line and the negative results to the
right. The udder organisms are characterized by the general lack of
ability to ferment the test substances. They fail almost without excep¬
tion to ferment raffinose and the polysaccharids, but show some tend¬
ency to attack the two alcohols. On the other hand, the 1 14 cultures

Fig. 6. Graphic representation of the characters of cultures of streptococci from the mouths of cows and

from infected udders.

from bovine feces fail almost entirely to utilize the alcohols, but show
exceptional activity in fermenting the more complex sugars and the
polysaccharids.

Table V. — Fermentation of test substances.

Origin of cul¬

Dex¬

trose.

Saccha¬

rose.

Lactose.

Raffinose.

Starch.

Inulin.

Mannite.

Glycerin.

ture.

| _

Milk:

Total .

P e rcentage
of total. . . .
Udder:

100

4.8

95- 2

2.3

97- 7

4.8

95- 2

69. 0

31-0

80. 9

19. 1

Total .

Percentage
of total. . . .
Feces:

100

78.4

21. 6

94-3

5- 7

100

27.4

72. 6

11. 6

88.4

Total .

1 14

1 12

114

112

P e rcentage
of total. . . .
Mouth:

100

98. 2

1.8

100

81. s

18. 5

52-5

47-5

17. 6

82. 4

18. 5

81.5

1.8

98. 2

Total .

Percentage
of total. . . .

100

89. 7

ii- 3

100

43-6

56-4

7- 7

92.3

25. 6

74*4

87. 2

12.8

2. 6

97-4

Mar. 25, 1914

Streptococci in Milk

505

The cultures from the mouth differ from those from the udder in the
higher percentages of raffinose, inulin, and mannite fermenters and in less
action on glycerin and gelatin. They are sharply differentiated from the
feces organisms in their general failure to ferment starch and the much
higher percentage of mannite fermenters.1

The milk cultures are distinguished by the comparatively small num¬
ber of saccharose fermenters, the failure to ferment raffinose, starch, and
inulin, and the active fermentation of both mannite and glycerin.

THE LIQUEFYING CULTURES

It will be noted that with the exception of a few obtained from milk,
all of the liquefying cultures came from the udder. If we consider the
11 gelatin-liquefying cultures as a group we obtain the data given in
Table VI, which shows that the liquefaction of gelatin is not an isolated
variation from the type but is correlated with an ability to utilize the
alcohols, mannite, and glycerin. This peculiar correlation between gelatin
liquefaction and glycerin fermentation was also noticed in the colon group.

Table VI. — Comparison of liquefying and nonliquefying cultures of streptococci from

the udder.

Item.

Gela¬

tin.

Dex¬

trose.

Sac¬

charose.

Lactose.

Raffi¬

nose.

Starch.

Inulin.

Mannite.

Glycerin.

“

Number of
cultures .
Per cent. . .

zoo. 0

100. 0

90.90

76,74

9.09

23. 26

100. 0

93.02

6. 98

100. 0

9.09

4- 65

90.90

95* 35

4* 65

100. 0

95- 35

M 00

00 Cl

Ov m r-"0

00 H

18. 19

8 3- 72

54- 54

45-46

100.0

Number of
cultures .
Per cent

The characters of the 11 cultures included in Table VI agree very
closely with the ‘ Group C” of the article by the writers on the lactic-
acjd bacteria.2 If we divide the udder cultures into gelatin-liquefying
and nonliquefying groups, we obtain figure 7, in which the cultures are
arranged as in figures 5 and 6. This gives two groups in each of which
the cultures show distinctive characters and remarkable uniformity.

We have, then, at least three sharply defined varieties: Two from
the udder, of which one has weak fermentative ability, attacking only
dextrose, saccharose, and lactose, with an occasional culture-producing
acid from mannite, inulin, or starch, and a second less numerous type,
which liquefies gelatin and ferments dextrose, saccharose, lactose,
mannite, and usually glycerin; and one from bovine feces, character-

1 Since this paper was written, a communication by C. A. Fuller and V. A. Armstrong entitled “ The
differentiation of fecal streptococci by their fermentative reactions in carbohydrate media ” has appeared
in the Jour, of Infect. Diseases, v. 13, no. 3, p. 442-462, Nov., 1913. The characteristics of their cultures
from bovine feces agree in all essential particulars with those found by the writers.

2 Rogers, L. A., and Davis, B. J. Methods of classifying the lactic-acid bacteria. U. S. Dept. Agr., Bur.
Anim. Indus. Bui. 154, 30 p., 6 fig. 1912.

5°6

Journal of Agricultural Research

Vol. I, No. 6

ized by its active fermentation of sugars and polysaccharides and gen¬
eral failure to ferment the alcohols, mannite and glycerin. The group
from the mouth has certain distinctive characters, but is not as clearly
defined as the other three groups. It will need additional study before
it can be described as a distinct variety.

If we consider the milk cultures individually, we find that two of them,
ik and il, clearly belong with the feces group. The one which liquefies
gelatin has the characters of the typical liquefying udder culture.
The remaining 39 cultures may be placed with the nonliquefying udder
organisms. If, however, we assume that the fermentation of mannite
and glycerin places the nonliquefiers with the liquefiers, an assumption

Fig. 7. — Diagram showing the fermentation reactions of two types of udder cultures of streptococci.

based on the possible variation of the liquefying power, we obtain a
division of the milk cultures as shown in Table VII and figure 8. Tljus,
we obtain two groups agreeing very closely with those into which we
were able to separate the udder cultures. This points very strongly to
the infected udders rather than to the feces as the source of chain-forming
cocci growing in lactose broth at 370 C.

Table VII. — Two possible groups of the milk cultures.

Significant charac¬
ters.

Total
num¬
ber of
cul¬
tures.

Gela¬

tin.

Dex¬

trose.

Saccha¬

rose.

Lac¬

tose.

Raffi-

nose.

Starch.

Inulin.

Man¬

nite.

Gly¬

cerin.

Gelatin -f .

I 1

Mannite + .

Glycerin + . .per ct. .
Gelatin — . . . .

l 12- 5

100

87-5

100

Mannite .

1 0

100

Glycerin — , . per ct

40. 6

3* 1

56-3

Mar. 35, 1914

Streptococci in Milk

507

The same test applied to the mouth cultures would show that almost
any individual culture could be included in the feces group. However,
almost any mouth culture would be an exceptional, not a typical, feces
culture. A culture fermenting saccharose, lactose, raffinose, and mannite
could be either from the mouth or from feces, but there is a high proba¬
bility that it would be of buccal origin. On the other hand, a culture
fermenting saccharose, lactose, raffinose, and starch, but failing to fer¬
ment mannite or glycerin, would almost certainly be of fecal origin.

RELATION OF THESE GROUPS TO NAMED VARIETIES

It would be difficult to identify all of these groups with previously
described species. Until the work of Gordon, few cultures were described

Fig. 8. — Diagram showing a possible grouping of the milk cultures of streptococci.

on the basis of the fermentation of a large number of test substances,
and in only a very few cases have the cultures been obtained from a
definite source. An exception may be made of the pathogenic bacteria
in which the cultures described have been selected from definite and very
similar sources. Among the streptococci we have an example in the
pus-forming organism generally described as Streptococcus pyogenes . In
Table VIII are compiled the typical reactions given for Streptococcus
pyogenes by three investigations. The reactions given by Andrewes and
Horder are compiled from a large number of cultures, and those given
by Gordon are from a number of his own cultures.1 2 Those given by
Bergey are the reactions of a comparatively few typical cultures.3 So

1 Andrewes, F. W., and Horder, T. J., A study of the streptococci pathogenic for man. Lancet, v. 2,
no. 11, p. 708-713; no. 12, p. 775-782; no. 13, p. 852-855. 1906.

Gordon, M. H. Report on an investigation of the fermentative characters of streptococci present on
fauces during scarlet fever. 40th Ann. Rpt. Local Govt. Bd. [Gt. Brit.], 1910-11, Suppl. Rpt. Med. Off.,
p. 302-31, 1911.

2 Bergey, D.H. Differentiation of cultures of streptococcus. Jour. Med. Research, v. 27 (n. s., v. 12),

no. 1, p. 67-77. 1912.

508

Journal of Agricultural Research

Vol. I, No. 6

far as it is possible to make comparisons, the reactions given agree very
closely with our nonliquefying udder cultures.

Table VIII. — Results of fermentation tests of Streptococcus pyogenes described in the

literature.

A still further comparison is possible by the tabulation of the fermen¬
tation reactions of five typical cultures of Streptococcus pyogenes obtained
through the courtesy of Prof. C. E. A. Winslow, of the American Museum
of Natural History. These results are given in Table IX. Although
some of these cultures have been grown on artificial media for many
years, they still exhibit the same general characters as our freshly isolated
udder cultures — namely, an ability to ferment dextrose, saccharose, and
lactose, general failure to ferment raffinose and the polysaccharids, and
an erratic tendency to ferment the alcohols. Unfortunately the gelatin
test was not made on these five cultures. The fermentation of glycerin
by three of the five indicates that they may have been of the liquefying
type. Savage in 1 76 cultures of streptococci isolated from cases of mastitis
found that 95 per cent liquefied gelatin.1 His cultures differed from
both the typical S. pyogenes and our liquefying cultures in that 49 per
cent fermented raffinose.

Table IX. — Results of fermentation tests of five cultures of Streptococcus pyogenes from
American Museum of Natural History {New York ) collection .

Source.

Dextrose.

Lactose.

Starch.

Inulin.

Glycerin.

New York Post Graduate Medical College (fatal sep¬
ticemia) .

3*90

3*8 5

3* 60

0. 20

0.18

0. 23

0.35

o- 59

Dr. Bien, Chicago. Ill. (abscess in erysipelas) .

5*45

3*25

• 13

.18

2- 55

2.34

Boston Board of Health (urine) .

3*85

4* 05

•45

3* 98

1-74

Johns Hopkins University .

6. 45

4- 95

4* 70

. 20

•45

.08

4* 91

I. 67

Michigan Agricultural College .

2. 50

1* IS

•OS

.09

•23

.19

VARIATION FROM TYPE IN THE UDDER ORGANISMS

The trouble from infected udders at both the Beltsville and Annapolis
farms was in the nature of an epidemic. The infection apparently
spread from cow to cow and became so severe that at Annapolis one or

1 Savage. W. G.. Report upon the bacteriology and pathology of “Garget*' (or mastitis) in cows. 37th
Ann. Rept. Local Govt. Bd. [Gt. Brit.], 1907-8. Suppl. Rept. Med. Off., pp. 359-424* *909.

Mar. 25, 1914

Streptococci in Milk

509

more cows were rendered useless. There was no apparent connection
between the two epidemics except that they occurred at about the
same time. We may assume that these epidemics originated in one
of two ways, either of which must admit more or less variation in phys¬
iological reactions from the original type. It may be possible that the
udders of one or more cows may have become infected by some of the
streptococci coming originally from the mouth, intestines, or other
sources. Under the influence of its new environment this organism may
have acquired pathogenic properties sufficient to produce the symptoms
observed in mammitis. Heinemann has shown that pathogenicity is a
property readily acquired when ordinary streptococci are grown in ani¬
mals.1 If these infecting organisms came from the mouth, the intes¬
tines, or the milk they must have acquired in a comparatively short time
an -entirely new set of biochemical reactions in addition to a variation
in pathogenicity. On the other hand, we may assume with much more
appearance of reasonableness that the infection spread from a single
infecting cell or aggregate of similar cells which already possessed patho¬
genic powers and general characters identical with those we have found
to be characteristic of the udder organisms. This assumption is in accord
with the established fact that streptococci from pathological lesions in
general have similar biochemical reactions. If the infection in these two
cases came from various sources, it must follow that growth under
similar conditions would produce uniform fermentation reactions in a
short time, a view held by Walker, who maintains that these reactions
may be varied almost at will and can only indicate the latest habitat of
the culture.2 If the infection came from a single cell, there must have
been some variation, since the fermentation reactions were not identical
at the time of this isolation.

In Table X are shown the varieties of nonliquefying udder cultures and
the number occurring in each of the two herds. There were seven varie¬
ties in all. The most numerous one ferments dextrose, saccharose, and
lactose only and occurred 24 times, equally divided between the two
herds. The next most numerous variation differed from the first in
failing to ferment saccharose and occurred 8 times. A third variation
fermented mannite in addition to dextrose, saccharose, and lactose and
occurred 4 times. The remaining four varieties evidently occur only
once or twice in every 40 cultures. Viewed from any standpoint it is
evident that these organisms are subject to variation from the type, but.
these variations are not of sufficient magnitude or frequency to detract
from the value of the physiological reactions as a means of establishing
true species.

1 Heinemann, P. G., The pathogenicity of Streptococcus lacticus. Jour. Infect. Diseases, v. 4. no. 1,
p. 87-92. 1907.

* Walker. E. W. A.f On variation and adaptation in bacteria, illustrated by observations upon strepto* *
cocci, with special reference to the value of fermentation tests as applied to these organisms. Proc. Roy.
Soc. [London], s. B, v. 83, no. 567, P. 541-558. 1911.

Journal of Agricultural Research

Vol. I, No. 6

510

Table) X. — Variation from type in nonliquefying udder cultures.

Significant characters.

Number of cultures
from herd at—

Total num¬
ber of cul¬
tures.

Dextrose.

Saccharose.

Lactose.

Raffinose.

Starch.

Inulin.

Mannite.

Glycerin.

Beltsville.

Annapolis.

H-

—

H-

—

SUMMARY

A collection of cultures of streptococci was made consisting of 42 cul¬
tures from milk which formed chains in lactose bile at 3 70 C., 51 cultures
from infected udders, 114 cultures from bovine feces, and 39 cultures
from the mouths of animals.

The morphology varied under different conditions and could not be
correlated with the source of the culture, except that the udder cultures
had a more marked tendency to chain formation than those from other
sources.

The ability of these cultures to liquefy gelatin and to form acid from
dextrose, lactose, saccharose, raffinose, starch, inulin, mannite, glycerin,
dulcite, and adonite was determined. Only one or two cultures utilized
adonite or dulcite.

When glycerin was attacked, the fermentation proceeded slowly, fail¬
ing to reach its maximum in 14 days, in contrast to the fermentation of the
sugars, in which the maximum was reached in two or three days.

A high percentage of the udder cultures failed to give the character¬
istic reduction in litmus milk.

Twelve cultures liquefied gelatin ; one of these came from milk and 1 1
from infected udders.

The cultures from feces were characterized by their activity in fer¬
menting the sugars, including raffinose, and their inability to utilize the
alcohols.

The mouth cultures fermented dextrose, saccharose, lactose, mannite,
and freqently raffinose, but were almost without effect on starch and
glycerin.

The udder cultures were characterized by the general lack of fermen¬
tative ability, which was limited almost entirely to dextrose, saccharose,
and lactose, with a comparatively small number utilizing mannite,
glycerin, and gelatin.

When the udder cultures were divided on the basis of gelatin lique¬
faction, two groups were obtained. The fermentative activities of one

Mar. 25, 1914

Streptococci in Milk

5ii

of these, which are similar to those of Streptococcus pyogenes , were limited
to dextrose, saccharose, and lactose, with an occasional culture ferment¬
ing mannite, starch, or inulin. The second group fermented the three
simple sugars, mannite, and usually glycerin and liquefied gelatin.

When the milk cultures were considered individually, it was found
that with the exception of two which clearly came from feces they could
be included in one or the other of the two groups into which the udder
cultures were divided.

Of the 41 nonliquefying udder cultures 24 gave identical reactions.
The remaining cultures differed from the type in one or two characters
only.

PRELIMINARY AND MINOR PAPERS

CRYSTALLIZATION OF CREAM OF TARTAR IN THE

FRUIT OF GRAPES

By William B. Alwood,

Chief , Etiological Laboratory , Bureau of Chemistry

During the chemical examinations made of the ripening fruit of grapes
in the Enological Laboratory, Charlottesville, Va., the writer was led to
conclude that the acid salt bitartrate of potassium was deposited from
the juice in quantity sufficient to sensibly affect the analytical results.
This led to the preparation of samples by the complete exhaustion of the
soluble constituents of the berries, with results which supported the
above conclusion.

The question of the character and location of the crystals of cream of
tartar in the berry presented itself as a matter of interest and possibly of
practical importance. The literature available did not furnish specific
information on this point. Babo and Mach, in their exhaustive treatise,
give but one brief reference to the occurrence of this salt in crystals in the
fruit.1

As soon as the fruit was well colored at Charlottesville in 1912 a series
of microscopic examinations was undertaken to determine whether
crystals of bitartrate of potassium occurred in the fruit. Portions of
Concord grapes were prepared and examined daily until the fruit was
ripe. Minute crystals varying much in shape and size were found in
great abundance in the soft cells lying just beneath the skin of the fruit.
Crystals were not present at any time in the pulp or compact portion of
the flesh in which the seeds are contained. Like examinations of Con¬
cord and Catawba were carried on at Sandusky, Ohio, in September and
October, 1912, and crystals of the same general type were found.

The fact that many of the crystals found in the berries did not conform
in type to crystals of the bitartrate prepared from pure cream of tartar
made it doubtful as to whether potassium bitartrate was deposited or not.
Therefore, the fruit was separated into portions for the purpose of a
chemical examination covering this point. The tough pulp containing
the seeds of 1,500 grams of ripe berries was separated from the hulls and
soft peripheral layer of cells which adhere to the hulls. This layer con¬
tains the coloring matter. The hulls and pulp were then carefully pressed
by hand and the juice of each recovered and held separately. This gave
three portions: (1) The pressed hulls, (2) the juice recovered from the
hulls, and (3) the juice recovered from the pulp.

In preparing the sample all the juice possible was recovered from the
sample of hulls and pulp; that is, they were entirely exhausted so far as
crushing and pressing could accomplish this result. The pressed hulls
were then carefully macerated in distilled water until the soluble organic
matter was exhausted. These portions showed on analysis the results
given in Table I.

1 Babo, A. F., and Mach, E. Handbuch des Weinbaues und del Kellerwirtschaft.
p. 16. Berlin, 1910.

Aufl. 4, Bd. 2,

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

(513)

Vol. I, No. 6
Mar. 25, 1914

E— 2

5*4 Journal of A gricultural Research voi, i, no. 6

Table I. — -4 na lyses of Concord grapes in IQI2, giving ihe percentage by weight of acids

and acid salts.

Portion analyzed.

Total acid.

Total tar¬
taric acid.

Free tar¬
taric acid.

Cream of
tartar.

Hulls exhausted with water .

0.429

. 141
1.065

ao. 589
•054
• 724

O. 08

. OO

. 20

0. 56
.07
• 59

Juice pressed from hulls .

Juice pressed from pulp .

° The results show for the samples of “hulls” a greater content of tartaric acid than the total titratable
acid of the samples. This is always the case in grape samples where the “acids other than tartaric” fall
below a certain proportion.

The results show that the juice pressed from the hulls is very low in
acid and acid salts, and that, while the organic matter remaining in the
hulls after pressure is less than half as acid as the pulp, it is rich in tar¬
taric acid and cream of tartar, in these regards nearly equaling the
percentage found in the juicy pulp. The actual weight of the pressed
hulls was 304 grams, or one-fifth of the original sample of fruit. From
the results given, it would appear that the hulls when pressed dry still
retained the crystals observed with the microscope, and actual observa¬
tion has demonstrated this fact. The results for tartaric acid and
cream of tartar settle the point as to the composition of these crystals.

Analyses of like import were made at Sandusky, Ohio, of samples of
Catawba and Concord grapes. The results show that the acid content
of the soft layer of cells attached to the hulls is proportionally richer in
tartaric acid and cream of tartar than the pulp.

In 1913 the microscopic examinations were begun much earlier, and
four varieties of grapes were included — Delaware, Concord, Niagara,
and Norton. The presence of crystals of bitartrate of potassium could
be observed before the berries were all colored, and the analyses of
partly ripe fruit confirm the results of 1912. These samples were sep¬
arated into two portions only, the hulls and the pulp, as noted above;
then each sample was completely exhausted of soluble organic matter
by repeated macerations and heating in distilled water. Table II gives
the results for one set of samples from each of two varieties.

Table II. — Analyses of grapes in IQ13, giving percentage by weight of acids and acid

salts

Concord.

Portion analyzed.

Total acid.

Total tar¬
taric acid.

Free tar¬
taric acid.

Cream of
tartar.

Hulls....
Pulp .

I- 43

al. II

• 79

.04

i- 33
.82

Niagara.

Hulls . .

0. 67

a°. 03

I. 18

Pulp .

.96

• 83

. 18

• 57

* The results show for the samples of “hulls ” a greater content of tartaric acid than the total titratable
acid of the samples. This is always the case in grape samples where the “acids other than tartaric ” fall
below a certain proportion.

There are crystals other than bitartrate present in the fruit, but this
paper is intended only to record an observation which may have peculiar
interest. Further details of the investigation will appear later.

THE REDUCTION OF ARSENIC ACID TO ARSENIOUS
ACID BY THIOSULPHURIC ACID

By Robert M. Chapin,

Senior Biochemist , Bureau of Animal Industry

While endeavoring to work out a practicable field method for the esti¬
mation of the total arsenic — that is, a method which should include both
arsenites and arsenates — in arsenical baths used for dipping cattle, studies
were made upon the effect of various reducing agents which are able to
absorb iodin in acid solution upon the well-known reversible reaction,
As(OH)3+ 2l-f 2H20^±As(0H)6+ 2HI. Unless the solution in which this
reaction is taking place is freely acidified with a strong mineral acid or
heated, the progress of the reaction from right to left is inconveniently
slow. It was found that the addition of sodium thiosulphate apparently
so greatly aided the reduction that it rapidly went to completion, even
in cold and but slightly acid solutions. From this observation it was
but one step to discover that the presence of hydriodic acid played no
part whatever, the reduction of arsenic acid to arsenious acid being
quickly and completely effected by treatment with a mixture of sodium
thiosulphate and mineral acid alone. *

It has long been known that arsenic, like some other metals, may be quan¬
titatively precipitated as sulphid by sodium thiosulphate in a boiling acid
solution. In the present case, however, provided the conditions are right,
there is no formation of arsenious sulphid.

The reactions which may follow from the acidification of a solution of
sodium thiosulphate are complex and variable, depending upon tempera¬
ture, dilution, relative proportions of thiosulphate and acid, and pos¬
sibly upon the order in which the admixture is made. The matter has
most recently been discussed by Stiasny and Das,1 who studied the reac¬
tions between such a mixture and potassium bichromate, a problem simi¬
lar in nature to the one here under consideration.

Preliminary experiments showed that (1) the rapidity with which the
reduction of arsenic acid progresses is mainly dependent upon the con¬
centration of hydrogen ions, the organic acids, except oxalic, operating
very sluggishly, and (2) the nature of the reactions probably depends to
a considerable extent upon whether arsenic or thiosulphuric acid is in
excess and is also varied by the order in which the three components,
arsenic acid, thiosulphate, and mineral acid, are mixed if the operation of
mixing occupies any considerable time.

The present series of experiments was limited to the study of the
reactions occurring when a mixture of arsenic acid, or arsenate, with
excess of sodium thiosulphate is acidified with an appropriate amount of
hydrochloric or sulphuric acid, such being the conditions which must
necessarily prevail in any method for the quantitative estimation of arsenic
which might be based on the reactions. The solutions employed were

1 Stiasny, Edmund, and Das, B. M. Reaction between sodium thiosulphate and a mixture of potassium
bichromate and sulphuric acid. A contribution to the chemistry of chrome tannage. Jour. Soo. Chem.
Indus., v. 31, no. 16, pp. 753~759- 1912.

(51s)

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Vol. I, No. 6
Mar. 25, 19x4
A— 4

5i6

Journal of Agricultural Research

Vol. I, No. 6

(1) a tenth-normal (oxidimetric) solution of arsenic acid prepared by
oxidizing arsenious acid with nitric acid and expelling excess of the latter,

(2) a tenth-normal solution of sodium thiosulphate, (3) a twentieth-normal
solution of iodin, free from iodate, and (4) normal hydrochloric acid.
The equivalents of the solutions were as follows:

Ten c. c. of the solution of arsenic acid reduced, after Williamson, with
hydrochloric acid and potassium iodid and then rendered alkaline with
an excess of sodium bicarbonate required 19.74 c* c* of the iodin solution.

Twenty c. c. of the solution of sodium thiosulphate required 39.50 c. c.
of the iodin solution. To the solution of sodium tetrathionate thereby
resulting there were added 10 grams of dry sodium carbonate and the
solution, loosely covered, was heated one hour upon a steam bath. It
was then cooled, diluted, acidified to litmus paper with acetic acid, and
without delay titrated with iodin solution, of which 39.45 c. c. were
required.

In the experiments to be described a measured quantity of arsenic acid
was diluted to 25 c. c. and was mixed — whether previously neutralized or
not appeared to be immaterial — with 20 c. c. of thiosulphate added from
a burette, and then with 10 c. c. of normal hydrochloric acid added from
a pipette. When containing moderate amounts of arsenic, the mixtures
remained perfectly clear for possibly 15 minutes, disengaging but a slight
odor of sulphur dioxid. After a variable time an opalescence would appear,
rapidly increasing and becoming yellow and accompanied by a pronounced
odor of sulphur dioxid. For quantitative work the action obviously must
be stopped before the separation of sulphur and arsenious sulphid becomes
perceptible. From the considerable number of experiments only enough
will be described to show the natuie of the reactions occurring.

Experiment No. 1. — Ten c. c. of the solution of arsenic acid, 15 c. c. of water, 20 c. c.
of the solution of sodium thiosulphate, and 10 c. c. of hydrochloric acid were mixed
and allowed to stand for 1% minutes. The solution was then titrated with the iodin
solution, using starch indicator (titration I), after which sodium bicarbonate was
added, avoiding unnecessary excess, and titration with iodin continued (titration II).
The end point of titration II was but briefly persistent, owing to the tendency of
sodium tetrathionate to be oxidized to sulphate by iodin in alkaline solution. Next,
10 grams of dry sodium carbonate were added and the solution, loosely covered, was
heated for 1 hour on a steam bath. Then it was cooled, somewhat diluted, acidified
to litmus paper by acetic acid, and immediately titrated again with iodin (titration
III). The results obtained were as follows:

Titration 1 . 20. 50 c. c. of iodin.

Titration II . 19. 75 c. c. of iodin.

Titration III . 35. 25 c. c. of iodin.

Experiment No. 2. —Experiment No. 1 was duplicated, with the single exception
that the mixture was allowed to stand but 2% minutes before titration I was started.
The results were as follows:

Titration 1 . 20. 55 c. c. of iodin.

Titration II . 19. 80 c. c. of iodin.

Titration III . 35. 55 c. c. of iodin.

Titration I removes the excess of reducing agent without affecting any arsenious
oxid present, provided a sufficient quantity of hydriodic acid be also contained in
the solution. To insure this condition, it is safer to add a little potassium iodid just
before beginning titration I, though in case of experiments Nos. 1 and 2 sufficient
was introduced during the titration itself. Titration II measures the arsenious acid
formed by the reduction of arsenic acid.

Comparing now the results of titration II with the iodin equivalent of the arsenic-
acid solution, it is evident that the reaction is quantitative as respects arsenic. Com¬
paring the sums of titrations I and II (x) 40.25 and (2) 40.35 c. c.) with the iodin
equivalent of the thiosulphate solution (39.50 c. c.), it appears that the formation of

Mar. 25, 1914

Reduction of Arsenic Acid

5i7

sulphurous acid is very slight and that the essential reaction involves the formation
of tetrathionic acid, as follows:

As(0H)5+2H2S203=As(0H)3+H2S406+2H20.

A notable formation of any other acids of sulphur would necessarily result in a
markedly higher figure for titration I .

Corroborative evidence of the essential transformation of thiosulphuric acid to
tetrathionic acid is given by titration III, for Stiasny and Das have shown that an
alkali tetrathionate, heated with sodium or potassium carbonate, is nearly quantita¬
tively reconverted to thiosulphate. Titration III shows reformation of a quantity
of thiosulphate equivalent to (1) 35.25 c. c. and (2) 35.55 c. c. of iodin solution, com¬
pared with an originally introduced quantity of thiosulphate equivalent to 39.50 c. c.
of iodin, which amount of thiosulphate, as already noted, after oxidation to tetra¬
thionate, digestion with sodium carbonate, and repeated titration, required 30.45
c. c. of iodin solution.

To further prove the actual reduction of arsenic acid and also that
such reduction is brought about by thiosulphuric acid in the absence
of hydriodic acid, the theoretically possible action of which is not rig¬
orously excluded by the conditions of experiments Nos. 1 and 2, the
following experiment was performed :

Experiment No. 3. — A mixture of arsenic acid, sodium thiosulphate, and hydro¬
chloric acid, made exactly as described in experiments Nos. 1 and 2, was allowed to
stand for five minutes. After the addition of methyl orange, normal caustic soda
was run in until only faint acidity remained, as shown by the orange tint of the solu¬
tion. After the addition of a little sodium acetate and a drop or two of acetic acid to
insure a distinctly acid reaction to litmus paper the solution was titrated cold with
uranium acetate, using potassium ferrocyanid as indicator. The end point was
reached upon the addition of 1 c. c. of uranium-acetate solution. Five c. c. of the
arsenic-acid solution was then added and titration continued. The end point was
again reached upon the addition of 10 c. c. more of uranium acetate, or a total of 11
c. c. Lastly, 5 c. c. of the arsenic-acid solution, treated in a parallel manner, but
without any addition of sodium thiosulphate, required 10.75 c. c* °f uranium-acetate
solution. The previous conclusions regarding the nature and extent of the action
upon arsenic acid were therefore confirmed.

As previously indicated, a small amount of the thiosulphuric acid
suffers decomposition into sulphur dioxid and, presumably, sulphur.
The sulphur does not become evident under the conditions observed,
being partly held in colloidal solution, but for the most part reacting
with tetrathionic acid to form pentathionic acid, as shown by Stiasny
and Das in their investigations already mentioned. The presence of
pentathionic acid was here shown in a similar manner on several of the
mixtures while they still remained clear by neutralizing with caustic
alkali, using methyl orange as indicator. As the neutral point was
reached, a distinct opalescence appeared which was not affected by
hydrochloric acid, but which was dissolved after a time by excess of
caustic alkali.

The action of thiosulphuric acid upon arsenic acid appears, therefore,
at least under the particular conditions studied, to be closely parallel
to the action of thiosulphuric acid upon bichromic acid as described by
Stiasny and Das.

For obvious reasons it is not likely that the reaction here noted,
apparently for the first time, will afford the basis for a desirable volu¬
metric method for use in the laboratory. It may be of value as a con¬
venient means for reducing arsenic acid to arsenious acid preliminary to
precipitation by hydrogen sulphid. As a basis for a field test, in default
of anything better, it does offer some promise, and experiments in that
direction are now under way.

INDEX

Page

Abronia cycloptera , in Tooele Valley, Utah . 387

salsa, in Tooele Valley, Utah . 387

Acid, arsenic, reduction to arsenious acid . 5 x 5—5 1 7

estimation of . 515

arsenious, reduction from arsenic acid . 5 1 5—5 1 7

benzoic, in soil . 357~3SS

metaoxytoluic , in soil . 3 58-3 59

Adaptation in seedlings of Hopi maize . 293-302

Adsorption by soils, selective . 179-188

African cherry orange. See Citropsis.

Agromyza amoena, syn. A. pusilla.

angulata, resemblance to A. pusilla.
blanda, syn. A. pusilla.

carbonaria, relation to A. pruinosa . 471

coquilletti , resemblance to A. pusilla . 84

diminuta, syn. A. pusilla.
exilis , syn. A. pusilla.
fuella, syn. A. pusilla.
fusio , syn. A. pusilla.

malampyga, var. marginalis, resemblance to A. pusilla . 85

orboria, syn. A. pusilla.

pruinosa, cause of pith-ray flecks in Betula nigra . 471-473

pumila , syn. A. pusilla.

pusilla . 67-74

distribution of . 62

enemies of . 76-83

food plants of . 63-64

parasites of . 76-83

strigata, syn. A. pusilla.

virens, resemblance to A. pusilla . 85

Agropyron spicatum, in Tooele Valley, Utah . . . 378, 387

Alfalfa. See medicago saliva.

Alkaloidal content of A tropa belladonna, variation in . 129-146

Alkaloids, percentage in leaves of Atropa belladonna . 132, 134-146

variation in leaves of A tropa belladonna . 141-145

Allenrolfea occidentals, in Tooele Valley, Utah . 413

Allium acuminatum , in Tooele Valley, Utah . 394

Almond, California desert. See Prunus fasciculata.
desert. See Prunus fasciculata.

Havard’s. See Prunus havardii.

Mexican. See Prunus microphylla.

Nevada wild. See Prunus andersonii.

Texas. See Prunus minutiflora.

Alternaria sp., isolation from Triticum sativum . 476

Alwood, W. B. (paper), Crystallization of Cream of Tartar in the Fruit of

. Si3S*4

America, South, potato weevils from . 347~352

Amsinckia tessellata, in Tooele Valley, Utah . 378

Journal of Agricultural Research,

Dept, of Agriculture, Washington, D. C.

Vol. I

Oct., 1913-Mar., 1914

520

Journal of Agricultural Research

Vol. I

Amygdalus andersonii, syn. Prunus andersonii.
fasciculate, syn. Prunus fasciculate,
fremonti , syn. Prunus eriogyna .
glandulosa , syn. Prunus texana.
microphylla, syn, Prunus microphylla.
minutiflora, syn. Prunus minutiflora.

texana, syn. Prunus texana . Page

Anogra albicaulis , in Tooele Valley, Utah . 378

pallida , in Tooele Valley, Utah . 378

Antennaria dimorpha , in Tooele Valley, Utah . 378

Anthonomous grandis, comparison with A. grandis , var. thurberiae . 91

thurberiae, danger from . 96

n. var . 90

Antigen from surra, used in diagnosis of dourine . 101-105

Apricot, desert. See Prunus eriogyna.

Arabis laevigata , food plant of Agromyza pusilla . 64

longirostris, in Tooele Valley, Utah . 378

Arizona, occurrence of a cotton boll weevil in . 89-98

Arsenic acid. See Acid, arsenic.

Arsenious acid. See Acid, arsenious.

Artemisia spinescens , in Tooele Valley, Utah . 394

tridentata, in Tooele Valley, Utah . 377-387

Aster pauciflorus , in Tooele Valley, Utah . 405

Astragalus arietinus, in Tooele Valley, Utah . 378

beckwithii , in Tooele Valley, Utah . 378

utahensis, in Tooele Valley, Utah . 378

A triplex canescens, in Tooele Valley, Utah . 378, 387

confertifolia , in Tooele Valley, Utah . 394-401

nuttallii , in Tooele Valley, Utah . 401

spatiosa, in Tooele Valley, Utah . . 406

Atropa belladonna , percentage of alkaloids in leaves . 132, 134-146

variation in alkaloidal content . 129-146

variation of alkaloids in leaves . 141-145

Avena sativa, imperfect fungi isolated from . 475-489

inoculation with imperfect fungi . 476-481

Bacterium aptatum . . . 189-206

comparison with B. phaseoli . 208

comparison with B. xanthocklorum . 209-210

comparison with Pseudomonas tenuis . 206-208

n. sp . 206

Bacterium Causing a Disease of Sugar-Beet and Nasturtium heaves, A (paper) . 189-2 10

Bacterium phaseoli , comparison with B . aptatum . 208

tumefaciens, causal organism of crown-gall of Carya illinoensis . 334-337

xanthocklorum , comparison with B. aptatum . 209-210

Ballard, W. S., and Volck, W. H. (paper), Winter Spraying with Solutions

of Nitrate of Soda . 437-444

Balsamorrhiza hirsute, in Tooele Valley, Utah . 378

sagittate, in Tooele Valley, Utah . 378

Barley. See Hordeum vulgare.

Bean, hog. See Hyoscyamus niger.

Beet, sugar. See Beta vulgaris.

Belladonna. See Atropa belladonna.

Bellflower. See Campanula trachelium.

Oct., 1913-Mar., 1914

Index

521

Page

Beilis perennis , food plant of Agromyza pusilla . 63

Benzene derivatives in soils . 357_364

Benzoic acid. See Acid, benzoic.

Beta vulgaris , bacterial disease of leaves . 189-210

food plant of Agromyza pusilla . 63

Betula nigra , cambium miner in . 471-474

infestation by Agromyza pruinosa . 473

Birch, river. See Betula nigra.

Bladder senna. See Colutea arborescens.

Blight, nursery, of Carya illinoensis . 305-312

twig, of Quercus prinus . 339-346

Boll weevil, cotton, occurrence in Arizona . 89-98

Brassica napus , food plant of Agromyza pusilla . 64

* injury by Agromyza pusilla . 74-75

okracea, food plant of Agromyza pusilla . 63

rapa, food plant of Agromyza pusilla . 63

Briggs, L. J. et al., (paper), Indicator Significance of Vegetation, etc . 365-418

Bromus marginatus seminudus, in Tooele Valley, Utah . 394

tectorum , in Tooele Valley, Utah . 378, 401

Brown, N. A. and Jamieson, C. O. (paper), Bacterium Causing a Disease of

Sugar-Beet and Nasturtium Leaves . 189-210

Buck, John M., Mohler, John R., and Eichhom, Adolph (paper), Diagnosis
of Dourine, etc . 99-108

Cabbage. See Brassica oleracea.

California desert almond. See Prunusfasciculata.

soil, composition of Triiicum sativum on . 278-282

Calliephialtes cgrbonarius , relation to Calliephialtes sp . 212

comstockii . 214

messor . 213— 214

Calliephialtes Parasite of the Codling Moth, The (paper) . 211-238

Calliephialtes pusio . 214

SP . 211-235

Cambium Miner in River Birch, The (paper) . 471-474

Campanula trachelium, food plant of Agromyza pusilla . 63

Capsicum sp., food plant of Agromyza pusilla . 63

Carduus scariosus, in Tooele Valley, Utah . 406

Carya illinoensis , diseases of . 303-33 8

anthracnose of . 319-330

crown-gall of . 334~337

nursery, blight of . 305-312

Castanea dentata , heart-rot of . 117-119, 121, 127

inoculation with Diplodia longispora . 341

twig blight of . 339

pumila, heart-rot of . 116-117, Ir9

Castilleja linariaefolia, in Tooele Valley, Utah . 378,387

Cells, tyloselike, occurrence in conifers . 461-462

Cerasus minuii flora, syn. Prunus minutiflora.

Cercosporafusca , causal organism of brown leaf-spot of Carya illinoensis . 3 12-3 19

emend, sp . 318-319

Chaenactis douglasii, in Tooele Valley, Utah . 378

Chapin, R. M. (paper), Reduction of Arsenic Acid to Arsenious Acid by Thio-
sulphuric Acid . 515—5x7

522

Journal of Agricultural Research

Vol. I

Page

Chemical characteristics of Triticum sativum , environmental influences on. . 275-292

Cherry orange. See Citropsis.

Chestnut. See Castanea dentata .

Chinquapin. See Castanea pumila.

Chrysocharis ainsliei, parasite of Agromyza pusilla . 79

farbesi , parasite of Agromyza pusilla. . 79

Chrysopsis villosa , in Tooele Valley, Utah . 378

Chrysothamnus graveolens glabrata, in Tooele Valley, Utah . 405

marianus , in Tooele Valley, Utah . 378, 394

nauseosus albicaulis , in Toole Valley, Utah . 378*387

pumilus , in Tooele Valley, Utah . 378, 387

Cirrospilus flavoviridis , parasite of Agromyza pusilla . 81

sp., parasite of Agromyza pusilla . 82

Citropsis, a New Tropical African Genus Allied to Citrus (paper) . 419-436

Citropsis articulata , n. comb . 433

gabunensis, n. comb . 430-432

Citropsis , grafting of . 435

hybridization of . 43 5-43 5

Citropsis mirabilisy n. comb . 432-433

Citropsis , new genus . 421

Citropsis Preussii , n. comb . 423-425

Schiveinfurthii , n. comb . 426-429

Citropsis, possible uses of . 434

Citrullus vulgaris , food plant of Agromyza pusilla . 63

Citrus, alliance to Citropsis . 419-436

Citrus articulata , syn. Citropsis articulata.

Cavalereiy relation to C. ichangensis . n

celebica , relation to C. ichangensis . ; . 10

histrixy relation to C. ichangensis . 10

Citrus Ichangensis, a Promising, Hardy, New Species from Southwestern China

and Assam (paper) . 1-14

Citrus ichangensis latipesf n. subsp . 11

macroptera, relation to C. ichangensis . 10

papuana , relation to C. ichangensis . 10

Cladosporium gramineum , isolated from Avena sativa . 476

Cleomeserrulata, in Tooele Valley, Utah . 406

Closteroceras utahemisy parasite of Agromyza pusilla . 81

Clover, red. See Trifolium pratense.
sweet. See Melilotus officinalis .
white. See Trifolium repens .
zigzag. See Trifolium medium.

Codling moth, Calliephialtes parasite of . . 211-238

Collins, G. N. (paper), Drought-Resisting Adaptation in Seedlings of Hopi

Maize . 293-302

Colutea arborescensy food plant of Agromyza pusilla . 63

Complement fixation, diagnosis of dourine by . 99-108

Complement-fixation test for dourine . 105-107

Coniothyrium caryogenum, causal organism of kernel-spot in Carya illinoensis. . 330-334

n. sp . 334

Cotton. See Gossypium barbadense.

Cotton-boll weevil, occurrence in Arizona . 89-98

Cowania siansburiana, in Tooele Valley, Utah . 378

Cowpea. See Vigna unguiculata .

Oct., 1913-Mar., 1914 Index 523

Page

Crepis glauca, in Tooele Valley, Utah. . 405

occidental is, in Tooele Valley, Utah . . 378

Cream of tartar, crystallization in the fruit of grapes . 5 13-5 14

Creosote, wood penetration of, affected by tyloses . . 464-467

Crown-gall of Carya illinoensis . 334“337

Cryptanthe multicaulis , in Tooele Valley, Utah . 394

Crystallization of Cream of Tartar in the Fruit of Grapes (paper) . 5 13-5 14

Crytanthe sp., in Tooele Valley, Utah . 378, 387

Cushman, R. A. (paper), Calliephialtes Parasite, etc . 211-238

Cypress spurge. See Euphorbia cyparissias.

Cysticercus cellulosae, comparison with C. ovis . 31

confusion with C. ovis . 15

syn. Taenia ovis.
ovipariens, syn. Taenia ovis.
oviparus , syn. Taenia ovis.
ovis. See also Taenia ovis.

Cysticercus Ovis, the Cause of Tapeworm Cysts in Mutton (paper) . 15-58

Cysticercus tenuicollis , comparison with C. ovis . 32-33

confusion with C. ovis . 17

syn. Taenia ovis.

Cysts, tapeworm, in mutton, Cysticercus ovis , cause of . 15-58

Dahlberg, A. O., and Rogers, L. A. (paper), Origin of Some of the Streptococci,

etc . 49I-511

Daisy, garden. See Beilis perennis.

Dandelion. See Taraxacum geniculata.

Delphinium burkei , in Tooele Valley, Utah . 378

Derostenus arizonensis, parasite on Agromyza pusilla . 80

diastatae , parasite on Agromyza pusilla . 80

functiventus , parasite on Agromyza pusilla . 80

pictipes , parasite on Agromyza pusilla . 80

varipes , parasite on Agromyza pusilla . . . 81

Desert apricot. See Prunus eriogyna.

Diagnosis of Dourine by Complement Fixation, The (paper) . 99-108

Diaulinopsis callichroma, parasite of Agromyza pusilla . 81

sp., parasite of Agromyza pusilla . 82

Diaulinus begini , parasite of Agromyza pusilla . 78

websteri , parasite of Agromyza pusilla . 79

Dibothriocephalus spp., comparison with Taenia ovis . 35

Diplodia longispora, causal organism of twig blight of Quercus prinus . 345-346

Dipylidium caninum , comparison with Taenia ovis . 34

Disease, bacterial, of leaves of Beta vulgaris and Tropaeolum majus . 189-210

Diseases of Carya illinoensis . 3°3~338

Distichlis spicata, in Tooele Valley, Utah . 405

Dodecatheon sp., in Tooele Valley, Utah . 405

Dourine, complement-fixation test for . 1 05-1 07

diagnosis by complement fixation . 99-108

Draba sp., in Tooele Valley, Utah . 378

Drought-Resisting Adaptation in Seedlings in Hopi Maize, A (paper) . 293-302

Echinococcus granulosus , comparison with Taenia ovis . 34

Eichhom, Adolph, Mohler, John R., and Buck, John M. (paper), Diagnosis of

Dourine, etc . 99-108

Elder, European. See Sambucus nigra.

524 Journal, of Agricultural Research voi. 1

Page

Elymus condensate , in Tooele Valley, Utah . 401

Emplectocladus andersonii, syn. Prune andersonii.
fasciculate, syn. Prunus fasciculata.

Emplectocladus, subgenus of Prunus .

Enemies of Agromyza pusilla . . 76-8 3

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Journal of agricultural research"

See other formats