1 ' V

• I ...

'^^r^

. : y

I 1 ■ > .

^ •

'•!? ■♦♦-,

5^ 'i .*»'

^* >

■^L

'■' ■^

•^'■"*^

^ "^ . -k-*^

^ ' lf-

^^L M

i?«;ay^

'-V 4 \

Library.

c

OXTRIBUTIOXS

FROM

The Xew York Botanical Garden

\^oGr-ME II

1 902- 1 904

Published by the Aid of the

Davii> Lvdig Fund

Bequeathed by Charles P. Daly

i

PRESS OF

THE NEW ERA PRINTING COMPANY

LANCASTER PA.

LI

TABLE OF CONTENTS i.;.

<-J A K' i-J o .\

1^0 '^No. 26. Chemical Studies of the Cocoa nut with some Notes on
the Chano^es chiring Germination, hy j. i:. Kiikwood
and William J. Gies.

[^o'^No. 27. Some Mt. Desert Fungi, by V. S. White.

\<^(^'; No. 28. Fossil Ferns from the Laramie Gump <>f Colorado, Ijy
Arthur noUick.

1^£);lNo. 29. The Polyporaceae of North America — L The Genus
Ganoderma, hy William Alphonso Murrill.

/7o'r^No. 30. Studies on the Rocky Mountain Flora — 1.\, l)y P. A.
Rydberg.

I'^o^No. 3r. A fossil Petal and a fossil Fruit from the Cretaceous
(Dakota Group) of Kansas, by Arthur Iloliick.

|9oSNo. 32. The Polyporaceae of North America — IL The Genus
Pyropolyporus, by William Alphonso Murrill.

|9^^ No. 33. Studies in the Asclepiadaceae — IV'-VII, by Anna Mur-
ray Vail.

j9(3S'No. 34. The Genus Riella, with Descriptions of new Species
from North America and the Canary Islands, by M.
A. Howe and L. M. Underwood.

|<]i)^No. 35. The Polyporaceae of North America— IIL The Genus
Fomes, by William Alphonso Murrill.

l<^o2)No. 36. Studies on the Rocky Mountain Flora — X, by Per Axel
Rydberg.

1^0 3 No. 37. Some Generic Segregations, by P. A. Rydberg.
|<^o^No. 38. The Polyporaceae of North America — IV. The Genus
Elfvingia, by William Alphonso Murrill.

/ 7^ 3No. 39. A Preliminary Enumeration of the Grasses of Porto Rico,
by George V. Nash.

/^<33No. 40. The Phyllodes of Oxypolis filiformis, a Swamp Xero-

phyte, by Rosina J. Rennert.
/ ^fl3No. 41 . The Polyporaceae of North America — V. The Genera

Cryptoporus, Piptoporus, Scutiger and Porodiscus, by

William Alphonso Murrill.

I^C^^^' 4^' "^ Revision of the Family Fouquieriaceae, by George V.
Nash.

I'^dS No. 43

V

/ 9 6)3^0. 44

/9o3No. 45

/^6 3 No. 46

/f^SNo. 47

/^^^No. 48
/•^^i/No. 49

/^^t/_No. 50

Some Correlations of Leaves, l:)y Daniel Trembly Mac-

Dougal.
Soil Temperatures and Vegetation, by Daniel Trembly

MacDougal.
Studies in Plant Hybrids : The Spermatogenesis of

Hybrid Peas, by William Austin Cannon.
Some Aspects of Desert Vegetation, by Daniel Trembly

MacDougal.
Anatomy and Physiology of Baccharis genistelloides, by

Elsie M. Kupfer.
Mutation in Plants, by Daniel Trembly MacDougal.
The Polyporaceae of North America — VI. The Genus

Polyporus, by William Alphonso Murrill.
The Spines of Fouquieria, by Winifred J. Robinson.

PUBLICATIONS

OF

The New York Botanical Garden

Journal of the New York Botanical Garden, monthly, illustrated, con-
taining notes, news and non-technical articles of general interest. Free to all mem-
bers of the Garden. Toothers, ID cents a copy; $i.oo a year. [Not offered in
exchange.] Vol. I, 1900, viii -j- 213 pp. Vol. II, 1901, viii -|- 204 pp. Vol. Ill,
1902, viii -|- 244 pp. Vol. IV, 1903, viii -{- 238 pp.

Bulletin of the New York Botanical Garden, containing the annual reports
of the I )irector-in Chief and other official documents, and technical articles embodying
the results of investigations carried out in the Garden. Free to all members of the
Garden ; to others, ^3.00 per volume. Vol. I, Nos. I-5, 449 pp., 3 maps, and 12
plates, 1S96-1900. Vol. II, Nos. 6-8, 518 pp., 30 plates, 1901-1903. Vol. Ill,
No. 9, 174 pp., 15 plates, 1903.

Memoirs of the New York Botanical Garden. Price to members of the
Garden, $1.00 per volume. To others, ^2.00. [Not offered in exchange.]

Vol. I. An Annotated Catalogue of the Flora of Montana and the Yellowstone
Park, by Dr. Per Axel Rydberg, assistant curator of the museums. An arrangement
and critical discussion of the Pteridophytes and Phanerogams of the region with
notes from the author's field book and including descriptions of 163 new species,
ix -(- 492 pp. Roy. 8vo, with detailed map.

Vol. II. The Influence of Light and Darkness upo*. Growth and Development,
by Dr. 1). T. MacDougal, first assistant and director of the laboratories. An ac-
count of the author's extensive researches together with a general consideration of
tlie relation of light to plants. The principal morphological features are illustrated,
xvi -|- 320 pp. Roy. 8vo, with 176 figures.

Contributions from tlie New^ York Botanical Garden. A series of tech-
nical papers written by students or members of the staff, and reprinted from journals
other than above. Price, 25 cents each. ^5.00 per volume.

Vol. I. Inclusive of Nos. 1-25, vi -f- 400 pp. 35 figures in the text and 34
plates. .

Vol. 1 1. Nos. 26-50, vi -f 340 pp. 55 figures in the text and 18 plates.

All subscriptions and remittances should be sent to

THE NEW YORK BOTANICAL GARDEN
BRONX PARK, NEW YORK CITY.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 26

CHEMICAL STUDIES OF THE COCOANUT WITH

SOME NOTES ON THE CHANGES

DURING GERMINATION

BY J. E. KIRKWOOD and WILLIAM J. GIES

NEW YORK
1902

[Reprinted from Bulletin Torhey Botanical Club, 29 : Sil-359, 20 June. 1902.}

Chemical Studies of the Cocoanut with some Notes on the Changes

during Germination *

By J. E. KiRKWOoD and William J. Gies

(With Plate 19)

[From the Laboratory of Physiological Chemistry of Columbia University, and
the New York Botanical Garden, New York.]

Synopsis Page.

I. Introduction 322

II. Chemical composition of the ungerminated cocoanut 325

A. Proportions of milk, endosperm and shell in the husked nut 326

B. Composition of the milk 3^^

C. Composition of the endosperm ^^l

a. General composition 33^

b. Fat 335

c. Crude fiber and carbohydrates 340

d. Proteids 34°

e. Inorganic matter (ash) 345

f. Enzymes 345

g. Average composition 345

D. Composition of shell and husk 34^

E. Cocoanut pearls 34°

III. Changes in the cocoanut during germination 349

A. Morphological changes 349

B. Chemical changes 35^

C. Enzymes 35"

* Preliminary accounts of some of the results of this research were given in the
Proceedings of the American Association for the Advancement of Science, — : 275.
1900, and in the Proceedings of the American Physiological Society. 1900 : American
Journal of Physiology, 5: 14. 1901.

The term " coco " appears to be derived from " coc " or " cocus," a local name
for the " Indian nut," the fruit of Cocos nucifera, given to it on account of a fancied
resemblance of the base of the endocarp, with its three circular impressions, to the
face of a monkey when it utters a cry having a sound like the word. See / /,
//. jg. The term "cocoa" should be carefully distinguished from "cacao," the
product of Theobroma cacao, from "coca" the derivative of E'jthroxylon coca, from
"coco," the coco-kola of commerce, and from "cocco" or "cocoa root" [Colorasia
esculenta ) .

[Issued 20 June] 321

322

KiRKWooD AND GiES : Chemical Studies

I.

Introduction

" Of the whole class of

seeds having the character

of luxuries rather than of

necessaries, the cocoanut is

by far the most important to.

mankind, whether considered

as a delicious and nutritious

food or as supplying valuable oil and

many other articles useful in social

life." *

The common cocoanut is derived
from Cocos nucifera, a species of palm
growing in practically all tropical coasts
and islands. The cocoa palm grows
naturally on the seashore or in its im-
mediate vicinity and does not bear well
when at a great distance from salt
water. (See analyses, p. 335.) At
maturity it has a cylindrical stem about
2 feet in diameter. At its apex the
tree carries a tuft of leaves, which are
about 12 feet long. These have num-
erous narrow, rigid and long leaflets.
The leaf, which may attain to 20 feet in
length, consists of a strong mid-rib,
whence numerous long acute leaflets
spring, giving the whole the appearance
of a gigantic feather. The flowers which
produce the nut are yellowish-white.
They are arranged in spikes, branching
from a central axis, and inclosed with
a tough spathe usually a meter or more

Fig. 1. Inflorescence of the cocoanut showing
spathe inclosing the spikes, each with numerous
male flowers above and a single female flower near
the base. X i- Winton.

* Smith : Food, 226. 1873.

OF THE .COCOANUT DURING GERMINATION

323

in length. Their appearance and arrangement is shown in Fig. i,
on the opposite page. The tree grows to a height of about 60-100
feet and usually bears 80-100 nuts arranged on the tree in bunches
of 10-20. It continues to bear during two generations of men.

The fruit is subtriangular-ovoid in form,
about I 2 inches long and 6 inches broad. It is
composed of a thick, fibrous " husk " (exocarp)
and thin, hard " shell " (endocarp), containing a
white fleshy seed, the " meat" (endosperm), with a
thin integument (testa). (See /v^. j, p. 324.) The
thick husk is remarkably adapted to the preser-
vation of the seed, whilst the nut is tossed about
by the waves until it reaches some shore, it may
be, far distant from that on which it grew.
While immature the nut is without the solid
endosperm, but is filled with a milky fluid. As
it ripens, however, the endosperm gradually de-
velops and the milky juice diminishes in quan-
tity. The temperature of this juice when fresh

Fig. 2. Half-
grown cocoanut fruit
with calyx, and axis

is always comparatively low. (See page 349 for from which the male

flowers have fallen.
X i- Winton.

further reference to the parts of the nut.)

Figures i, 2 and 3 are from cuts loaned to
us by Dr. A. L. Winton, who used them lately in the account of his
very valuable histological study of " The Anatomy of the Fruit of
the Cocos nudfera."^- We are greatly indebted to Dr. Winton
for his kind assistance.

The cocoanut forms the chief food of the inhabitants of Ceylon,
the South Sea Islands, the coast of Africa and other tropical coasts
and islands. t The flesh is not only eaten as it comes from the
tree, both ripe and unripe, but it is also prepared and served in
various ways. In India the " copra" is much used as an ingre-
dient of curries. It forms an accessory part of the diet, and is found

* Winton : American Journal of Science, IV. 12 : 265. 1901.

f The cocoanut is agreeable to the taste of various domestic and other animals, and
is eagerly eaten by them. The cocoanut-crab [Birgus latro, suborder Macrura ; an-
omalous form, approaching the Brachyura and closely related to the hermit crab) feeds
almost entirely on the kernel of the cocoanut. Its powerful claws enable it to easily
peel off the husk and open the hard shell.

324

KiRKWooD AND GiES : Chemical Studies

in many of the confections, of civilized man all over the globe.*
Immoderate use of the fruit, which according to the people of
the tropics is highly refrigerant, causes, it is said, rheumatic and
other diseases. t The milk is considered an agreeable cooling
beverage in the tropics. It has been known for some time that
irritation of the mucous membrane of the bladder and urethra is
caused by drinking too freely of the cocoanut milk.^ It is strongly

Fig. 3. Ripe cocoanut. .5", lower part of axis forming the stem ; A, upper end of
axis with scars of male floweis ; Epi, epicarp ; Afes, mesocarp with fibers ; End, endo-
carp or hard shell ; T, portion of testa adhering to endosperm ; Alb, endosperm sur-
rounding cavity of the nut ; K, germinating eye. X \- Winton.

diuretic. Parisi has used the cocoanut therapeutically as an an-
thelmintic with uniformly satisfactory results. § He states that the
meat of the nut is a powerful taenicide, the milk sharing the prop-

*In the Annals of the New York Academy of Science, 13: 490, 1900-1901,
the following may be found : " Dr. Gies in answer to a question stated that the food
content of the cocoanut is small." This answer is quoted incorrectly. The question
referred to the nitrogenous food content. It was stated on that occasion that the
** content oi p'roteid food-stuff is small." See page 340.

t Pavy : A Treatise on Food and Dietetics physiologically and therapeutically
considered, 488. 1878.

{Curtis: Annals" of the New York Academy of Sciences, 13: 490. 1900-1901.

\ See Liebreich : Encyklopaedie der Therapie, i : 744. 1896.

OF THE CoCOANl-T DURING GERMINATION 325

erty.* The cocoanut lias been used as a vermifuge in India for
probably forty generations by the beef eaters of that country and
is there well known as a means of expelling the flat worm.t The
small, green and immature nut is grated fine for medicinal use,
and when mixed with the oil of the ripe nut becomes a healing
ointment.

The fibrous husk (coir) is widely used for the construction of
ropes, brushes, bags, matting, etc. The compact fleshy edible
portion (copra), closely lining the hard shell and which is entirely
fluid or gelatinous when young, contains a large proportion of fat,
which is extracted and used for various commercial purposes, such
as the manufacture of fine soaps and candles and frequently as a
substitute for butter. Cocoanut oil and resin melted together
yield a substance capable of being used with success in filling up
the seams of boats and ships, and in tropical countries for cover-
ing the corks of bottles as a protection against the depredations
of the white ant. A quart of the oil may be obtained from six to
ten nuts. The hard shell is easily polished and lends itself to the
formation of various utensils and ornaments. It also has a high
fuel value.

Although considerable is known of the constituents of the
cocoanut, of its nutritive value and commercial uses, little has
been done to ascertain the nature of the changes which the nut
undergoes during germination. At the suggestion of Dr. Mac-
Dougal we have undertaken such a study, more especially from
the chemical standpoint, and although our work in this particular
connection has not been quite as fruitful as we had hoped it might
be, our results are not without some interest.

II. Chemical Composition of the ungerminated Cocoanut
Before beginning our work on the germinating seed we felt it
desirable to make ourselves thoroughly familiar with the chemical
qualities of the ungerminated nut. This seemed all the more
desirable because of the incomplete as well as the disconnected
chemical data thus far recorded in this connection. This purpose
was accomplished in a large number of analyses of numerous

* United States Dispensatory, 1619. 1899.

t American Journal of the Medical Sciences, 67 : 281. 1889.

326 KiRKWOOD AND GiES : Chemical Studies

samples. We record the more important of these results," with
comparative data from the work of others, on the following pages.

Most of the nuts subjected to the analyses referred to farther
on were furnished to us for this work by Hon. Wm. Fawcett and
the United Fruit Co., who sent them in their husks from Jamaica.
They were ripe, fresh and of about the average size. A few de-
terminations were made with material from nuts bought in the
markets in this city — source unknown, though doubtless-of West
Indian origin. These were of ordinary size, appeared to be ripe
and fresh, and gave essentially the same analytic results as those
obtained directly from Jamaica.*

We wish at the outset of this paper to thank Dr. MacDougal
not only for the supply of material with which he favored us, but
also for the suggestions which led us to undertake this work and
for the kind encouragement he has given us from the beginning.

Proportions of Milk, Endosperm and Shell in the husked
Nut. — The weights and proportions of the main parts of the nut
without its husk were carefully ascertained in special observations,
as well as incidentally in other experiments. f The milk was
removed as indicated on page 328. The empty nut was quickly
broken with a hammer, the endosperm and germ, with the thin
seed coat, carefully and promptly removed with a knife, and the
fresh moist parts weighed at once. The results given on the
opposite page were obtained in this connection. J

The only results recorded on these gross relationships that we
have been able to find were those obtained in a single experiment
by V. Ollech, and those by Bachofeu.§ The parts of a single
cocoanut, except the milk, were dried in the air by v. Ollech. ||

*A few showed signs of deterioration, such as "popping" on opening, free acid
in the milk, etc. These were, of course, discarded.

•f- The weight of the fibrous husk varies considerably, as the amount of moisture in-
creases by absorption or decreases by evaporation. The weights of the other parts are
ordinarily not subject to such fluctuations.

J The weights of the germ and the thin seed coat enveloping the endosperm were
included with the latter.

gSee also Berzelius : Lehrbuch der Chemie. Translated by Wohler, 7: 533.

1838.

II v. Ollech : Quoted by Konig, Die menschlichen Nahrungs- und Genussmittel,
etc., 2: 495. 1893.

OF THE COCOANUT DURING GERMINATION
Freshly importkd Nuts (from Jamaica).

327

t

Weights in Grams.

Percentage of total Weight
of husked Nut.

Milk.

.a
B

Kndosperm

Endosperm

3

Fruit
without

Shell.

with licrm
and

Milk.

Shell.

with Germ
and

Milk.

Vol.

Specific
Gravity.

Husk.

Integument.

30.2

Integument.

51-7

c.c.

150

I

»4S

255

437

153

18. 1

IOI8

2

771

198

379

194

25.7

49.2

25-1

190

1017

3

658

168

371

119

25-5

564

18.1

; "7

1020

4

718

199

351

168

27.7

48.9

234

164

IOI9

5

597

152

327

118

255

54-8

19.7

"3

1022

6

463

127

251

85

27.4

54-2

184

83

1019

7

622

195

334

93

3^-3

53-7

15.0

90

1023

8

563

144

329

90

25.6

58.4

16.0

87

1027

9

633

166

374

93

26.2

59.1

14-7

90

1027

lO

530

15c

282

92

29.4

53-2

174

90

I02I

II

637

150

363

124

235

57.0

19-5

121

1024

12

497

144

267

86

29.0

53-7

17-3

85

IOI4

»3

538

162

283

93

30.1

52.6

17-3

90

1021

14

413

123

256

34

29.8

62.0

8.2

33

1030

IS

5"

158

309

44

30-9

60.5

8.6

43

1037

i6

578

190

320

68

32.8

55-4

11.8

67

IOI6

17

568

142

350

76

25.0

61.6

134

74

)02t)

i8

495

140

293

62

28.3

592

12.5

60

1024

19

813

221

392

200

27.2

48.2

24.6

194

I02I

20

758

208

393

157

27.4

51-9

20.7

150

1022

21

584
6og

148

339

97

254

58.0

166

94

1028

Aver.

169

333

107

27.8

55-2

17.0

104

1023

Nuts from the Market (New York City).

I

1070

250

558

262

234

52.2

24.4

254

IOI8

2

1009

251

506

252

24.9

50.1

25.0

246

IOI5

. 3

728

202

417

109

27.7

57-3

150

106

1027

4

800

226

450

124

28.2

56.3

15-5

120

1026

5

688

191

385

112

27.8

56.0

16.2

no

IOI5

6

565

131

316

118

23.2

56.0

20.8

116

IOI7

7

639

210

382

47

: 32.9

59.8

7-3

46

1024

8

638

210

3"

117

32-9

48.8

18.3

"5

IOI7

9

480

125

304

51

26.0

633

10.7

48

1034

10

561

158

307

96

28. 2

54-7

17.1

92

1024

II

733

204

414

"5

27.8

56.5

15-7

no

1024

12

762

176

380

206

23.1

49.9

27.0
17.8

202

1020

Aver.

722

194

394

134

27.1

551

130

I02I

They represented the following proportions of the total weight,
which was 1,133 grams :

Fibrous Husk. Shell. Endosperm with Germ.* Milk.

30.45 per cent. 19.59 per cent. 37.78 per cent. 12.18 per cent.

Of the total weight of the husked nut, which, by calculation, must

* Including, doubtless, the seed coat as well.

328 KiRKWooD AND GiES : Chemical Studies

have been 788 grams, the percentages of the parts were (calcu-
lated by us) :

Shell. Endosperm with Germ. Milk.

28.17 P^r cent. 54.32 per cent. 17-Sl psr cent.

These results, it will be observed, harmonize closely with the aver-
ages of our own determinations.

The data obtained by Bachofeu in this connection will be
found in the table on page 335.*

Composition of the Milk. — The milk was poured from the
nut through an opening made in the "eye" of the fertile carpel
(see page 350) with a cork -borer. Extraneous matter could easily
be kept out of the milk by this procedure and, besides, the fluid
could be obtained when desired in a perfectly fresh, unevaporated
condition.

The milk was found to be faintly turbid and opalescent in each
case, and always contained a few oil globules and occasionally
crystalline matter. It was acid in reaction to litmus although, as
shown by lacmoid, no free acid was present in the normal fluid.
The reaction is due to acid phosphate. Both alkali and earthy
phosphate are present. The latter can be precipitated, in part at
least, on boiling. An abundant precipitate of phosphate is obtained
when the milk is made alkaline. The average specific gravity,
determined with the aid of a hydrometer, was,, as already noted on
page 327, 1,023 and 1,021. The average specific gravity of the
mixed milk of i 5 nuts not included in the table on that page was
1,023. Of eight additional nuts not referred to there, and ex-
amined at another time, the figures for the mixed milk were

1,022.

The milk reduces Fehling's and Nylander's solutions and it fer-
ments. It contains some monosaccharide which, from the char-
acters of the phenylosazone derivatives, appears to consist of
either dextrose or galactose, probably of both. Disaccharide in
the form of cane-sugar is also present in good quantity, as might
be inferred from the sweet taste of the milk.

* Results having some relation to these are given by Atwater : Report of the
Storrs (Conn ) Agricultural Experiment Station, 123. 1899. Hammerbacher (Land-
wirtschaftlichen Versuchs-Stationen, etc., 18: 472. 1875) found that the endosperm
of two nuts weighed 835.8 grams ; the milk, 303.95 grams. See also, pages 331 and 356.

OF THE Cocoa NUT during Germination

329

On standing the milk turns sour, becomes thicker, and has
much the odor and pliysical appearance of soured cow's milk.
The milk ferments readily. As it does so the acidity increases
with a production of acid from the sugar. Alcohols are also pro-
duced in the process. The distillate from the fermented milk has
an agreeable taste and an alcoholic odor.*

Chlorides are prominent with phosphates among the inorganic
substances of the milk. It contains only a very small quantity
of proteid, coagulating above 80° C, and also traces of a proteose-
like body. Very faint biuret and xanthoproteic reactions were
obtainable with the fresh fluid. A snow-white precipitate consist-

General Composition of the Mu.k

Milk Used.

Percentage of Fresh Milk.

Percentag

Organic
Matter.

'e of Solids.

No.

Specific
Gravity.

Grams.

Water.

Solid Matter.

Inorganic

Total.

4.48

4-57

Organic.

Inorganic.

Matter.

I -a
b

IOI9

28.815

27.280

95-
95

52
43

3-98
4-05

0.50
0.52

88.84
88.60

II. 16
11.40

z-a

b

1020

25 403

27-837

95
95

28
44

4.72
4.56

4.27
4.14

0.45
0.42

90.58
90.70

9.42
9-30

b

1022

30-382
28.528

94
94

73
62

5.27
5-38

4.78
4.90

0.49
0.48

90.57
91.02

9-43
8.98

4-a
b
c

IOI6

25.958
25.823
26.298

95-
95
95

73
65
68

4.27

4-35
432

3.88
3-96
3-91

0.39
0.39
0.41

90.81
90.92
90.57

9.19
9.08

9-43

5-a
b

I02I

29.416
29.467
24.667

95
95
95

II

23

24.

4-89
4-77
4.76

4-47
436
4-35

0.42
0.41
0.41

91-38
91-39
91-31

8.62
8.61
8.69

6-a
b

1024

23.119
23.886

95
95

44
2,2,

456
4.67

3-82
3-92

0.74
0.7s

83.68
84.04

16.32
15.96

b

1028

22 540
26.690

94

94

80
94

520
5.06

4.18

1.02

80.38

19.62

2,-a
b

1027

28.722
28.409

26.847

95
94

02

97

4.98
5-03

4.21
4.26

4.21

0.77
0.77

0.56

84-45
84.69

88.47

15-55
15-31

Aver.

1022

95

•23

4-77

"53

ing in part of earthy phosphate is obtained on warming the milk
on the water-bath at 70° C. The filtrate from this product when
boiled yields a delicate turbidity of coagulated proteid which be-
comes flocculent on addition of a slight excess of acetic acid.
The filtrate from this coagulum gives only a very faint biuret re-
action. Cocoanut milk is said to contain malate of lime.t

*Cocoa beer, containing 3.4 per cent. "Extractive," has been made by Calmette :
Chemisches Centralblatt, 2 : 394. 1894.

t Harley and Harley : Proceedings of the Royal Society of London, 43 : 464.
1887-88.

330 KiRKwooD AND GiES : Chemical Studies

On evaporation to a small volume on a water-bath the fresh
milk becomes darker in color, takes on an odor characteristic of
sugar syrups and looks not unlike molasses. Cane-sugar crystal-
lizes from it in abundance on cooling.

The analytic data given on page 329 were obtained in our
study of the general composition of the perfectly fresh milk of the
Jamaican nuts.*

Percentage results in this connection had been obtained pre-
viously as follows :

Organic

Inorganic

Nitrogenous

Kitrogen-free

Carbo-

Water.

Solids.

Matter.

Matter.

Substance.

Fat.

Extractives

hydrates

9i.5ot

8.50

7.31

1.19

0.46

0.07

6.78

91-371:

8.63

7-5°

113

0.38

O.I I

7.oi§

These results were obtained with milk from nuts grown in the
eastern hemisphere. The milk from the Jamaican nuts appears, as
we have seen, to contain less solid matter, both organic and inor-
ganic. This difference is emphasized by Hammerbacher's T[ ob-
servations on the specific gravity of cocoa-milk. He describes the
milk as a colorless, slightly opalescent fluid with a specific gravity
at 20° C. of 1044.** The milk from two nuts weighed 303.95
grams. From the nitrogen-free extractive substance in 77.8 grams
of milk contained in a third nut, 0.8504 gram of dextrose was ob-
tained. When milk was warmed with dilute sulphuric acid an odor
of volatile fatty acid became perceptible. A crystalline barium salt
was prepared from the distillate of the acidified milk which was found
to consist of barium propionate.

The milk contains a small amount of diastatic ferment and also
oxidase.ft We were unable to detect any other enzymes.

The following results were obtained by van Slyke;|;| in his com-
parative studies of the milk of six unripe nuts and of one ripe one :

* The methods of analysis used for this and similar purposes, throughout our
work, were those commonly employed in the laboratory. See Vandegrift and Gies :
American Journal of Physiology, 5 : 287. 1901.

■\ Hammerbacher : Landwirtschaftlichen Versuchs-Stationen, etc., 18 : 472. 1875.

J Konig : Menschlichen Nahrungs- und Genussmittel, etc., 2: 308. 1893. See
also Bizio : Pharmaceutisches Centralblatt, 756. 1833.

§ Including 4.42 per cent of cane-sugar. See page 328.

\ Hammerbacher, loc. cii.

** See our large number of determinations of specific gravity on page 327. Also
references on pages 328 and 329.

•ft Hunger: Journal of the Society of Chemical Industry, 20 : 1030. 1 901.

JJ Van Slyke : Chemisches Centralblatt, i: 595. 1891. Compare with results on
page 329.

OF THE COCOANUT DURING GERMINATION

331

Constituents.

Weight in grams. . .
Specific gravity . . .
Water (per cent. )
Total solid matter (%)

Inorganic substance.

Glucose

Cane-sugar. . .

"Albuminoid " . .

Fat . .

Milk of unripe Nuts.

230.5
1,024.6

94-37 '
5 -^'3
0.575
4.58

Trace
0.120
0.084

378.6
1,023.0
94.48
5-52
0.635
3.83
Trace.
0.126
o.ioo

347.0
1,022.3

94 59
541
0.675

3-45
Trace.
0.I14
0.138,

383-7
1,023.0
94.89 I

5-M I
0.611

4.06

Trace.

0.205

o 131 '

5

6

350.0

3300

1,022.1

1,021.5

95-27

9643

4-73

3-57

0.658

0.602

4- 36

3-56

Trace.

Trace.

0.140

0095

o.t45

0.120

Average t-6.

336.6
1,022.8
95.01

4-99
0.626

3-97
Trace.

o.«33
0.120

Milk of
ripe Nut.

109.6
1,044.0
9»-23

8.77
1.06.
Trace.

4.42

0.291

0.145

The chief chemical differences induced by growth, a.s indicated
by the above results, are an increase in the proportion of solid
matter, including ash, fat and nitrogenous substance. Glucose
almost entirely disappears from the milk of the ripe nut, cane-
sugar replacing it — a fact evidencing synthetic production of disac-
charide from monosaccharide.

Hammerbacher, believing that the endosperm develops directly
from the milk, determined the quantitative relationships of the
saline matters contained in each part from the same nut. He gives
the following as his percentage results :

Ash of the Milk.

Ash of the Endosperm.*

Potassium,

5S.200

43882

Sodium,

0.728

8.392

Calcium,

3.679

4628

Magnesium,

6.606

9-438

Chlorine,

10.373

13-419

Phosphoric acid,

20.510

16.992

Sulphuric acid.

5-235

5-091

Silicic acid.

0.500

102.331

102.342

Minus oxygen

for chlorine, 2.338

3-024

99-993 99-318

The above results indicate a particular increase of the content of sodium chloride
in the ash of the developing endosperm and a corresponding decrease of potassium
phosphate. See pages 322 and 335. The amount of silicic acid in the endosperm
is also noteworthy. See page 335.

Endosperm. General Co'nposition. — The pure white kernel

* Compare with results of Bachofeu's analysis, given on page 335. Our own re-
sults were the same as these qualitatively. See also Schaedler Technologic der Kette
und Oele des Pflanzen- und Thierreichs, 840, 1S92, who found 3.60 per cent, of iron
in the ash of the endosperm in addition to the above constituents.

332

KiRKwooD AND GiEs : Chemical Studies

or "meat" of the nut is fibrous in structure, closely lines the
shell, is from i to 2 cm. thick, and contains a very large propor-
tion of fat. It is the part used most frequently for dietetic pur-
poses. It possesses a characteristic and pleasant odor and is
very agreeable to the taste. The endosperm cells do not contain
starch granules, but fat needles and proteid lumps are present in
them. The proteid particles are partly crystalline.*

After the kernel has been finely divided in a meat chopper, the
resultant hash may be subjected to increasing pressure, when an
General Composition of the Endosperm

Endosperm
used.

P<

:rcentage of fresh Endosj

;rm.

1

Percentag

^e of Solids.

Organic
Matter.

No.

Grams.

Water.

Solid Matter.

Inorganic

Total.

Organic.

51.19

56.79
52.34

Inorganic. 1

b
c

8.467

9.728

10.900

47.70
42.10
46.60

52.30
57.90
53-40

I. II
III
1.06

97.88
98.09
98.01

2.12
1. 91
1.99

2-a

b
c

II.S85
12. 151
11.707

48.31
48.90
52.29

51.69
51.10

47.71

50.65
50.01
46.61

1.04
1.09
1. 10

98.01
97.87
97.69

1-99
2-13
2.31

3-^
b

c

8.762
8.185
8.923

43-90
47.73
46.31

56.10
5227
53.69

55.20
51.20

52.71

0.90
1.07
0.98

98.39

97-95
98.18

1. 61

2.05
1.82

4-a
b
c

II 511

9-501
9.244

47.89
46.90
4750

52 11

53-IO
52.50

51-05
52.05
51-43

1.06
1.05

1.07

97.97
98.02

97.96

2.03
1.98
2.04

5-«
b

8.942
9.312

42.80

43-79

57.20
56.21

56.17
55.21

1-03 1

1. 00

i

98.21
98-23

1.79

1.77

6-a
b

10.214
10 624

50.30
48.70

49.70
51-30

48.68
50.28

1.02
1.02

97.95
98.02

2.05

1.98

7-a
b

10.746
10.142

10.052

42.21
39.60

57.79
60.40

53.69

56.83

59-46
52.66

0.96
0.94

98-34
98.45

98.07

1.66
1-55

Aver.

46.31

I -03 i

1.93

oily juice is obtained from it. The filtrate from this turbid mixture
has a higher specific gravity than the milk of the nut, is acid in
reaction, reduces Fehling's solution, contains a dextrin-like body
and the milk salts, gives the proteid color reactions, yields coag-
ulable proteid, and on dilution with water becomes turbid from
precipitated globulin.

The data given above were obtained for general composition of
the endosperm immediately after the nuts were opened. f

* See pages 342 and 352.

f The methods were the same as those employed with the milk. The thin seed-
coat was trimmed off and the pieces of kernel cut into small, thin pieces with a knife.
The material was taken from all parts of the nut.

OF THE COCOANUT DURING GERMINATION'

iVMi

Comparison of the averages given on the opposite page may
be made with the following previously recorded results for the
fresh endosperm from nuts of eastern origin : *

Fresh Endosperm

Water. Total Organic

Solids. Matter.

46.64 53.36 52.39

Inorganic

Matter.

0.97

Dry Kndosperm

Organic Inorganic

Matter. Matter.

98.20 1.80

The agreement is seen to be very close.

By reason of the dietetic and commercial values of the various
constituents of the endosperm of the cocoa fruit, numerous prod-
ucts of the kernel have been made and analyzed. The air-dried
endosperm, or so-called " copra," is shipped in large quantities from
the tropics. Cocoa-oil is obtained from the copra by various
methods in countries distant from the tropics, the solid residues
remaining after extraction serving various purposes. This residue
makes up the so-called "cocoa-cake" obtained in the process of
expressing the oil at various degrees of temperature. It is also
eround into "cocoa-meal." In both forms, the residual substance

Water.

Total
Solids.

Organic Matter.

Products Analyzed.

Nitrog-
enous
Sub-
stance.

Fat.

N-free
Extrac-
tive.

Crude
Fiber.

In-
organic
Matter.

Air-dried endosperm or copra. f
Endosperm, perfectly dried. J
Endosperm, free from fat and

water. J
" Cocoa-cake." ?
" Cocoa-meal." |1
♦' Cocoa-meal," after extraction

of oil. Tf

5-81

94.19' 8.88

67.00
67-35

11.00
10.88

1-85

12.44
I5.II

46.25
3S-70

35-34
64

4.06

5.42

16.69
14.40
17.40

45

1. 81
1.83

5-57
5.90

7-32
5 95

10.30
II. 12

4-55

100.00

89.70

88.88
95-45

31-49
19.70

17.94
23.20

1875-

Also

* Hammerbacher : Landwirtschaftlichen Versuchs-Stationen, etc., 18: 472.
See also Bizio : Pharmaceutisches Centralblatt, 757. 1833.

\ Konig : Menschlichen Nahrungs- und Genussmittel, etc., 2 : 652. 1893.
p. 308.

J Hammerbacher : Landwirtschaftlichen Versuchs-Stationen, etc., 18 : 472. 1875.

^Dietrich und Konig: Zusammensetzung und Verdaulichkeit der Futtermittel;
2 : 1031. 1891.

II Dietrich und Konig : i7>u/., l : 725.

Tl Schaedler : Technologic der Fette und Oele des Pflanzen- und Thierreichs,
624. 1892. (a) For references to digestibility and nutritive value of cocoa-cake see
results of experiments on pigs and sheep given by Dietrich and Konig, 2 : 1031, 1036,
^040, 1 123. {i) Compare above results with the table for general composition on
the opposite page.

334

KiRKWooD AND GiES : Chemical Studies

is used as food for cattle and as a fertilizer, having special value in
both these connections.* It is sometimes also used illicitly as a
food adulterant.

The analytic percentage results on page 333 have been reported
by various agricultural chemists for such products from nuts grown
in the eastern hemisphere.

The following summary of facts connected particularly with food
value was given several years ago by Woods and Merrill : f

Edible portion,
As purchased,

Without milk, as purchased,
Cocoanut milk, as purchased.
Shredded cocoanut,
Shredded cocoanut,
Edible portion,
Cocoanut milk,

in

V

a

u

•0

V

T3

>1

J3

u

Pi

V

1

V
1

1

1

14. 1

5-7

50.6

27.9

1-7

48.8

7.2

2.9

25.9

14-3

•9

37-3

8.9

3(^

31.7

17-5

I.O

92.7

•4

i-S

4.6

.8

4-3

6.5

63.7

24.1

1.4

2.8

6.0

SCO

39-0

1.2

5-8

8.9

67.0

16.5

1.8

91-5

•5

.1

6.8

1.2

V O

V "
3 •-

^ 3 in

1! g..H

2,986

1,529
1,872

97

Through the kindness of Dr. MacDougal we have been able
to examine an account of " The cocoanut and plant vitality " in the
Bulletin of the Botanical Department of Trinidad (July, 1900, p.
249). Reference is therein made to the report of Bachofeu in the
Queensland Agricultural Journal for April, 1900. Bachofeu says :
"Though there exist several analyses of parts of the cocoanut, no
one seems to have undertaken the task of getting a complete
analysis made with the view of ascertaining the actual demand
made by the cocoanut upon the mineral constituents of the soil."

The results obtained by Bachofeu for a single nut are so com-
plete, and so general in their interest and application that we quote,
on page 335, his general summary in its entirety. |

Bachofeu' s results indicate that sodium chloride and potassium
phosphate are the chief inorganic matters drawn upon in the de-
velopment of the cocoanut — chemical data in harmony with the fact

*v. Knieriem : Chemisches Centralblatt, 2: 672. 1898.

f Woods and Merrill : Bulletin, Maine Agricultural Experiment Station, No. 54 ;
81. 1899.

J The analyses were made in Ceylon. Native nuts were used.

OF THE COCOANUT DURING GERMINATION

335

that tiic cocoa palm docs not thrive away from the coast or where

See second table, page 331.

salt is lacking in the soil

Bachofeu's Analysis of the Cocoanut

Total weight in lbs.

" " in per cent.
^ j Moisture in per cent.

\ l>ry matter in per cent.
Pure ash in per cent., containing viz :

Silica, SiOj.

Oxide of iron and alumina, FejOjAl.^Og.

Lime, CaO.

Magnesia, MgO.

t Potash, K.p.

Soda, Na.^0.

f Potassium chloride, KCl.

Sodium chloride, NaCl.

Phosphoric acid, PjOj.

Sulphuric acid, SO3.

I Containing total potash, KjO.
* Containing nitrogen, N.

Husk.

Shell.

Kernel.

2.702

0.546

0.875

57-28

"-59

18.54

65.56

15.20

52.80

34.44

84.80

47-20

1-63

0.29

0.79

8.22

4.64

I-3I

0.54

1-39

0.59

4.14

6 26

3.10

2.19

1.32

I 98

30.71

45.01

45-84

3-19

15.42

Milk.

0-593
12.58

45-95
1.92

3-13

100.00

15-56
4.64

5-75

99-99

30-71
0.137

45.01

O.IOO

1304
5.01

20.33

8.79

99-99

54.05
0.504

0.38

2.95
Trace.

7-43
3-97
8.62

41.09

26 32

5.68

3-94

100.00

34-54

Thus of the more important ingredients of the soil 1,000 nuts remove the following ;

In Lbs.

Husk.

Shell.

Kernel.

Milk.

Total Lbs.

Nitrogen, N.

3-7017

0.5460

4.4100

8.6577

Phosphoric acid, P^Oj.

0.8456

0.0735

1-4053

0.1279

24523

Potash, KjO.

13-5255

0.7127

3-7362

0.7783

18.7527

Lime, CaO.

1.8234

0.0991

0.2143

0.1674

2.3042

Sodium chloride, NaCl.

.20.2375

0. 2464

0.3563

0.5431.

21.4233

Fat. — The striking chemical characteristic of the endosperm is
its large content of oil. This may readily be extracted with fluids
like ether. It can also be obtained in large proportion by pres-
sure, particularly at the tropical temperatures. The fat has the
consistence of butter in northern countries and possesses, when
fresh, a fragrant and characteristic odor and an agreeable taste. It
is snow white, sometimes cream-colored and readily crystallizes in
large rosettes from the molten condition or from its alcohol or
ethereal solutions. These crystals closely resemble those of pal-
mitic acid. They melt at about 20-23° C, and congeal again several
degrees below the melting point. They are fairly soluble in cold
alcohol. Although cocoa-fat differs somewhat in composition in
different countries, it has been found that the variations are compara-
tively slight. The temperature at which the oil is expressed influences

336 KiRKwooD AND GiES : Chemical Studies

these variations by increasing or decreasing the proportion of fats
melting only at higher temperatures. These facts account for the
variations in the figures given for melting point. Its specific grav-
ity is 0.9 + .

On heating to about 170° C, the oil gives off the odor of lactic
acid ; at a temperature of 300° C. acrolein may be detected. On
long-continued heating with nitric acid the following dibasic acids
are formed : succinic, adipic, pimelic, suberic and azelaic. Nitro-
caproic acid is also formed. * The oil is very soluble in all of the
well-known fat solvents. It contains some free fatty acid, but con-
sists chiefly of glycerides of caprylic, lauric, myristic and palmitic
acids, t Glycerides of caproic and capric acids are present in ap-
preciable quantity ; also a trace of stearin and some olein. | The
fat dissolves readily at a comparatively low temperature in an equal
quantity of glacial acetic acid. Such a solution becomes turbid at
40° C.§ By reason of its content of lower fatty acid radicles
cocoa-oil has a high saponification value. Cocoa-oil is particularly
resistent to the hydrating effect of superheated steam. ||

The following data were obtained for the percentage fat-content
in the fresh endosperm. The method of determination used was
Dormeyer's :^

Gen'l
123456 7 Average.

Fresh endosperm, a. 38.27 40.01 36.71 35.10 34.60 38.90 38.60
/^. 36.14 40.54 35.02 34.90-34.10 40.70 38.40
Average, 37.20 40.28 35.87 35.00 34.35 39.80 38. 50 37.29

The ether extracts containing the oil were free from lecithin

*Schaedler: Technologic der Fette und Ode des Pflanzen- und Thierrichs, 843.
1892.

f Ihe presence of palmitin (tri) is doubted by Ulzer : Chemisches Centralblatt,
II : II43. 1899.

J The so-called " cocinic acid" or " cocostearic acid" derivable from " cocin "
or " cocinin " is, like the latter, a mixture. The former is a mixture of some of the
above fatty acids ; the latter of their glycerides. See Oudemans : Chemisches Central-
blatt, 192. 1861.

^ Valenta. Quoted by Vaubel : Physikalischen und chemischen Methoden quan-
titativen Bestimmung organischer Verbindungen, i : 162. 1902.

II Klimont : Journal of the Society of Chemical Industry, 21 : 126. 1902.

^ Dormeyer: Jahresbericht iiber die Fortschritte der Thier-Chemie, 26 : 42. 1896.
The fresh tissue was finely divided and weighed, then dried to constant weight at
100-105° C. , and all of it extracted with anhydrous sulphuric ether. The usual amounts
of tissue were used.

Grams.

Per Cent

I5.I4S8

60.595

9.5282

38.113

0.5596

2.238

OF THE COCOANUT DURING GERMINATION 337

and could be almost entirely .saponified, ilammcrbacher * in the
saponification of 25 grams of the pure oil obtained the following
results :

Fatty acids convertible into insoluble lead salts,
Fatty acids convertible into soluble lead salts,
Glycerin,

Total, 25.2366 100.946

The excess in weight of products is explained by the addition
of hydroxyl groups in the cleavage of the triglycerides. Konig
had previously found the glycerin content of cocoa-fat to be 2.08
per cent. Hammerbacher therefore concludes : " It follows from
these results that this v^egetable fat consists in greatest part oi free
fatty acid."

That there is some error in this conclusion, however, is evident
from the results of later work. Benedikt f reports the glycerin
content of cocoa-oil to be 1 3.3-14.5 per cent. Crossley and Le
Suer found that the content o{ free fatty acid in terms of oleic acid
varied between 2.50 and 8,86 per cent.;}:

Stellwaag || studied the fat extracted from cocoa cakes. This
oil was rancid, of course. He found the quantity o{ free fatty
acid to be only 9.84 per cent. The fat from the ether, extract
melted at 23° C. The saponification figure was 244.4. The
extract contained 81.14 per cent, of neutral fat. The amount of
unsaponifiable matter was o. 5 i per cent. The molecular weight
of the fatty acids was given as 207.^

Studied through the oleo refractometer of Amagat and Jean,
cocoa-oil is found to rotate to the left like an animal fat.**

The composition of cocoa-oil as determined by Konig ft 's :

c. H. o.

74.15 per cent. 11-73 P^"" '^^"t- 14.12 per cent.

* Hammerbacher : Landwirtschaftlichen Versuchs-Stationen, etc., 18:472. 1875.

I Benedikt und Zsigmondy : Chemiker Zeitung, 9 : 975. 1885.

% Crossley and Le Suer. Quoted by Hopkins : Oil-Chemists' Handbook, 38, table
iv. 1900.

II Stellwaag : Landwirtschaftlichen Versuchs-Stationen, etc., 37 : 135. 1S90.

\ See also Konig, Menschlichen Nahrungs- und Genussmittel, etc., 2 : 389. 1893.
** Blyth : Foods, Their Composition and Analysis, 359. 1896.
tt Konig : loc, cit., 2 : 3S5. See also, Brandes, I'harmaceutisches Centralblatt,
751- 1838.

338 KiRKWOOD AND GiES : Chemical Studies

The following facts regarding cocoa-oil have been compiled
from various sources. They may be compared with similar data
for other fats and oils given in the standard works of Konig,
Staedeler, Lewkowitsch and others :

A. The heat of combustion of cocoa-oil is 9,066 small calories
per gram.* It is as low as that of any other fat ; slightly lower
than butter. This is due to the fact that it contains a large pro-
portion of fatty acids of low molecular weight.

B. Melting point is at 24° C. Congealing temperature is 22-
23° C. Fatty acids from it melt at 24.6° C. They congeal at

19° c.t

C. Saponification value = 257. 3-268. 4!

D. Iodine number = 9.O-9.5 ; same for its fatty acids = 8.5-

9-0.§

E. Specific gravity = 0.9 11 5 at 40° C.||

F. Acid value = 9.95-35.21.

G. Reichert-Meissl figure = 7.4 ; Hehner = 88.6-90.5.
H. Barium figure (Konig-Hart) = 117-120.

I. Molecular weight of the mixed fatty acids = 196-21 1.

The use of cocoa-fat and other cheap vegetable oils as a
substitute for butter among the poorer classes has been in-
creasing. Cocoa-fat is better adapted for cooking than for table
use. It is frequently employed as an adulterant of ordinary
butter. Prepared cocoa-fat makes a fairly good substitute for com-
mon butter. The fresh material becomes rancid after a time,
because of its accumulating content of free fatty acid resulting from
bacterial agency. Volatile acids are formed. Its tendency to
rancidity is not as great, however, as that of animal fats. The
fatty acid present in the fat to begin with can easily be removed
with insoluble basic compounds, such as magnesia. By this means

* Merrill. Quoted by Sherman and Snell : Journal of the American Chemical
Society, 23 : 166. 1901.

\ Konig : Menschlichen Nahrungs- and Genussmittel, etc., 2 : 322. 1893.
J Konig : /die/.

I Benedikt. Quoted by Vaubel : Physikalischen und chemischen Methoden
quantitativen Bestimmung organischer Verbindungen, 2 : 235. 1902.

II Values given after E-I inclusive are quoted by Hopkins : Oil-Chemists' Hand-
book, 38, table iv. 1900. See also Lane : Journal of the Society of Chemical
Industry, 20 : 1083. 1901.

01- THE COCOANUT DURING GERMINATION 339

a " butter" is made from this oil which has the merit of enduring
hot climates without becoming rancid. This i)roduct has been
recommended for military and naval uses.*

Among the prominent commercial products is the cocoa-butter
made in Mannheim, Germany, f Konig ;}: found this product to
have the following percentage composition :

Nitrogenous
Water. Solids. Organic Matter. Inorganic Matter. Fat. Fatty Acid. Substance.

0.15 9985 99.848 0.002 99.848 trace. trace.

It has been stated that cocoa-butter is not very easily digested
and that it does not agree with sick people. § The recent re-
searches of Bourot and Jean, || however, show that a specially
prepared cocoa-butter melting at 3 1° C. and containing only a trace
of free fatty acid, is quite as easily and completely digested as
ordinary butter.TJ

We have already alluded to some of the commercial uses to
which cocoa-fat is put. Soaps made from it combine with or hold
an unusual amount of water while still retaining special hardness,
one pound of the oil yielding three pounds of soap.** It is thus
well adapted for the preparation of toilet soaps. The soaps made
from cocoa-oil are characterized by great solubility in salt solution
and can be precipitated from such fluid only by the addition of a
very large excess of salt. The so-called "marine" or "salt water
soap" has the property of dissolving as well in salt water as in
fresh water and is made of cocoa-oil and soda. ft

*Rusby: Reference Handbook of the Medical Sciences, 3 : 164. 1901.

■j-See Leffman and Beam : Select Methods in Food Analysis, 182. 1901.

+ Konig : Menschlichen Nahrungs- undGenussmittel, etc., 2 : 309. 1893. See also
Schaedler, Technologic der Fette und Oele des Pflanzen- und Thierrcichs, 1340. 1892.

g Liebreich : Encyklopaedie der Therapie, i : 744. 1896.

II Bourot und Jean : Jahresbericht uber die Fortschritte der Their-Cheniie, 26 :
58. 1896. See also V. Knieriem, Chemisches Centralblatt, 2 : 672. 1898.

^ "Cocoanut cream," a dietary product much used in the tropics, is made by grat-
ing the endosperm and squeezing through cloth the fluid from the finely divided
material. In a warm climate the resultant mixture contains much oil and is a very
delicious accessory food. Besides the oil, the " cream " contains chietly carbohydrate
and proteid. See page 332 for references to similar fluid obtained from the endosperm
by pressure in our own experiments.

** Ebermayer : Physiologische Chemie der Pflanzen, 344. 1882. See also Joss,

Pharmaceutisches Centralblatt, 449. 1834.

ft See Schaedler, Technologic der Fette und Oele des Pflanzen- und Therreichs,
1178-1188, 1892, where may be found the results for percentage composition of the
sodium soap, given at the bottom of the next page :

340 KiRKWOOD AND GiES : Chemical Studies

The harder fats of the oil make excellent candles. They are
used also as constituents for suppositories and related therapeutic
products. Medicinally the oil is employed repeatedly as a substi-
tute for lard, ohve oil and cod-liver oil. It is also made the chief
substance by bulk in various salves and in cold cream, pomade and
similar cosmetic preparations. In ointments and cerates it is
especially valuable because of its ready absorption when rubbed
on the surface of the body ; further, it takes up an unusual amount
of water — a useful quality when it is desired to apply saline solu-
tions externally. It shows little tendency to produce chemical
changes in substances with which it may be associated.

Crude Fiber and Carbohydrates. — Cellulose is a prominent con-
stituent of the endosperm. Associated with the fibrous elements
is a polysaccharide, present in comparatively large quantity. This
substance is only slightly soluble in water, is insoluble in alcohol,
but readily soluble in salt solution. It is precipitated along with
globulin when saline extracts of the kernel are dialyzed ( page
341). The gum is readily transformed into sugar by the action
of diastase or ptyalin.

The fluid pressed from the finely divided endosperm contains
a slight amount of reducing sugar — dextrose. Galactose appears
to have been identified also.* Cane-sugar is also present.

The following results were obtained in our determinations of
the percentage content of crude fiber in the fresh tissue : f

12 345 General Average.

Fresh endosperm, a 3.96 3.20 2.98 3.40 2.78

b a,.ii 3.80 3.12 3.52 2.98

Average, 4.08 3.50 3.05 3.46 2.88 3.39

Proteids. — That the meat of the cocoanut contains at most
only a very small amount of proteid matter is seen at a glance
from the following percentage results for content of nitrogen. |

Water. Fatty Acid. Sodium Oxide (combined). Sodium Oxide (free). Other Salts. Residue.
58.74 . 32.82 4.26 1.50 2.26 0.42

See also the Dispensatory of the United States of America, 1899 : 1619, for ref-
erences to objectionable chemical qualities of some cocoa-soaps.

* Green: Soluble Ferments and Fermentation, 100. 1899.

■\ Determinations were made, after the fresh weighed material had been dried and
thoroughly extracted with ether, by the method adopted by the Association of Official
Agricultural Chemists : Bulletin, Division of Chemistry, U. S. Department of Agricul-
ture, 46 : 26.

% In these determinations the Kjeldahl method was employed.

OF THE COCOANUT DURING GERMINATION 341

General
* * 3 4 S 6 7 Average.

Fresh eiulosperm, rt 0.657 0.734 0.806 0.738 0.766 0.776 0.701
b 0.740 0.781 0.756

Average, 0.657 0-734 0.806 0.738 0.753 0.778 0.729 0.742

The fresh endosperm contains an average of 0.74 per cent, of
nitrogen which, multiplied by the usual factor, 6.25, would indi-
cate 4.63 per cent, of " albuminoid." Some of this nitrogen,
however, is undoubtedly closely associated with the fibrous ele-
ments. Much of it probably is in the form of nitrogenous ex-
tractive.* Some of the nitrogenous substance is soluble in 95 per
cent, alcohol.

. The proteid present in the endosperm appears to consist chiefly
if not exclusively of globulin and proteose (globulose ?), the globu-
lin predominating in quantity. f We have made several samples of
cocoa globulin by the method used by Osborne for the preparation
of edestin — in general as follows : % The kernel was run through
a hashing machine and the finely minced substance freed from fat
by repeated extraction in ether for several days. The ether ad-
herent to the tissue was evaporated at room temperature and the
ether-free tissue then extracted in lo-per-cent. salt solution for 24-
48 hours. The saline extract was then filtered off and globulin
thrown from its solution either by the dilution process, by dialyz-
ing for several days in running water, or by treatment with am-
monium sulphate to complete or half-saturation. The deposit of
globulin resulting thereby always contained an appreciable amount
of gummy carbohydrate. The carbohydrate admixture was elimi-
nated by subjecting the deposit to the action of diastase or ptyalin
for 24-48 hours, in the presence of thymol at 45° C. in neutral
fluid, during which time it was transformed into soluble reducing
sugar. § The globulin residue left behind after this treatment was

* The factor 6.25 is here too large, also, because the proteids present contain about
l8 per cent, of nitrogen. See pages 343 and 344. Stutzer found that, of the total
nitrogen of cocoa-cakes, from 1.8 to 6.9 per cent, was contained in non-proteid substance.
Quoted by Dietrich and Konig : Zusammensetzung und Verdaulichkeit der Futtermittel ;
2 : 987, 1380. 1891.

f The amount of nucleoproteid must be very slight.

\ Osborne : See various papers in the Journal of the American Chemical Society
since 1894.

\ Similar difficulty was experienced by Osborne, who got rid of the gum by repeated
dialysis and precipitation with ammonium sulphate. Joutnal of the American Chem-
ical Society, 17 : 429, 539. 1895.

342

KiRKwooD AND GiEs : Chemical Studies

further purified by re-solution and re-precipitation. For quantita-
tive analysis some of the final product was washed in water, alco-
hol and ether, and dried at iOO°-i05° C. to constant weight.

Sometimes the globulin prepared in this way was both crystal-
line and amorphous. At other times it was entirely crystalline.
Triangular, hexagonal and rhombohedral forms were frequently
seen, although octahedra predominated.* The crystals so closely

Fig. 4. Crystals of cocoa edestin.

resemble those we have repeatedly made from hempseed and lin-
seed by the same method, and are so like those given by Osborne
for edestin, t that we felt satisfied from the beginning our globulin
would prove to be of the edestin type. Careful study of the re-
actions of the substance convinced us of this fact, for it gives all of
those attributed to edestin by Osborne.

* The large proportion of gum extracted by the saline solution made it difficult not
only to prepare the proteid in pure form but to obtain it quantitatively. Besides, the
edestin passed in part into an insoluble modification during the manipulations. An
appreciable loss resulted, therefore, in each preparation. We obtained as much as 25
grams of the purified product from the kernels of twelve nuts.

f Osborne : Journal of the American Chemical Society. See also his paper on
crystalline vegetable proteids in the American Chemical Journal, 14 : 28. 1893.

OF THE COCOANUT DURING GERMINATION

343

On the oppo-site page we give a microphotographic view of
edestin crystals from our second preparation .Mthough not the
purest, we have selected this preparation for this purpose because
its crystals are mostly rhombohedra. These forms rarely occur in
abundance in edestin precipitates, octahedra being more commonly
obtained. Most of the larger masses among the crystals shown
here are " rounded " octahedra ; not in perfect focus because they
are thicker than the rhombohedra. The smaller particles consist
of globular matter and crystal pieces.

The crystals given in Fig. 4. were photographed for us by the
writer's colleague, Dr. Edward Leaming, who cordially gave us
the benefit of his large experience. We wish here again to extend
to Dr. Leaming our sincere thanks for his valuable assistance.

That the substance under discussion is edestin is further shown
by the results of analysis. We append our results for nitrogen
content, as determined by the Kjeldahl method, calculated for ash-
free substance :

Percentage of Nitrogen in Cocoa Edestin

Preparation .

I

2

3

4

S

Analytic results.

17.87
17.77
17-79
17.81

17-85
17.96

17.91

17.66
17.69
17.78

18.14
IS. 21
18.18

18.23
1S.20
18.28

Average.

17.91

17.71

18.18

18.24

Ash.

0.41

0.13

1.12

1.90

1.84

Preparations I, 2 and 3 contained amorphous material, possibly some of the
gummy matter referred to on page 342, in spite of our efforts to completely remove it.
Preparations 4 and 5 were obtained from I and 3 by further treatment with diastase
and by recrystallization by dialysis from lo-per-cent. salt solution. They were practi-
cally wholly crystalline.

The above results show that the globulin separated from the
cocoanut by the methods here employed is edestin.*

This same proteid of the cocoanut was examined by Ritt-
hausen, t who termed it conglutin without really identifying it with
that substance. His analyses gave it a content of nitrogen of
1 7.87-1 7.9 1 per cent. Chittenden and Setchell % referred to it by

*The edestin from barley contains 18.10 per cent. N. That from maize 18.12
per cent.; from rye, 18.19 per cent.; wheat, 18.39 P^^ '^^"'- Osborne : Journal of the
American Chemical Society, 17 : 547. 1895.

t Ritthausen : Jahresbericht uber die Fortschritte derThier-Cheraie, 10 : 18. 1880.

X Chittenden and Setchell : Quoted by Chittenden, Digestive Proteolysis, 32. 1895.

344

KiRKWOOD AND GiES : Chemical Studies

the name of phytovitellin. The composition they gave for it is in
general accord with that of edestin (nitrogen content = 18.40 per
cent.), and as they obtained it partly crystallized in octahedra, Os-
borne * has lately suggested that the substance is edestin. The
results we have obtained confirm Osborne's deduction.

The proteose to which we have already alluded was obtained
from the globulin filtrate. The latter was freed from traces of
globulin by the coagulation method, the hot filtrate evaporated to
a small bulk on the water-bath and the proteose precipitated and
purified by the usual method, t About four grams were obtain-
able from fifteen nuts. The product contained both proto and
deutero forms. Some heteroproteose was also detected in the
products formed on dialysis and a trace of dysprotose was
obtained.

The following results for nitrogen content in the ash-free sub-
stance were obtained by the Kjeldahl method :

Percentage of Nitrogen in Cocoa Proteose

Preparation.

I

2

3

General Average.

Analytic results.

18.67
18.50
18.58

18.48
18.46
18.40

18.57
18.61

18.54

Average.

18.58

18.45

18.57

18.53

Ash.

1. 71

1.08

1. 21

1-33

These results differ only slightly from those reported by Chit-
tenden and Setchell.| This difference may be explained by the
fact that mixtures of proteoses have been analyzed in each case
by Chittenden and Setchell, and by us. Their preparation of pro-
teose contained 18.25 per cent, of nitrogen.

In his volume entitled Digestive Proteolysis, Chittenden gives
the analytic results for eleven different proteids and the proteoses
derived from them (page 67). For seven of these the nitrogen of
the corresponding proteose is somewhat higher than that of the
original proteid. Analysis of our own preparations has shown the
percentage of nitrogen to be greater in the proteose than in the
globulin, a result in accord with the majority rule just noted.

* Osborne : Journal of the American Chemical Society, i8 : 13. 1896.

f Mac Dougal : Practical Text-book of Plant Physiology, 164. 1901.

J Chittenden and Setchell : Quoted by Chittenden, Digestive Proteolysis, 32. 1S95.

OF THE COCOANUT DURING GERMINATION 345

There appeared to be only a very slight amount of an albumin
in our extracts — a coagulable substance which was not precipitated
from its neutral solution when the latter was half-saturated with
ammonium sulphate.*

Osborne's methods of extracting glutenin and gliadin f in dilute
alkali and acid, and in dilute alcohol, after the removal of globulin,
proteose and albumin as above described, gave mere traces of pro-
teid substances in solution, derivatives, doubtless, of the proteids
already referred to, which perhaps had not been completely re-
moved from the residual tissue ; or possible nucleoproteid.

Peptone could not be detected in any of our extracts. |

As/i. — Composition is referred to on pages 331 and 335.
Qualitatively our results were the same as those there given.

Erizymcs. — Water, salt solution and glycerin each failed to
extract appreciable quantities of either proteolytic or adipolytic
enzymes from the endosperm of the fresh, ungerminated nut,
although an active amylolytic ferment was extracted by all of
these fluids. The large quantities of fat and fatty acid in the endo-
sperm suggest that an emulsifying ferment may be present. This,
however, may be localized in the germ, increasing to physiolog-
ical quantity and activity only in the process of germination (see
page 358). The proteoses present in the endosperm seem to
imply the presence of a proteolytic ferment. Possibly, however,
the proteoses represent a residue from which the globulin was
derived by reverse process. §

We have already referred to the fact that oxidase has been de-
tected in the milk. Traces of it are also contained in the endo-
sperm.

Average Composition. — The average results of our analyses of
the endosperm are summarized in the following table, which pre-
sents the data obtained for the composition of the fresh tissue and
the dry solid matter derived from it (constant weight at 100-
105° C).

* Cohnheim : Chemie der Eiweisskorper, 150. 1900.

f Osborne and Campbell : American Chemical Journal, 15 : 392. 1893.

X Small quantities of non-proteid nitrogenous substances were detected by Ritt-
hausen : Chemisches Centralblatt, 230. 1880. Compare, also, with recent results
respecting proteoses obtained by Bokorny : Chemisches Centralblatt, i : 1167. 1902.

I See recent papers in the Zeitschrift fiir physiologische Chetpie by Schulze and
Kutscher and their associates.

346

KiRKWooD AND GiES : Chemical Studies

Percentage Composition qf the Endosperm

Constituents.

Water.
Solids.

Inorganic matter.
Organic matter.

Fat (substance soluble in ether).
Crude fiber (cellulose).

Proteid (NX5-5)t
Soluble carbohydrate, non-nitrogenous
substance, extractive, etc. (by differ-
ence).
Nitrogen.

. Fresh Endosperm.

Dry Endosperm.*

46.31
53-69

1.03
52.66
37-29

1-93
98.07

69-45

3-39
4.08

6.31
7.60

7.90

0.742

14.71
1.382

The previous results obtained by Hammerbacher| for the
fresh endosperm from nuts of eastern origin were as follows :

Non-nitrogenous
Water. Solids. Inorganic Matter. Organic Matter. Fat. Crude Fiber. Proteid. Extractive.

46.64 53.36 0.97 52.39 35.93 2.91 5.49 8.06

Composition of Shell and Husk. — We have already alluded
to some of the uses to which the shell and husk of the cocoanut
are put by reason of the chemical and physical qualities they pos-
sess. Some facts regarding their chemical composition were given
in the table on page 335.

The following percentage results of elementary analysis of the
powdered shell were obtained by Baumhauer;§ all samples having
finally been thoroughly extracted in alcohol and ether, and then
dried at 120°- 150° C:

Cocos nucifera.

Cocos la^idea.

I. Extracted in
boiling water.

2. Extracted in
boiling water,
dilute alkali and
acetic acid.

3. Extracted in
boiling water,
concentrated
alkali and acetic
acid.

4. Extracted in
alkali and in
chlorine water.

5. Same

treatment

as I.

6. Same

treatment

as 4.

c.

H.
Ash.

52.99
5.88

1-43

47.19
6.09

46.27
5-81
I.OO

43-72

6.II

52.20
5.80
0.22

44.20
6.24

0-55

* According to Dietrich and Konig ( Konig, Menschlichen Nahrungs- und Ge-
nussmittel, etc., i : 6i2. 1893) the rt2;--(/;7V(/ substance contains the following in per-
centage of the total dry weight :

Total Substance Soluble in Water. Proteids Soluble in Water. Sugar (Sugar- Yielding Substance).
15.16 2.27 9.25

f The factor 5.5 is used because the proteids of the endosperm contain 18 per cent,
of nitrogen. See references in this connection on page 341, footnote.

J Hammerbacher : Landwirtschaftlichen Versuchs-Stationen, etc. , 18:472. 1875.
§ Baumhauer : Pharmaceutisches Centralblatt, 601. 1 844.

OF THE COCOANUT DURING GERMINATION 3^7

Nitrogen was detected in small amount in the powders which
had not been treated with alkali. The alkaline extracts contained
substance, precipitable by acetic acid, with the following percentage
composition: From Cocos nucifcra, C= 50.04, H = 5.81, ash =
4.45 ; from Cocos lapidea, C = 52.15, H *= 5.93, Ash = i.oo.

Tromp de Haas and Tollens * were able to show the presence
of a large amount of pentosane (xylan) in the endocarp, the pow-
dered material yielding an abundance of xylose on hydration in 4
per cent, sulphuric acid. Mannose was absent from the acid solu-
tion from which the xylose had been crystallized. After xylan
had been completely removed from the shell-powder by the above
method, dextrose was derived from the residue on treatment with
sulphuric acid in the usual manner.

In his very complete histological studies of the cocoanut, Win-
ton t recently called attention to the fact that both the husk and
shell contain a brown substance which is quickly changed to a red-
dish color by caustic potash, but is unaffected by alcohol, ether or
any of the specific reagents for proteids, fats or resins. He also
states that no immediate effect is produced by ferric chloride solu-
tion, but on long standing the color is changed to olive green.
Winton has pointed out the presence of minute silicious bodies
among the fibers of the husk.

Winton, Ogden and Mitchell % give the following percentage
data for the composition of the shell :

Water,
Solids,

Organic matter.
Inorganic matter,
Soluble in water,
Insoluble in HCl,
Ether extract.
Non-volatile,
Volatile,

During germination 1
soon begins to decay. See page 351.

* Tromp de Haas and Tollens: Chemisches Centralblatt, 2 : 359. 1895.

t Winton: American Journal of Science, IV. 12: 265. 1901. Facts are also
given regarding the use of powdered cocoa-shell and the husk as adulterants of ground
spices.

+ Winton, Ogden and Mitchell : Report of the Connecticut Agricultural Experi-
ment Station, 2 : 210. 1898.

7-36

Alcohol extract.

1. 12

92.64

Reducing matters calculated

as

99.46

starch,

20.88

0.54

Starch,

0.73

0.50

Crude fiber,

56.19

0.00

Nitrogen,

0.18

0.25

Albuminoid (NX 6 25),

I 13

0.25

Quercitannic acid.

1.82

0.00
shel

1 remains unaltered.

The husk

348 KiRKWooD AND GiES I Chemical Studies

CocoANUT Pearls. — Within the nut there is occasionally found
a small stony substance of a bluish white color, a kind of vege-
table bezoar, called in India calappa, which is eagerly purchased
by the Chinese, who ascribe great virtues to it as a sort of amulet
to preserve them from diseases. The cause of its formation in the
nut is unknown.

According to Harley and Harley * these pearls, like those of
molluscan origin, appear to consist almost entirely of calcium
carbonate, with water and organic matter in smaller proportion.
Riedel, quoted by Harley and Harley, states that in 1886, while
in North Celebes, he found a pearl in the endosperm of the cocoa-
nut. One such a pearl was pear-shaped in form and 28 mm. long.

We are greatly indebted to Dr. D. Morris, Imperial Commis-
sioner of Agriculture for the West Indies, for the following very
interesting quotation from a letter to Dr. MacDougal :

" More than two hundred years ago Rumph, an eminent bot-
anist in the East, sent as a present to the Grand Duke of Tuscany
a ring in which a cocoanut pearl had been set. Further, Rumph
himself described cocoanut pearls in his great work with consid-
erable minuteness and gave illustrations of two of them. One
was perfectly round, the other was oval or egg-shaped. * * *
Travelers in the Philippine Islands have heard of cocoanut pearls,
but seldom or ever have seen them. The natives, it is said, keep
"cocoanut stones" as charms against disease and evil spirits.
The rajahs, we were told, highly prized them and wore them as
precious stones. It was only a few years ago that real cocoanut
pearls were at last brought to England. One is now at the Mu-
seum at the Royal Gardens at Kew, brought by Dr. Hickson.
It is almost egg-shaped, perfectly white, and composed almost en-
tirely of carbonate of lime. It has, in fact, a somewhat similar
composition to the pearl of the oyster, and yet there is little doubt
it is a purely vegetable product." f

* Harley and Harley : Proceedings of the Royal Society of London, 43: 464.
1887-88.

t '• Besides these cocoanut pearls," quoting further from Dr.. Morris' letter to Dr.
MacDougal, '• Rumph describes what he calls ' Melate ' pearls taken from the flowers
of a Jasmine ; and a ' Champake ' pearl taken from the flower of a Michelia. If we
had not already seen the pearl of the cocoanut it would have been impossible to believe
that there were such things as Jasmine and Michelia pearls * * * Of their composition,
mode of occurrence and true nature we have yet to learn." See the article by Harley
and Harlev referred to above.

OF THE COCOANUT DURING GERMINATION 349

See also, on the subject of cocoanut pearls, the Proceedings of
the Boston Society of Natural History, 1861 and 1862 ; The
Tropical Agriculturalist, 1887; Nature, 1887.

III. Changes in the Cocoanut during Germination

The nuts for our studies of the changes occurring during
germination were obtained fresh, fully developed and with their
husks still on them, directly from Jamaica. Immediately after their
arrival at the New York Botanical Garden they were imbedded in
earth until they were nearly covered. The earth was kept saturated
with water and a tropical temperature was maintained. These con-
ditions closely approximated those attending normal germination.

Morphological Changes. — Nearly four months elapsed before
the shoots began to appear through the husks, the fibers of the
husk having been pressed aside in their upward progress. At
this stage the stem of the shoot was an inch or more in diameter
at the " root-crown," sharply tapering toward the point of pene-
tration at the surface of the husk. As the growth proceeded it
seemed to gradually become more and more rapid, and by the end
of a year the plants had attained the height of two or three feet,
with a stem about an inch in diameter throughout most of its length.
By this time the part of the husk under the earth had decayed
considerally; it became softer and more porous, and several stout
roots had developed through it and penetrated the soil to the
depth of a foot or more.*

The appearance of the nuts and their plants at this period of
their growth is shown in the cut on page 350. Unless other-
wise stated, the chemical analyses reported farther on were made
of the parts at this stage of their development.

It may not be amiss, in describing the morphological changes
induced in the nut during the process of germination, to also
briefly review, at the same time, the more important facts regard-
ing structure of the nut as it exists in the ungerminated condition.

The entire fruit is, strictly speaking, intermediate between a
nut and a drupe — a "drupaceous nut." The outer covering,

* For facts regarding germination and cultivation see I'.ailey : Cyclopedia of
American Horticulture, I : 341-343. 1900. Also, Wittmack, I.. : Die Keimung der
Cocosnuss. Ber. d. deut. bot. Ges. 14 : I45. l8j6.

350

KiRKWOOD AND GiEs : Chemical Studies

F'iG. 5. Germinating cocoanut at the end of a year, showing plumule and roots,
with husk little altered except where it was in contact with the earth.

usually removed before the nut appears upon the market, is a
thick fibrous layer comprising the exocarp, the epicarp consist-
ing of a smooth, thin, tough coat of a brownish or grayish color.
(See pp. 323 and 324). The endocarp, or what is commonly known
as the shell of the nut, is composed of three carpels whose lines
of fusion are always apparent. The nut lies in the husk with the
end containing the " eyes " toward the pedicel. Each carpel con-
tains an " eye," so-called, and under one of these three eyes, the

OF THE COCOANUT DURING GERMINATION 351

softest, is the germ imbedded in the endosperm. The fertile carpel
may be recognized from the fact that it has the greatest degree of
divergence between the longitudinal fusion lines of the carpels.
The true integuments of the ovule are reduced to a thin brown
coat closely adhering to the abundant endosperm.

The embryo is a cylindrical body about 8 mm. in length lying
below one of the natural openings of the endocarp and in a line
perpendicular to the exterior surface of the endosperm. When
germination begins the embryo elongates and, having pushed
through its thin coverings, begins to enlarge at both ends.
From the outer end arises the plumule and the roots ; the inner
end is an extension of the true cotyledon and is developed into a
special absorbing organ. See //. ig.

The absorbing organ is of a soft spongy texture and all
through it are the ramifications of vascular strands which converge
to the narrow " neck, " which connects the absorbing tissue with
the stem. The cotyledon, and by this term, hereafter, we shall
mean the part of that structure specialized for absorption, can
attack only the part of the endosperm to which it is contiguous.
In the earlier stages of growth this absorption is confined to the
part nearest the young shoot, which we may hereafter refer to as
the proximal end of the nut. Finally, however, the cotyledon
fills the entire cavity of the nut and somewhat thins the endosperm
distally, also.

The milk may persist in the nut until the cotyledon has almost
filled the cavity. After germination has proceeded for some time
the milk becomes insipid to the taste, and contains fragments of
cellulose and large drops of floating fat. In nuts in which germi-
nation has continued for a year the cotyledon has entirel}'' filled
the cavity, but usually there is still left a third to a half of the endo-
sperm undigested. This residual portion in normal cases is little
affected, except that it is softened superficially, and to the taste
suggests nothing different from the meat of the ordinary ungermi-
nated nur.

In its natural development the roots of the plant soon take
firm hold of the soil and, long before the endosperm is com-
pletely absorbed, junction between the shoot and the absorbing
organ is broken, the husk decays and the plant enters an inde-

352 KiRKWOOD AND GiES : Chemical Studies

pendent career. Neither the husk nor the shell appears to serve
any other than passive mechanical function, and only the constitu-
ents of the endosperm and milk, so far as the nut is concerned,
nourish the young plant before it finds in the soil the elements
provided there in abundance for its growth to maturity.*

In our microscopic studies, particularly of the cotyledon,
pieces of the fresh part were "fixed" in a mixture of glacial
acetic acid (^) and 70/0 alcohol (2^). After remaining in this
fluid for a few hours the pieces were transferrecJ to 70^ alcohol
and later to 85/0 alcohol, in which they were kept. Sections were
cut with a razor and mounted in glycerin. Treated with iodine,
such sections of the cotyledon showed an abundance of starch in
all cells except those of the outermost layer or epidermis. This
outer layer stained yellow with iodine. That it contained an abund-
ance of oil was shown by its deep black reaction with osmic acid.
Large globules of oil are, however, distinctly visible in the epidermis
under the microscope without the aid of osmic acid. Oil may
also be found in the subepidermal layers, but it rapidly diminishes
in quantity as the distance from the outside increases and as the
starchy deposit accumulates. Needle-like crystals may be very
readily found in the epidermal cells. These resemble crystals of
tripalmitin, but the fact that so much oil appears in globules, and
that the breaking down of fats must occur to a large extent in
these cells, would suggest that they are palmitic acid rather than
the fat itself.

The above facts make it appear that the starch is formed indi-
rectly at least from the oil.f See references under enzymes,
pages 345 and 358.

See//, /p for drawings of parts mentioned above.

Chemical Changes. — The following summary gives briefly
the effects of germination on the individual nuts examined :

I. Not Analyzed. — A. Development had proceeded for nearly
six months. The plumule protruded six inches above the husk.
Roots had developed through the husk — two were about a quarter
of an inch in diameter. The stem was very thick at the " root-

* Note references on page 357 to the functions of the husk in holding water and
possibly furnishing nutrient matter in its decay.

f Ebermayer : Physiologische Chemie der Pflanzen, 347. 1882.

OF TF1E COCOANUT DURING GERMINATION 353

crown " ; sharply tapered to the point of surface penetration. The
fibers about the stem were tightly pressed together. The neck of
the absorbing organ was very woody and fibrous in character.
Absorption of the endosperm at the proximal end was quite
marked ; distal portion undiminished. Milk cavity largely filled by
the absorbing organ. A small space at the distal end remained,
containing viscid white material full of large oil globules ; quantity
less than lo c.c. — doubtless concentrated milk. It was strongly
acid in reaction from acid phosphate, reduced Fehling's solution,
gave only a faint biuret reaction and was free from starch. The
inner surface of the endosperm in the distal portion was soft, having
the consistency of lard.

The weight of the whole shoot, minus the roots, in the fresh
condition was 28.1 grams. Dry, the weight was 4.25 grams or
1 5. 1 per cent, of the fresh substance, indicating a presence of 84.9
per cent, of water in the original plant.

B. This nut, although germinating for the same period of time,
was not quite as far advanced as the previous one, having shoots
that were just emerging from the husk. The stem was thicker,
however. In most respects its internal condition was exactly the
same as that of the first. The fluid in the distal cavity was less in
quantity, not as turbid, contained less oil — otherwise was the same
as that of the previous nut.

The weight of the entire plumule was 38.8 grams. Dried,
it weighed 5.75 grams. Thus it contained 14.8 per cent, of solid
matter and 85.2 per cent, of water.

II. Afialyzed. — i. Growth continued for eight months. The
cotyledon entirely filled the cavity. About half of the endosperm
was absorbed ; practically all of that proximally except a thin
layer. The distal residue of endosperm appeared to be normal in
taste and appearance except on the surface, where the soft layer
previously referred to — one fourth the entire thickness — could
again be seen. The outer surface of the cotyledon, that part in
contact with the endosperm, was much corrugated ; the whole
organ, solid but spongy, sweet and agreeable to the taste, pyriform.
In the tables on pages 354 and 355 the results of our analyses for
this nut are indicated by the numeral i.

354

KiRKWooD AND GiES : Chemical Studies

General Compo.sition of the ]

['arts of the

GERMINATED COCOANUT

Percentage of fresh Tissue.

Percentage of
Solids.

Parts of the Nut and its Plant.

Water.

Solid Matter.

Or-
ganic

Matter.

In-

Total.

10.90

12.29

8.38

8-59
II. 01

13-93

Or-
ganic.

99.14
99.10
99.00
99.11
99.20

98-94

gg.o8

In-
organic.

0.86
0.90
1. 00
0.89
0.80

1.06
o.g2

I-31

1.07

1.74
1.42

1.38

0.87
0.84

0.85

organic
Matter.

I. Cotyledon.

A. Central, vascular portion : rt— I
" " " b — I

<c it " c 2

« «« '< d—2

(< K " c 3

Central, between center
and surface : /— 3

89.10

87.71
91.62
91.41
88.99

86.07

92.07
92.25
88.10
89.71
92.77

92.36

gi.2i

91.30

93-85
90.93
93-24

7-93

7-75
11.90
10.29

7.23

7.64

Average.

8g.i5

10.85

8.79

B. Outer, corrugated portion : a — i
K •' '« b — 2

<( <( •' c 3

Outer portion — neck : d — 3

84-95
82.79
80.83
78.98

8i.8g

19.09

23.42

21.25

31-65
30-36
28.68

25-77

1505
17.21

19.17
21.02

18. II

80.91
76.58

78.75

68.35
69.64

71.32
74-23

98.69

98-93
98.26

98.58
g8.62

99-13
99.16

99-15

8.70

6.15
9.07
6.76

Average.

92-33

98.92
98.90

g8.9i

7-67

II. Residual endosperm.

C. Proximal portions after

much absorption : a — 2

^—3

1.08

X.IO

Average.

i.og

D. Medial portions : a — I

/; — I
c — 2

'^-3

99-03
99.06
99.25
99-23

0.97
0.94
0.75
0.77

0.86

98-58
98.65

98-95
98.96

1.42

1-35
1.05

1.04

Average.

2g.i2
46.08

70.88

99-14

98.78

1.22

E. Distal, normal portion : a — 3

53-92

99.02

0.98

g8.i2

1.82

III. Stetn of the plant.

F. Base, " root crown," with

petioles at lowest parts : a — I
^— I
C — 2

86.21
86.51

85-15
84-31

13-79
13-49
14-85
15-69

98-95
98.70

98.84

98.68

1.05
1.30
I.16
1.32

92.37
90.38
92.06
91.60

7-63
9.62

7-94
8.40

Average.

85-55

14-45

98.79
98.80

1. 21

gi.6o

93-20

8.40

G. Parts above the base, with

more petioles : a — 2

b-i

82.47
79.87

17-53
20. 13

1.20

6.08

Averse.

81.17

18.83

98.57
98.63

98.75
98.63

98-65

IV. Petioles.

H. Alone, or with young

leaves : a — I
b—\

^—3

83-63
8255
82.13
82.17

16.37

17-45

17.87

17.83
17-38

1-43
1-37
1-25
1-37

1-35

91.27
92.15
9301
92.34

g2.ig

8.73

7.85

6-99
7.66

Average.

82.62

7.81

OF THE COCOANUT DURlNf. GERMINATION

355

General Composition of the Parts ok the germinated CocoANur. — Continued

Percentage of fresh Tissue. Percentage of

Solids.

Parts of the Nut and its Tiant.

Water.

Solid Matter.

Or-

Inrir-

Total.

25.34
28.01
27.40
27.49
31-55
29-35

28.19
12.78

Or-
ganic.

98-35
98.10

98.34
98.41

97-96

98-39

98.26
98.50

98.46

98-67

98-43
98.83

98.60

98.80
98.85
98-50
98.77

In-
organic.

I-6S
1.90
1.66

1-59
2.04
1. 61

1.74

e.-inic ganic
Matter. Matter.

1

\'. Leaves.

I. Mature or nearly so : a — I

b—l
C — 2
d—Z

/-3

74-66

71.99
72.60

72-51
68.45
70.65

71.81

87.22

93-49
93.20

93-93
94.19

93-52

94-52

6.51
6.80
6.07
5.81

6. 48
548

Average.

93.81

6.19

J. Very youngest : a — 3

1.50

1-54
1-33
1-57
1. 17

1.40

86.67

13.33

VI. Roofs.

K. Short, not developed outside

of husk, with soft tips : a — I

b—2

c — 2

87.08
89,89
86.41
87-46

87.71

12.92
10. 1 1

13-59
12.54

12.29

22.08

17.35
18.91

20.53

88.09
86.89
88.44
90.70

II. 91

13.11

11.56

930

Average.

88.53

11.47

Z. Parts of longer roots, taken

near the stem : a—\

b 2

d-Z

77.92
82.65
81.09
79-47

80.28

1.20

1-15
1-50
1.23

94-59
93-34
92.09
94.00

5-41
6.66

7.91

6. CO

Average.

19.72

98.73

1.27

93.51

6.49

M. Parts of longer roots, taken

outside the husk : a — i

b I

C 2

81.70
84.64
82.79

83.04

18.30
15-36
17.21

6.96

97-94
98.47
98.40

2.06

1-53
1.60

88.76
90.05
90.74

11.24

9-95
9.26

Average.

98.27

1-73

89.85

10.15

2. This nut represented germination after ten months and was
essentially like the preceding in all respects. The roots were
thicker and a number of good sized ones had not yet proceeded
through the husk. Their ends were soft and watery, rounded and
blunt. Analyses of this nut are referred to in the tables on pages
354-355 by the numeral 2.

3. Essentially the same as No. 2, both in stage of develop-
ment and conditions of parts, although the time of germination
was about two months longer. Analyses of the parts of this nut
are referred to in the tables on pages 354-355 by the numeral 3.

This nut had germinated for just about a year. The follow-

356 KiRKWOOD AND GiES : Chemical Studies

ing weights of the fresh parts were very carefully taken ; the data
for nitrogen were determined by the Kjeldahl method.*

Percentage of Percentage

Total Weight. of Nitrogen.

Weight

in Grams.

Cotyledon,

196

Central part,

Cortical layer,

Endosperm,

155

Distal portion,

Proximal portion,

Shell,

161

Stem,

16

Lower part,

" Root crown,"

Roots,

41

Inside of husk,

Outside of husk,

Petioles and young leaves,

58

Petioles,

Old Paves and petioles,

71

Total weight,

698

Shell and contents.

512

Whole plant.

186

Co'yledon and endosperm.

351

28.1

22.2

230
2-3

5-9

0.14
0.31

0.65
0.93

0.70

0-53

0.27

0.54

8.3 0.29

0.39

10.2 0.45

73-4
26.6

50.3

The tables on pages 354-355 give all our results for general
composition of the parts of the germinated nut.f Numerous
deductions may be drawn from these rc-^ults, as to growth and
metabolism.

The central part of the cotyledon, with its vascular network,
contains more water and less solid matter than any other part of
the germinated nut. The proportion of solid substance in it in-
creases toward the corrugated epithelium, being greatest in the
"neck," where the structure is fibrous and woody.

That the absorbing organ completely takes up the milk is very
evident from the way it fills the cavity and from its own composition,
but it is likewise apparent from our results that water is also with-

* The husk was not weighed because it was decayed underneath and water-logged.
Nitrogen was not determined in the shell because its substance remains unaltered dur-
ing germination.

t The methods of determination were the same as those used previously. The
roots, outside of the husk, which had been in the wet soil, were hurriedly rinsed with
water to remove inorganic matter then wiped dry with a towel and at once cut into
thin cross sections for analysis.

OF THE COCOAXUT DURING GERMINATION 357

drawn from the residual endosperm, this absorption beinj^ greatest
at the proximal end of the nut, where absorption was begun in the
first place, and least at the distal end, where it had hardly com-
menced. The roots also are seen to have absorbed considerable
moisture.

The lowest part of the stem contains almost as little solid
matter and is nearly as watery as the cotyledon. The percentage
of water in the stem diminishes as the distance away from the " root
crown " increases. The watery condition of the lower part of the
stem is increased, doubtless, by the fact that the surrounding husk
is impregnated with water, thus favoring direct absorption by
osmosis and at the same time preventing evaporation from the sur-
face of the growing tissue.

The amount of solid matter in the petioles is also compara-
tively slight, little more than in the lower part of the stem. In
the leaves the water is greatest in the youngest, as would be ex-
pected ; least in the oldest — those most exposed to the air.

The roots at the tips are soft and watery, but the older they
become the more solid matter they develop and the more woody
material they accumulate.

In the distribution of the inorganic matter in the fresh parts it is
noticeable that the portion of saline substance increases with a
decrease of water and vice versa, as in the cotyledon, in the residual
endosperm and throughout the plumule. This condition is such as
might be expected. The relation of the inorganic to the organic
matter in each part, however, is variable. The substance of the
cotyledon and the stem contains a greater portion of salts than
that of the endosperm and the leaves, the roots likewise holding
a fairly large amount of saline matter. The substance of the endo-
sperm contains least of all, from which fact it is quite clear that
the inorganic matter of the plumule has been absorbed, not only
by the cotyledon from the milk, but also by the roots from the
fluid in the husk and the surrounding earth.

At the beginning of germination the inorganic matter and
water of the milk are doubtless sufficient for the changes that
occur, the organic matter coming chiefly from the endosperm.
Some time before the cotyledon fills the milk cavity and completely
absorbs the milk, the roots have begun to take water and inorganic

358 KiRKWooD AND GiES : Chemical Studies

matter from the fluid in the husk — possibly also organic substance
from the disintegrating husk fibers — and thus they absorb new
nourishment from a large supply. Growth of the plumule is con-
sequently favored. The plumule soon reaches such a height and
development as to enable it to make increasing contributions to
the plant metabolism from the gaseous products the air affords.
By this time the whole growth has become practically independent
of the reserve material of the seed.

Enzymes. — We made only a few preliminary studies of enzyme
distribution. Extracts were made in water, dilute salt solution
and glycerin. The indicators used in nearly all the experiments
were prepared from the materials in the nut itself.

The extracts of the cotyledon were acid to litmus (phosphates),
though, as indicated by lacmoid, they contained no free acid.
Diastatic ferment was found to be distributed in abundance in all
parts of the cotyledon. Oxidase was also present. Only the very
slightest proteolytic action was manifested by the cotyledon ex-
tracts, even when they were obtained in particularly concentrated
form. In some experiments the results were entirely negative, how-
ever. Cellulose-dissolving and fat-splitting enzymes were not
detected in either the cotyledon or the residual endosperm, al-
though we cannot be sure that in our few experiments they have
not escaped us.* Germination progresses so slowly that possibly
some of the enzymes are present in only very minute quantity at
any one time — in such amount, perhaps, as to be undiscoverable
by the methods commonly employed for ferment detection. We
did not examine the parts of the plumule in this connection.

At this point, before we were able to come to any very definite
conclusions as to the enzymes present and before we could de-
termine the distribution of proteids, fats, carbohydrates, etc., in the
parts of the plant, we were obliged to discontinue our work. The
writer hopes to extend these experiments on the germinated cocoa-
nut to a consideration of related problems of nutrition.

* See our references to enzymes on page 345. Lipase seems to have been found
in the germinating cocoanut by Lumia : Jahresbericht iiber die Fortschritte der Thier-
Chemie, 28 : 724. 1898.

OF THE COCOANUT DURING GERMINATION oOl*

l<;xplaiiatioii or I'late 19

Fig. I. An end view of the cocoanut, without its husk, showing tlic three ciir-
pels and the " eyes." The fertile carpel is the one included in the largest angle.

Fig. 2. A sectional view through the end of an ungerininated nut, with its husk
removed, showing the form and location of the germ imi)e(lde<l in the endosp< rni
(under the micropyle). The shell is indicated by the heavy outline.

Fig. 3. Another sectional view similar to that of Fig. 2 showing development of
the absorbing organ after germination had proceeded for a few weeks. The incipient
stem and roots are to be seen. The cotyledon has enlarged within and without the
shell.

Fk;. 4. A longitudinal section through the whole nut somewhat to the side of the
median line. It shows the nut imbedded in its fibrous husk and the conditions found
after germination had gone on for about five months. The absorbing organ has filled
about two-thirds of the cavity of the nut but has absorbed comparatively little of the
endosperm. The stem has proceeded upward through the husk, the roots downward
through the husk into the soil. The leaves have been cut oft" above the husk. (See
page 350 for appearance of full plumule. )

Fig. 5. A section through the layer of the cotyledon normally in contact with
endosperm, showing the corrugated, villiform arrangement of the absorbing epithelium.
This section having been treated with osmic acid shows the localization of the fat
globules.

Fig. 6. A section through that part of the cotyledon given in Fig. 5. This sec-
tion, treated with iodine, shows the localization of starch.

Fic;. 7. A few cells from the absorbing epithelium of the cotyledon after enlarge-
ment. This figure shows the large clear oil globules and the darkly staining starch
grains in the subepidermal cells. The crystals seen in most of the cells appear to con-
sist of fatty acid, possibly palmitic.

Pig. 8. Enlarged starch granules from subepidermal cells.

Bui-L. ToRREY Club, 29

Pi.. 19.

COCOS NUCIFERA.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 27

SOME MT. DESERT FUNGI

BY V. S. WHITE.

NEW YORK
1902

[Reprinted from Bulletin Torhky Botanical Club, 29: fi5(W6.}, 25 September, 1902.)

Some Mt. Desert Fungi

15y v. S. WllITK

The country in the neighborhood of Bar Harbor forms a
specially interesting collecting ground for the mycologist because
of the quantity of wooded land which remains in spite of the
constantly increasing clearings and improvements. The woods
are moist and kept in this condition even in very dry weather by
numerous springs, underground streams and rivulets.

A system of paths originally begun by some of the first Boston
visitors, modeled on that in u.se about the forests of Thiiringen
near Eisenach is maintained by the Village Improvement Society.
The paths extend many miles in various directions ; pocket maps
are to be had on which the paths are marked in different colors
to correspond with the colors painted upon the trees at short
intervals or on occasional signposts ; rude arrows formed by forked
white birch branches indicate the entrace to these paths and are
also placed at points of intersection, so that there can be no
danger of going astray or unnecessarily out of one's way except
from heedlessness. These paths though kept free from under-
brush and occasionally cleared of dead wood, do not interfere
with the natural beauty and wildness of the place and render the
task of the collector a far easier and pleasanter one than it would
otherwise be. The present list of fungi collected in the vicinity
of Bar Harbor, Mount Desert, Maine, from 8 July to 15 Sep-
tember 1 90 1, by the writer have been determined principally by
Professor Charles H. Peck. Professor F. S. Earle of the New
York Botanical Garden, Professor Underwood of Columbia Uni-
versity, Dr. W. A. Murrill of New York, arid Mr. H. J. Banker of
California, Pa., have also aided the writer in making determina-
tions. Dried specimens with notes and in some cases water-color
sketches have been deposited with the New York Botanical
Garden. It is hoped that in the future this list may be greatly
enlarged as the fungus flora becomes better known. The present
list may serve as a preliminary supplement to Rand's Flora of
Mount Desert Island from which the fungi were omitted. The

550

551 White : Mt. Desert Fungi

one hundred and twenty species marked with an asterisk (*) are
not found in Ricker's List of Maine Fungi (April 1902).

Preliminary List of Bar Harbor Fungi

MYXOMYCETES
Arcyria nutans (Bull.) Grev.
Arcyria incarnata Pers.
DicTYDiuM cancellatum (Batsch) Macbr.
FuLiGO SEPTiCA (Link) Gmel. Common.
Leocarpus fragilis (Dicks.) Rost.
Lycogala epidendron (Buxb.) Fr.

ASCOMYCETES
Family Hypocreaceae
Cordyceps entomorrhiza (Dicks.) Fr. Only one specimen
found.

Hypocrea rufa (Pers.) Fr. One form.

Hypomyces lactifluorum (Schw.) Tul. On Lactaviiis piper-
atiis. Plentiful.

Nectria episphaeria (Tode) Fr.

Family Chaetomiaceae
*Chaetomium olivaceum C. & E.

Family Diatrypaceae
Diatrypa stigma (Hofifm.) De Not.

Family Xylariaceae

Hypoxylon cohaerens (Pers.) Fr. Common.
Hypoxylon coccineum Bull. Very plentiful.
Xylaria polymorpha (Pers.) Fr.

Family Mollisiaceae

* MOLLISIA MELALEUCA Fr.

Family Helotiaceae
Helotium citrinum (Hedw.) Fr.

Family Pezizaceae

Lachnea hemisphaerica Wigg.
Lachnea scutellata (L.) Sacc.

White : Mr. Dksekt Fungi 552

Dasyscvfha Agassizii (B. & C.) Sacc. Very plentiful in cer-
tain localities.

Peziza badia Pers. Widely distributed.

Family Geoglossaceae

Geoglossum hirsutum Pers. P^ound only once but growing
in large quantities.

Leptoglossum luteum (Peck) Sacc. Common on moss-
grown decayed stumps.

Leotia LUBRiCA (Scop.) Pcrs. Very plentiful in damp woods.

BASIDIOMYCETES
Family Pucciniaceae
Gvmnosporangium clavipes C. & P. On Ju7iiperns Virginiana.

P^amily Auriculariaceae

Auricul.aria Auricula-Judae (L.) Schroet. Growing in
large quantities in certain localities.

Family Tremellaceae
Tremellodon gelatinosum (Scop.) Pers. Plentiful.

Family Thelephoraceae

Corticium effuscatum C. & E.
CoRTiciUM INCARNATUM (Pcrs.) P"r. Plentiful.
Cr.\terellus cornucopioides (L.) Fr. P^ound only once but
in large quantity on hard-packed woody path.

*Craterellus lutescens (Pers.) Fr. Common.

Stereum hirsutum (Willd.) Pr. Plentiful.

Stereum ochraceoflavum Schwein.

Thelephora Schweinitzii Peck. Growing in large quantities.

Family Clavariaceae.

Clavaria cristata Pers. Common.
*Clavaria circinans Peck. Common.
Clavaria fusiform is Sowerby. Fairly abundant.
Clavaria inaequalis Miill.
Clavaria pistillaris L. Few specimens found.

553 White : Mt. Desert Fungi

Family Hydnaceae.

* Hydnum albonigrum Peck. Found only in one locality.

* Hydnum boreale Banker, sp. nov. Plant white to greenish,
moist; pileus thick, convex, undulate, 12-15 cm. wide; stem
short, thick, flattened, the longer diameter three times the shortest
(2 cm. by 7 cm.), 7 cm. long; teeth short, 5 mm. long, 0.5 mm.
wide, the points rounded and somewhat knobbed when dry, decur-
rent, brownish with white tips ; substance zonate within, the upper-
most layer white spongy thick, the lower strata becoming darker,
harder, and more compact, in the stem deep lavender color ; my-
celium purple, persistent ; odor unpleasant.

In damp rich soil and leaf mould.

Hydnum carbunculus Seer. [= H. fcrrugineuin Fr.) The
peppery taste of this species has not been previously noted and is
an important character.

* Hydnum humidum Banker, sp. nov. Plant yellowish, be-
coming brown, hygrophanous, pileus irregular, spongy, scrobicu-
late, 9 cm. or more wide, stem stout, brownish, about 2 cm. long,
1.5 cm. thick ; teeth long, 5 mm., brownish with white tips ; sub-
stance corky, hard when dry and gray like weathered wood but
with streaks and spots that appear darker and cartilaginous, which
cut as if gummy, zonate internally ; odor strong but not unpleas-
ant ; taste mild.

In hard-packed wooded path. The fresh plant exudes water
freely when pressed between the fingers.
Hydnum repandum L. Fairly common.

* Hydnum scrobiculatum Fr.
Hydnum zonatum Batch.

Family Polyporaceae.
Daedalea confragosa polyporoidea Peck. Plentiful.
Favolus alveolarius (DC.) Quel. Fairly abundant.
PoLYPORUS adustus (Willd.) Fr.

* PoLYPORUS CARPI neus Sower,

POLYPORUS CARNEUS NcCS.
PoLYPORUS CHIONEUS Fr.
PoLYPORUS CONNATUS Wcin.

* POLYPORUS coNFLUENS (A. & S.) Fr. Plentiful.
PoLYPORUS ELEGANS (Bull.) Fr. Fairly abundant.
PoLYPORUS fomentarius (L.) Fr.

White : Mt. Desert Fungi 554

POLYPORUS LEUCOPHAEUS Moilt.
POLYPORUS PERENNIS (L.) Fr.
POLYPORUS PICIPES Fr.

PoLYPORUS PERGAMENUS Fr. Very common.
PoLYPORUS ScHWEiNiTZii Fr. Plentiful.

POLYPORUS SCHWEINITZII DUALIS Peck.

PoLYPORUS SPLENDENS Peck. [P. oblectaus of Ricker's list, but
that species is not likely American.)

PoLYPORUS SULFUREUS (Bull.) Fr.

PoLYPORUS VERSICOLOR (L.) Fr. Very common in all localities.

PoLYPORUS VAPORARIUS Fr.

Lenzites sepiaria Fr.

Trametes ciNNABARiNA (Jacq.) Fr.

* Trametes rubescens Fr.

Family Boletaceae

Boletixus pictus Peck. Common in damp places, Sphag-
num, etc.

* Boletus albus Peck. Only one plant seen.
Boletus Americanus Peck. Common in damp places.
Boletus chromapes Frost. Fairly common but scattered,

rarely more than one plant in a locality.

* Boletus cyanescens Bull. Common.

* Boletus edulis Bull. Plentiful.

* Boletus edulis clavipes Peck. Common.

* Boletus felleus Bull. Rather scarce.
Boletus granulatus albipes Peck. Plentiful.

* Boletus miniato-olivaceus sensibilis Peck. Common.

* Boletus ornatipes Peck. Very common in woods.

* Boletus piperatus Bull. Common in woods and by road-
sides.

* Boletus purpureus fumosus Peck, var. nov. Pileus convex,
velvety, smoky brown, flesh creamy-yellow, changing- to bluish-
green where wounded ; tubes depressed about the stem, yellowish-
green, the mouths small, brownish-red, the dissepiments slightly
Recurrent on the stem, changing to greenish-blue where wounded ;
stem tapering upward, yellow above, colored like the pileus below,
bright yellow within, changing to bluish-green where wounded ;
spores lo ft long, 5// broad.

555 White : Mt. Desert Fungi

Pileus 4-14 cm. broad; stem 4-12 cm. long, 1-2.5 cm. thick
at the top, 2.5-5 cm. thick at the base.

Woods. July.

This Boletus agrees so closely with B. piirpiircns in all its char-
acters except color that it has seemed best to consider it a mere
variety of that species, but future investigation may show it to be
a distinct species. It does not appear to have any purple hues in
either pileus or stem.

* Boletus Russellii Frost. Under pines in dry soil ;
scarce.

Boletus scaber Frost. Very common and variable.

* Boletus scaber mutabilis Peck. Common.

* Boletus scaber niveus Peck. Common.

* Boletus scabripes Peck, sp. nov. Pileus thick, broadly con-
vex, firm, dry, glabrous, reddish-brown with a bloom ; tubes con-
vex in the mass, adnate or slightly depressed around the stem,
brown, with very small round darker mouths ; stem stout, equal,
solid but spongy within, grayish-white, adorned with numerous
small projecting black points; spores oblong, fusiform, 12-15 ;«
long, 4-5 a broad.

Pileus about 15 cm. broad; stem about 10 cm. long, 2-3 cm.
thick.

Rich soil in woods. July. A fine large species but with a
disagreeable taste. The black points on the stem are suggestive
of those on the stem of B. scaber, but they are smaller and more
numerous. In drying the plants become black and exude a black
juice with a strong odor.

* Boletus subglabripes Peck. In damp shady soil ; only
two plants found.

Boletus subluteus Peck. Very plentiful all summer.

Family Agaricaceae

( Ca7itharcHcae)

Caxtharellus aurantiacus (Wulf ) Fr. Plentiful.
Caxtharellus cibarius Fries. Common and widely dis-
tributed.

* Caxtharellus flaviceps Peck.

Caxtharellus floccosus Schwein. Very plentiful in a few
localities.

White : Mr. Desert Funcji 556

Cantharellus multiplex Undciw. Two specimens. Speci-
mens attaining a large size, some measuring nearly 17 cm.
across.

Trogla CRISPA (Pers.) Vr. Plentiful in a few localities.

* Paxillus atrotomentosus Fr. Found plentifully in one
locality ; a very showy fungus.

* Paxillus involutus (Batsch.) Fr. Plentiful.

* Phelloporus rhodoxanthus (Schwein.j Peck. Only two
plants found ; in widely different localities.

( Copriucac)

* Coprinus atramentarius (Bull.) Fr. Plentiful.
Coprinus atramentarius siLVESTKis Pcck. Found only in

one locality but abundant there.

* CoPRixus PLUMBEUS Pcck. A large quantity in damp

weather.

( Hygropliorcae)

Hygrophorus Cantharellus Schwein. Fairly common.

* Hygrophorus chlorophanus Fr. Abundant in restricted
localities.

* Hygrophorus coccineus Schaeff. In large quantities in
damp weather.

Hygrophorus conicus (Scop.) Fr. Abundant.
Hygrophorus miniatus Fr. F'ound once or twice only.

* Nyctalis asterophora Fr. Growing on much-decayed
specimens of Rnssiila and Lactarius pipcratus. This is a very
strange species, in which the lamellae are almost entirely unde-
veloped and the entire top of the fungus bears the spores.

{Lactarieae)
Lactarius affinis Peck. Pale form ; plentiful.
Lactarius cinereus Peck. Only two or three plants found.
Lactarius deceptivus Peck. Plentiful.

Lactarius deliciosus (L.) Fr. Fairly common in certain
localities.

* L.\ctarius fuliginosus Fr. Plentiful.
L.\CTARius GRiSEUS Peck. Plentiful in damp woods.

557 White : Mr. Desert Fungi

Lactarius hysginus Fr. Plentiful in damp woods.
Lactarius insulsus Fr. Only one plant seen.

* Lactarius maculatus Peck. Only two specimens found ; a
showy plant.

* Lactarius piperatus (Scop.) Fr. Common.

* Lactarius sordidus Peck. Found only once but then in
large quantities.

Lactarius subdulcis (Bull.) Fr. Plentiful and widely distrib-
uted.

* Lactarius torminosus (Schaeff.) Fr. Found twice.
Lactarius theiogalus (Bull.) Fr. Found in only one local-
ity growing in leaf mould.

Lactarius uvidus Fr. Only a few plants found, some grow-
ing on wood.

Lactarius volemus Fr. Common.

* RussuLA ADUSTA (Pers.) Fr. Common.

* RussuLA brevipes Peck. Plentiful.

* RussuLA BASiFURCATA Peck. Found several times.
RussuLA decolorans Fr. Plentiful and widely distributed.

* RussuLA EMETicA Fr. Common.

* RussuLA FiNGiBiLis Britz. Only found three times.
RussuLA FOETENS (Pcrs.) Fr. Common.

* RussuLA FRAGiLis (Pers.) Fr. Only a few plants found.
RussuLA INTEGRA (L.) Fr. Plentiful.

* RussuLA LUTEA (Huds.) Fr. Found only one.

* RussuLA MARIAE Peck. Fairly widely distributed.

* RussuLA NIGRICANS (Bull.) Fr. Found abundantly in a few
localities.

* RussuLA PULCHRALis Britz. Plentiful in one locality.

* RussuLA PALUSTRis Pcck. By roadside ; found several
times.

RussuLA PURPURiNA A. & S. Wliitc-stemmed form ; found
only once but in large quantities.

* RussuLA puELLARis UMBONATA Peck. Only one plant
found.

* RussuLA ROSACEA Fr. Abundant.

* RussuLA scABRiuscuLA Peck. Apparently scarce.

* RussuLA SORDIDA Peck. Plentiful.

White : Mr. Desert Fungi ooS

{Maras7)ticae)

Lentinus lepideus Fr.

Marasmius androsaceus (L.) Fr. On hemlock needles after
rain in large quantities.

* Marasmius confluens Pcrs.

Marasmius oreades Fr. Forming large nearly complete fairy
rings.

Marasmius rotula (Scop.) Fr. Plentiful.

* Mar.\smius viticola B. & C. Plentiful.
Panus stypticus (Bull.) Fr. Common.

{Agaficcac)
{Evclatae. With neither volva nor annul us)

* Clitocybe ectypoides Peck. Plentiful on damp banks.

* Clitocybe infundibuliformis Schaeff. Common.
Clitocybe laccata Scop. Common.

* Clitocybe marginata Peck, sp. nov. Pileus fleshy, rather
thick, subcampanulate, becoming conv'cx, obtuse or broadly um-
bonate, glabrous or nearly so, dry, bay red verging to mahogany
color, the margin at first involute, flesh yellow ; lamellae narrow,
close, decurrent, yellowish, reddish on the edge ; stem nearly
equal, stout, hollow, glabrous, shining, yellowish marked with
reddish longitudinal lines ; spores subglobose, 5 // long, 4-5 n
broad. Pileus 5-8 cm. broad; stem 5-8 cm. long, 6-12 mm.
thick. Cespitose ; growing around decaying stumps. Septem-
ber.

A peculiar species well marked by the reddish edge of the
lamellae, reddish lines of the stem and yellowish flesh of the
pileus. It is referable to the second subsection of the Disciformes
according to arrangement in Syllogc Fungoruvi.

Clitopilus caespitosus Peck. Found only once but in large
quantities.

* Clitopilus carneoalbus With. Fairly abundant in certain
localities.

* Clitopilus prunulus Scop. Very common in meadows.
CoLLYBiA ABUNDANS Peck. Growing in large clusters.

* CoLLYBiA COLOREA Peck. Growing in large quantities, singly
and in clusters.

559 White : Mt. Desert Fungi

CoLLYBiA DRYOPHiLA Bull. Common.

* CoLLYBiA FAMiLiA Peck. Growing in great quantities over
old decaying stumps.

* CoLLYBiA PLATYPHYLLA Fr. Common.
CoLLYBiA RADiCATA Relh. Common and variable.

* CoLLYBiA RADICATA FURFURACEA Peck. Found Only oncc ;
seven or eight fine plants growing about a much-decayed stump.

Crepidotus applanatus Pers. Common.

* Crepidotus malachius B. & C.
Crepidotus versutus Peck.

* Entoloma cuspidatum Peck. Growing in large quantities.
Entoloma Grayanum Peck. Only a few plants seen.

* Entoloma Peckii Burt. Plentiful.

* Entoloma salmoneum Peck. Plentiful.

* Entoloma sinuatum Fr. One cluster of five specimens
found ; a fine large species.

* Leptonia asprella Fr. Short-stemmed form ; fairly common.

* Leptonia grisea Peck. Four or five plants found.

* Leptonia Formosa Fr. Fairly common.

* Leptonia serrulata (Pers.) Fr. A beautiful species which
appears to be rare.

* Leptonia subserrulata Peck. A few specimens found in
grass two hundred feet from the beach.

Mycena corticola Schum. A beautiful little fungus 2 cm.
high, growing on wood ; only occasionally seen.

* Mycena haematopoda Pers. Plentiful in certain localities.

* Nolanea conic a Peck. Appears to be scarce ; growing on
wood.

Omphalia campanella Batsch. Growing as usual in large
quantities on decaying wood.

* Omphalia scabriuscula Peck. Plentiful.
Panaeolus campanulatus (L.) Sacc.

* Panaeolus fimicola Fr. Common after rains.

* Pleurotus ostreatus (Jacq.) Fr. Plentiful, growing in dense
clusters.

* Pluteus cervinus Schaeff. Very common and variable ; a
white variety found and also a scaly one corresponding to Agarictis
ngois Pers.

White : Mr. Desert Fungi 500

* Pluteus granularis Peck. Only one plant found.

* Pluteus umbrosus Pers. Found twice on decaying wood.

* Tricholoma ALBOFLAViDUM Peck. Plentiful.
Tricholoma alhum (Schaeff.) Vr. Growinj.^ sinjijiy in leaf

mould.

Tricholoma Davisiae Peck. Plentiful.

"* Tricholoma fallax Peck. On dead leaves and twij^s after
rain.

* Tricholo.ma fu.mosoluteum Peck. Abundant.

* Tricholoma lugubre Peck. Plentiful.

* Tricholoma rutilans Schaeff. Appears to be rare.

* Tricholoma variegatum Scop. Found only twice ; a jiretty

species.

{]\'/atac. Provided with a veil, but no volva)

Cortinarius armillatus (A. & S.) Fr. Fairly abundant.

* Cortinarius Berlesianus Sacc. & Cub. Only two plants
found.

* Cortinarius corrugatus Peck. Found twice but in large
quantities.

* Cortinarius erraticus Peck. Plentiful ; growing in large
semicircles in damp leaf mould.

* Cortinarius multiformis Fr. Plentiful.

* Cortinarius similaris Peck. Two plants seen.
Cortinarius torvus Fr. Plentiful in one locality.
Cortinarius violaceus (L.) Fr. Abundant.

* Cortinarius Whiteae Peck, sp. nov. Pileus fleshy, hemi-
spherical, becoming nearly plane with a lobed wavy or irregular
margin, dry, glabrous, subpruinose, reddish or brownish-orange
verging to tawny ; lamellae subdistant, deeply and broadly emar-
ginate, reddish-brown becoming brownish-cinnamon ; stem long,
equal, solid, fibrous, colored like the pileus, adorned with darker
fibrous lines or striations ; spores subglobose, 7-8 // long, 7 n
broad.

Pileus 6-12 cm. broad; stem 7-12 cm. long, 1.5-2 cm.
thick.

Woods. August. A large species intermediate between
Dcrmocybc and Tclamonia, related to the former by its dry pileus
and to the latter by its general a.spect and stout solid stem. It
gives me pleasure to dedicate it to its discoverer, a most enthu-

561 White : Mt. Desert Fungi

siastic mycologist from whose specimens, notes, and colored sketch
of the fresh plant the description has been derived.

Flammula flavida Pers. On fallen spruce in large quan-
tities.

* Flammula granulosa Peck, sp. nov. Pileus thin, hemispher-
ical, becoming convex, dry, densely squamulose, tawny ; lamellae
thin, rather broad, subdistant, adnate or slightly decurrent, pale
yellow becoming tawny ferruginous ; stem fleshy, rather slender,
stuffed, glabrous or sometimes minutely squamulose, colored like
the pileus ; spores elliptic, generally uninucleate, 7-8 11 long, 4-5
fj. broad.

Pileus 1.5-3 cm. broad; stem 2.5-3 cm. long, 1.5-3 '^"^•
thick.

Decaying wood. August. This is a small species but easily
recognized by the squamules of the pileus which give to its sur-
face a granular appearance similar to that of the pileus of Lepiota
granulosa . At the base of the stem there is a colored membra-
nous disk by which the plant is attached to its place of growth.

* Flammula sapinea Fr. Appears to be scarce.
Flammula spumosa Fr. Plentiful.

* Hypholoma capnoides Fr. Found only once ; growing
singly and in clusters.

* Hypholoma hydrophilum (Bull.) Sacc. A single cluster of
five or six found in a hollow at foot of maple in decaying wood and
leaf mould ; spores a rich purple red.

Hypholoma incertum Peck. Plentiful in paths singly or in
clusters.

* Hypholoma perplexum Peck. Plentiful.

* Inocybe rimosa (Bull.) Fr. Two plants found in deep leaf
mould in woods.

* Naucoria erinacea Fr. Only a few plants found.
Naucoria semiorbicularis Bull. Large quantities found in

lawn.

* TuBARiA furfuracea Pcrs. A few specimens found in de-
caying leaves in damp woods ; a pretty little species.

{Anmilatae. Provided with a veil which remains as a
permanent annulus ; no volva)

* Agaricus abruptus Peck. Many very large specimens found.

White : Mr. Desert Fun(;i 502

Agaricus campestris L. Abundant in certain localities.
Armillaria mellea Vahl. Only one cluster found, which is
unusual for this species.

* Lepiota ILLINITA Fr. Found in only two localities.

* Lepiota naucinoides Peck. A single plant found in lawn.

* Pholiota agericola Peck. Only two plants found.

* Pholiota caperata Pers. Plentiful.

* Pholiota cerasina Peck. In large clusters on decaying

spruce.

* Pholiota squarrosoides Peck. Growing in large clusters;

plentiful in certain localities.

Stropharia Howeana Peck. Only a few plants found in rich
woods in leaf mould.

Stropharia semiglobata Batsch. Common in lawns.

* Stropharia stercorea Fr. Abundant after rain.

( Volvatac. Provided with a volva ; usually also
with a veil and annul us)

* Amanita crenulata Peck. Only a few specimens found.

* Amanita floccocephala Atk. Three or four plants seen.
Amanita Frostiana Peck. Abundant.

*Amanita mappa Fr. Growing singly in rich soil.

Amanita muscaria (L.) Fr. Abundant.

^Amanita muscaria Formosa Gonn. & Rab. Fairly common.
Amanita phalloides Fr. Abundant.

*Amanita rubescens Fr. Abundant.

Amanita spreta Peck. Rare.

*Amanita verna Fr. P'airly common.

*Amanitopsis vaginata (Bull.) Roz. A white variety also
found.

*Amanitopsis vaginata fulva (Schaeff.). Common.

*Amanitopsis vaginata crassivolvata Peck, var. nov. Pileus
broadly convex or nearly plane in the mature plant, sub-
umbonate, glabrous, viscid when moist, widely sulcate striate on
the margin, creamy yellow ; lamellae broad, close, free, white,
sometimes tinged with yellow ; stem long, equal, hollow or stuffed
with cottony fibers, colored like but a little darker than the pileus,
volva very thick, loose, white and woolly ; spores globose, 8-10 }i
broad. Pileus 7-10 cm. broad ; stem 15-20 cm. long, about 2
cm. thick.

563 White : Mt. Desert Fungi

In damp soil. Although the sheathed Ainanitopsis is a very-
variable species, this large form is so well marked by the very-
thick felty volva that it seems best to separate it as a variety. It
does not closely sheath the base of the stem as in the common
form but spreads away above and assumes the shape of a wine
glass or goblet.

* Amanitopsis volvata (Peck) Sacc. Growing singly ; three
specimens found.

Family Sclerodermataceae

Scleroderma vulgare Hormen. Appears to be rare.

* Scleroderma verrucosum (Pers.) Bull.

Family Lycoperdaceae
BoviSTA PiLA B. & C. Common in fields.
BoviSTA plumbea Pers. Common in sandy ground near the
ocean.

* Geaster fornicatus (Huds.) Fr. Found in one locality
only.

* Lycoperdon atropurpureum Vitt;
Lycoperdon gemmatum Batsch. Common.

* Lycoperdon Peckii Morg.
Lycoperdon pyriforme Schaeff Common.
Lycoperdon separans Peck.
Lycoperdon subincarnatum Peck.
Lycoperdon Turner: E. & E.

Family Phallaceae

* DicTYOPHORA duplicata (Bosc.) E. Fisch. Plentiful in
woods in one locality.

* Mutinus caninus (Huds.) Fr, A few specimens only found.

New York, N. Y., July 1902.

CONTRIBUTIONS FROM THE NEW VOHK BOTANICAE

GARDEN— No. Js

FOSSIL FERNS FROM THK LARAMIE GROUP

OF COLORADO.

By ARTHTR HOLLK'K,

N E W V () R K
11102

[Reprinted from Torreya, Vol. -', No. 10, October, 1902.]

FOSSIL FFRXS FROM THE LARAMIE GROUP ni'

COLORADO. *

Hv Arthur Holuck

(Plates 3 and 4)

Some twelve or thirteen years ago an extensive collection of
fossil plants, from the Laramie (Upper Cretaceous) Group of
Colorado, was made by Messrs. George Hadden and R. C. Hills,
for the late Dr. J. S. Newberry. This collection is now in the
paleobotanical museum of the New York Botanical Garden, and,
although partly labeled, was never reported upon by Dr. New-
berry.

Included in the collection are a few ferns, most of which are
more or less rare and some of them apparently represent unde-
scribed species or varieties. Of these the following have been
selected as noteworthy :

Anemia supercretacea sp. nov.

General form of frond, also nervation, unknown ; pinnae deli-
cate, narrowly conical in outline, gradually tapering to the tips;
pinnules entire, lower ones spatulate, distinct, somewhat decurrent
along and forming acute angles with the rachis, upper ones often
more pointed or becoming confluent and forming toothed or cren-
ulated tips to the pinnae. Plate 3, Figs. 6, 7.

In reddish shaly sandstone, Florence, Colo.

Anemia robusta sp. nov.

General form of frond, also nervation, unknown ; pinnae (?)
linear in outline, about 3 cm. in width ; pinnules entire, ovate to
subspatulate, with blunt wedge-shaped tips, about 2.5 cm. in

* Read before the Botanical Society of America, Pittsburg Meeting, July i, 1902.

145

146

length by 12 mm. in width, more or less confluent or decurrent
along the rachis, each one provided with a weak midvein. Plate
3» Fig. I.

In grayish sandstone, Florence, Colo.

AsPLENiUM MAGNUM Knowlton, Monog. U. S. Geol. Surv. 32 :

667. pi. 7p. /. 5-8 a

I have figured these specimens for the reason that they show
some slight differences or variations from the type figured by
Knowlton. In ours the pinnae are more deeply dissected, so
that the pinnules are more distinctly separated and are merely
confluent close to the rachis. They are also somewhat more
pointed and falcate in outline. Plate 4, Figs, i , 2.

In general appearance they both resemble Aspiduim Kcnnerleyi
Newb. (Boston Jour. Nat. Hist. 7:513. 1863; Monog. U. S.
Geol. Surv. 35: 11. pi. 16. f. 7,5. 1898), which differs chiefly
in the pinnules, for the most part, being finely denticulate near
their extremities, although "sometimes entire," according to the
description.

Fig. I in reddish shaly sandstone, Fig. 2 in black carbonaceous
shale, Florence, Colo.

Acrostichum Haddeni sp. no v.

General form of frond unknown ; pinnae (?) apparently 20 cm.
or more in length by about 3 cm. in maximum width, narrowed
to acute tips and with coarsely crenulate-serrate margins ; nerva-
tion consisting of a series of alternate pinnately arrranged veins,
extending from the rachis to the extremities of the serrations,
with the spaces between occupied by a network of fine nerves.
Plate 4, Figs. 3-6.

Named for Mr. George Hadden, the collector.

Figs. 3, 5 and 6 in grayish sandstone, Walsenberg Colo.; Fig.
4 in gray shale, F^lorence, Colo.

Polystichum Hillsianum sp nov.

General form of frond unknown ; pinnae (?) linear-oblong in
outline, about 3 cm. in maximum width, narrowed to 18 mm. at
the base, deeply pinnatifid ; pinnules entire, linear to subfalcate
in outline, blunt-pointed or obtuse at apex, alternately disposed
and confluent along the rachis ; nervation pinnate, consisting of

147

;i veries of primary nerves extending from the rachis to the c.\
tremities of the pinnules and numerous secondary nerves, once
forked. Tlate 4, Fig. 7.

Named for Mr. R. C. Hills, largely through whose efforts ihi
collection was acquired.

In reddish shaly sandstone, Florence, Colo.

Gleichenia rhombifolia sp. nov.

General form of frond, also nervation, unknown ; pinnae
linear in outline, acute, about 5 mm. in width, composed of sub-
quadrate pinnules, the lower ones distinct, acute and curved in-
ward at the tips, the upper ones becoming conlluent or closely
appro.\imated and all of nearly uniform length, giving to the
pinnae a delicate, ribbon-like appearance. Plate 3, Fig. 3.

In reddish shaly sandstone, Florence, Colo.

Gleichenia delicatula Heer (?), Fl. Foss. Arct. 3-: 54.
pi. g.f. lie; pi. lo.f. 16, f J
This specimen agrees quite satisfactorily with Heer's Fig. i le
above quoted, but not so closely with his other figures, which
show pinnae broader at their bases and not so uniform in width
throughout as in ours. Further than this, if we determine these
to be identical this determination would infer a very extensive
vertical range for the species, a range which we would hardly be
justified in assuming without question, upon such meager evidence
as that afforded by the single fragment represented by our speci-
men, although there would be no inconsistency in identifying a
species from the Kome beds with one from the I.^ramie. Plate

3. Fig- 4-

In reddish shaly sandstone, Florence, Colo.

Pecopteris (Cheilanthes) sepulta Nevvb. (?) Monog. U. S.
Geol. Surv. 35 : 12. pi. 62./. 5, ja, 6. 1898

Pecopteris (Phcgopteris) sepulta Nevvb. Proc. U. S. Nat. Mus.
5: 506. 1883.

After some hesitation I have finally decided to refer our speci-
men provisionally to this species. The differences are slight and
the resemblance is so close that a distinction .seems hardly to be
warranted. Plate 3, Figs. 5, 5a.

148

Newberry's figures show the pinnae at the extremity of the
frond to be either entire or wavy-margined and confluent along
the rachi's, while lower down they are subdivided into more or
less distinct pinnules and are apparently separated from each
other, although the basal pinnules are attached to the rachis.

The only difference on the part of our specimen is that the
basal pinnules are free ; but this may be and probably is due to
its representing a yet lower portion of a frond than is depicted in
either of Newberry's figures, which in themselves indicate a tran-
sition to a base similar to ours.

The reference to the Paleozoic genus Pecopteris will doubtless
be criticised on general principles, but rather than add to the
synonymy I have thought it best to adhere to the name last used
by Dr. Newberry.

In grayish sandstone, Walsenberg (?),* Colo.

Stenopteris (?) cretacea sp. nov.

General form of frond unknown, but apparently large and
strong ; each pinna or branch consisting of a broad-winged
rachis, with relatively remote, entire, strap-shaped pinnules, each
of which is traversed by a strong midvein from base to apex ;
secondary nervation unknown.

The fragmentary nature of our specimen renders accurate com-
parison difficult and it is possible that it might equally well be
considered under the genus Thaumatoptcris, hence the generic
reference is questioned.

In reddish shaly sandstone, Florence, Colo.

Explanation of Plates
Plate 3. Fig. I. Anemia robus/a sp. nov.; Fig. 2. S/eiiopt'-ris (?) cretacea sp
nov. ; P"ig. 3. Gleihcuia rhoinbifolia sp. nov. ; Fig. 4. Gleichenia delicatula Heer
(?); Figs. 5, 5a. Pecopteris ^ Cheilanthes) scpiilta Newb. (?); Figs. 6, 7. Anemia
supe<-cretacea sp. nov.

Plate 4. Figs. I, 2. Asplenium magnum Knowlton ; Figs. 3-6. A rostichiivi
//addein sp. nov. ; Fig. 7. Polysti hum I/illsianum sp. nov.
New York Bot.anical Garden.

* The label denoting the locality was not found, but the matrix is lithologically
identical with that of the Walsenberg sj. ecimens.

TOKKKYA, Vol.. 2.

Pl.ATK 4.

><*

V-

FOSSIL FERNS FROM COLORADO.

ToRREYA, Vol. 2.

Plate 3.

V

•A

-^

2

/:

x\

v,^

/\

A

FOSSIL FERNS FROM COLORADO.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 29

THE POLYPORACEAE OF NORTH AMERICA.
I. THE GENUS GANODERMA.

By WILLIAM ALPHONSO MURRILL.

NEW YORK
1902

[Reprinted from the Bulletin of thb Torrhy Botanical Club, 39: Oct., 1902.]

The Polyporaceae of North America. I. The Genus
Ganoderma-''

By William Am'Honso Murrill

It is evident, even to the beginner in mycology, that the present
classification of the higher fungi is not satisfactory. This is par-
ticularly true of the group containing Polyponts and its allies,
which is at present very imperfectly divided into genera, and con-
tains many species which are confused and uncertain, because of
the incompleteness of the original descriptions. Before the group
can be reduced to order, full descriptions must be had of all its
species, the type specimens in Sweden, England, Germany, and
elsewhere must be examined, and sufficient knowledge gained of
the life and habits of each plant to determine its natural position
in the group. In this paper an attempt is made to present in
order the members of the " Incidiis " group of Polyponis occurring
in North America. A large number of the specimens examined
have been collected during several years past by Professor Lucien
M. Underwood, and by various contributors to his private herba-
rium. The extensive literature and large collections of the New
York Botanical Garden and Columbia University have also been
most generously placed at the writer's disposal.

I desire also to express my appreciation of the many courtesies
extended me by Professor Paul Magnus, Mr. P. Hennings, Pro-
fessor Kjellman, Professor T. M. Fries, Professor H. Von Post,
Mr. Lars Romell, Mr. W. B. Hemsley, Mr. George Massee and
others, while studying forms of this group in Germany, Sweden,
and England.

History of the Genus

The species upon which this genus was founded was first
described in its immature form as an agaric, but when perfect
specimens were obtained it was placed in Boletus, a genus es-
tablished by Dillenius in 17 19 to include all pore-bearing fungi,
and later adopted by Linnaeus, although Micheli had erected

* Read by invitation before the Botanical Society of America, Pittsburg, Tune,
1902.

599

600 MuRRiLL : The Polyporaceae of North America

the genus Poly poms as early as 1729. Polyporus was used by
Persoon in his Synopsis only as a subgenus under Boletus, from
which position it was later raised by Fries to equal rank
with Boletus and made to include the woody forms of pore-
bearing fungi. In 1 85 I, Fries tried the experiment of breaking
Polyporus up into Forties, Polystictus, and Poria, a division which,
though discontinued in his later works was resurrected by Cooke in
his Praeciirsores in 1885 and adopted by Saccardo and others since.
When a systematic arrangement of the Polyporaceae of Fin-
land was undertaken by Karsten (Rev. Mycol. 1881), one of the
new genera established was Ganodcrnia, which was based upon
the laccate character of pileus and stipe and embraced Polyporus
lucidus only. In Karsten's " Finlands Basidsvampar " published
in 1889, Ganoderma is characterized as follows: " Basiderna 4-
sporiga, n. klotrunda, cystiderna icke anmarknings varda. Spo-
rerna aggrunda eller elliptiska, vartiga, gulbrunaktiga." Although
no mention is here made of the laccate pileus and stipe, no
other species are included with G. lucidum in the genus. As
a synonym of Ganoderma, Karsten here mentions Placodes Quel.,
which genus was erected in 1886 to include a variety of forms
which were " covered with a hard crust, without zones or concen-
trically sulcate, persistent, woody." Along with P. lucidus in one
of the subdivisions are placed also P. dryadeus, P. resinosus, P.
erubescens, and P. helveolus.

The transfer of G. lucidum from Ganoderma Karst. to Phaeoporus
Schroet. m Schroeter's flora of Silesia was a violation of modern
principles and consequently could not meet with general favor. In
1887, Patouillard, following the lead of Quelet, extended the range
of Ganoderma to include all forms of Polyporaceae with colored
spores, adhering tubes and shining crusted pilei ; and in 1889 he
published a partial monograph of the genus as extended, listing
forty-eight species, which were arranged in subgroups upon spore
characters. As to the wisdom of so broad a treatment of the
genus mycologists appear to be divided. The following synopsis
includes only those species which, while essentially alike in internal
structure, possess a shining varnished surface produced by the
exudation and hardening of a reddish juice similar to that found in
G. pscudoboletus.

MuRRiLL : The Polyporaceae of North America <;i»1

Synopsis of tlie Nortli American HpecleH

1. Context pallid to tawny. 2.
Context umbrinous-chestnut. 5.

2. Spores over 6 // long, verrucose, ferruginous. j.
Spores less than 6/z long, smooth, pale; pileus less than 3 cm. broad, margin en-
tire ; tubes not stratified. 4. G. parvulum.

3. Context ochraceous to fulvous ; plants stipitate or sessile, growing on deciduous

trees. 4.

Context pallid; plants usually stipitate, annual, growing on hemlock, i. G. tsttiae.

4. Plants stipitate, or rarely sessile, perennial ; margin of pileus truncate at maturity.

2. G. pseudoboletus.
Plants sessile, annual ; margin of pileus acute. 3. G. sessile.

5. Spores smooth, pale yellowish brown. 6.
Spores roughly echinulate, dark brown ; pileus sessile, very thick, its diameter

decreasing downward ; context less than one third the length of the tubes.

5. G. Oerstedii.

6. Pileus zonate, even ; tubes not stratified. 6. G. zonatum.
Pileus sulcate, azonate ; tubes stratified. 7. G. sulcatum.

I. Ganoderma tsugae sp. nov.

A conspicuous reddish-chestnut fungus growing on dead or
dying hemlock. Pileus corky to woody, fan-shaped, convex
above, concave below, 4-20 x 5-25 x 1-4 cm.;* surface gla-
brous, uneven, concentrically sulcate, laccate, lustrous, yellowish-
red to mahogany-colored, at length black ; margin light-yellow,
acute, becoming concolorous, truncate, and marked with many
shallow furrows, often undulate and at times more or less lobed :
context soft-corky, radiate-fibrous, white or nearly so, 1-3 cm.
thick; tubes annual, 0.5-0.75 cm. long, 6-4 to a mm., brown
within, mouths circular or polygonal, white to light-cinnamon,
edges obtuse, becoming acute : spores ovoid, obtuse at the
summit, attenuate and truncate at the base, verrucose, yellowish-
brown, 9-1 1 X 6-8 fi : stipe lateral, ascending, frequently forked,
cylindrical, equal, 2-20 x 1-4 cm., resembling the pileus in color,
surface and context.

This fungus occurs only on Tsuga Cattadefisis, upon the decay-
ing trunks and roots of which it is very abundant. In West Vir-
ginia and New York it has been found full grown as early as June,
but it may not reach maturity until autumn, when it speedily de-

* The.se figures indicate length, breadth and thickness respectively. When the stipe
is lateral the pileus is an outgrowth from it and length is measured in the direction of
this growth. When the pileus is sessile, the tubercle marking the beginning of growth
may be considered a lateral stipe much reduced. When the question of length is
settled, breadth and thickness need no explanation.

602 MuRRiLL : The Polyporaceae of North America

cays or falls a prey to insects. So subject is it to insect attack
that recognizable specimens are rare in the herbarium. As its
ordinary host is confined to America, there is no reference to it in
European literature, unless the plants found on Picca cxcclsa by
Karsten in Sweden belong here rather than with G. psciidoboletus.
Extensive collections of various forms of this species were
made by Professor Underwood at Syracuse, N. Y., in July 1884,
and at West Goshen, Conn., August 1896. Miss A. M. Vail
brought specimens from Tyringham, Mass., in August 1897,*
Professor G. F. Atkinson has figured the species in plate 66 of
his work on mushrooms. Wherever the hemlock grows it appears
to be common. It is nearly related to G. pseiidobolctus, but is
annual, grows on coniferous trees, decays soon after maturity, and
is usually much lighter in weight and paler in substance.

2. Ganoderma pseudoboletus (Jacq.)
AgaricHS pseudoboletus Jacq. Flor. Austr. i : 26-27. //. ^/.

1773-

Boletus rugosus Jacq. Flor. Austr. 2 : 44. //. i6g. 1774.

Boletus lucidus 'Ltyss. Flora Halensis, 300. 1783.

Boletus obliquatus Bull. Herb, de la France,//. 7. 1780; //.

459- 1790.

Poly poms lucidus Fries, Syst. Myc. i : 353. 182 1.

Polyporus laccatus Pers. Myc. E!ur. 2: 54. 1825.

Poly poms Curtisii Berk. Kew Gard. Misc. i : 10 1. 1849.

A large fungus common on decaying trunks and stumps of
deciduous trees conspicuous on occount of its brilliant varnished
appearance. Pileus corky to woody, usually kidney-shaped,
convex above, concave below, 2—20 x 2—25 x i— 5 cm., yellow to
reddish-chestnut or black ; surface glabrous, shining, laccate,
broadly sulcate and usually marked with concentric lines or bands
of a darker color ; margin white or light yellow, sterile, thin,
acute, becoming truncate, sulcate, and concolorous as new strata
are added ; context soft-corky or woody, radiate-fibrous, concen-
trically banded, ochraceous above, tawny next to the hymenium :
tubes one- to many-layered, the strata varying in distinctness,
0.5-2 cm. long, 3-5 to a mm., brown within ; mouths circular to

* During the summer of 1902 fine specimens have been collected in the hemlock
grove of the New York Botanical Garden by Messrs. Burnham and George, in
Connecticut by Miss White, and in Ohio by Professor A. D. Selby.

MuRRiLL : The Polyporaceae of North America 603

hexagonal, white or yellow, at length brown, dissepiments c-ntire,
obtuse, becoming acute : spores ovoid, obtuse, at the summit, at-
tenuate and truncate at the base, yellowish-brown, verrucose, 9—1 1
X l-^ !>■ '■ stipe lateral, excentric, central, or wanting, erect to as-
cending, 0-30 X 0.5-4 cm., equal, irregular, or enlarging above,
concolorous, glabrous, shining, laccate, the substance similar to
the context and darker at the center.

On living or dead trunks, stumps, or roots of oak, alder, hazel,
maple, willow, honey-locust, sweet-gum, and beech in Sweden,
Germany, Bavaria, France, England, America and Australia.
American material has been examined from New York, New Jersey,
Delaware, Pennsylvania, Ohio, Indiana, Kentucky, Michigan, Vir-
ginia, North Carolina, South Carolinia, Georgia, Florida, Alabama,
Mexico, and Nicaragua.

This fungus has been found fossil in the lake dwellings of
Switzerland and has been known for a long time on account of its
conspicuous coloring caused by a thick glutinous juice which ex-
udes from its surface and dries upon it as a thin lustrous coating.
Albino or semi-albino forms occur where the coating is lacking or
incomplete. In age the varnish disappears and the pileus takes on
a grayish weather-beaten appearance. As the young pileus begins
to form at the end of the stipe it is white or yellow in color with-
out varnish and somewhat resembles an unexpanded agaric. It
is this stage that Jacquin figured and described in his Flora i\us-
triaca as Agariciis pscudoboletus. Several immature plants were
found by him in a grove growing about the base of a dead oak
trunk. The description he gives is quite a good one and, taken
with the fine colored plate, leaves no doubt as to the identity of
the specimens. The succeeding year he collected several mature
plants which he described as Boletus riigosiis as follows : —

" Fungus speciosus putridis arborum truncis innascitur, totus lignoso-coriaceus et
persistens. Stipes durus, inaequalis, badius, vernice veluti obductus, calamum vel
policem crassus radone voluminis ipsius fungi, pileum gerit plerumque subdimidiatum,
dum laterali ejusdem parti adnecitur. Hie superne planus est, rugosus primum, ex
rubro badius et nitidissimus, tandem hepaticus minusque nitens. Corticis pauca
substantia est interne coriacea, holoserisea, cinnamomea, tenax atque ad fomitem
apta. Substantia tubulosa concolor, crassa, a corticosa separabilis, subtilissime
porosa ; subtus punctata, in principio pallens, sensim magis cinnamomea ; ad cibum
inepta. Fungi duo, ex eodem loco exorti, et majores, in tabula proponuntur,
hinc atque illiuc spectati. Turn fungulus minor; et fungi pars, ut pateant in-
teriora."

604 MuRRiLL : The Polyporaceae of North America

The specimen upon which Leysser's Bolehis liicidus is founded
was collected by Curtis on a decaying hazel stump near London
in November, 1780, and figured in Flora Londinensis, 4 : //. 22^.
Leysser's description is as follows : —

" Boletus lucidus stipitatus, pileo coj-iaceo castaneo liicido sulci s circttlaribtis, poris
miniitissimis, albis, stipite laterali.^''

Specimens collected by Rev. M. A. Curtis in South Carolina
seemed to Berkeley sufficiently distinct from Leysser's plant to
constitute a new species, which he thus describes in Kew Garden
Misc. I : 101. 1849 : —

" Pileus excentric, soft-corky, sulcate, zonate, ochroleucous, in places sanguine-
lacquered ; stipe elongated, rugose, sanguine-lacquered ; hymenium from white to
ochraceous, pores punctiforra."

A Special study of many specimens called Poly poms Curtisii
Berk, shows them to be only variations of G. psetidoboletns due to
age, rapidity of growth, and perhaps to differences in the host.
The yellowish form so common in the southern part of the United
States is figured by BuUiard and Gillet as also occurring in France,
while farther north the color is usually darker and more lustrous,
the texture firmer, and the hymenium less stratose. These differ-
ences are not sufficiently constant, however, to enable one to
separate the forms geographically.

Another interesting variation observed by Langlois in Loui-
siana is so distinct from the ordinary form of the plant that it
might easily deceive the skilled mycologist. Specimens in the
Underwood herbarium have the stipe exactly central, the pileus
very even and thin, at first deeply infundibuliform with white
margin, at length becoming nearly plane, reddish-brown, and pol-
ished, with the margin concolorous.

3. Ganoderma sessile sp. no v.

A large sessile plant, with wrinkled varnished cap and acute
margin, found on decaying deciduous trees. Pileus corky to
woody, dimidiate, sessile, imbricate or connate at times, conchate,
thickest behind, thin at the margin, 5-15 x 7-25 x 1-3 cm.,
yellow to reddish-chestnut, at length opaque dark brown ; surface
glabrous, laccate, shining, radiate-rugose, concentrically sul-
cate, usually marked near the margin with alternating bay and
tawny zones ; margin very thin and acute, usually curved down-

MuRRiLL : The Polyporaceae of North America 005

ward, often undulate, not becoming truncate, white, at length con-
colorous : context soft-corky or woody, radiate-fibrous, concen-
trically banded, ochraceous-fulvous ; tubes 0.5-2 cm. long, 5-3 to
a mm., brown within, mouths circular or angular, white to gray-
ish-brown, edges thin, entire : spores ovoid, obtuse at the summit,
attenuate and truncate at the base, verrucose, yellowish-brown, 9-
1 1 X 6-8 //.

This .species occurs on deciduous trees and has the general
habit of G. pscndoboletus, from which it differs in being annual and
sessile, with a very acute margin and a more rugose surface. So
far as I have been able to determine, it does not occur in Europe,
and has not yet been figured. Specimens have been collected by
Professor L. M. Underwood on decaying oak at Greencastle, Ind.,
Oct. 1894, and [at White Plains, N. Y., May 1897, and on dead
sycamore at Fort Lee, N. J., May 1899, and March 1901.
Plants collected by Morgan in the Miami Valley, Ohio, Dec.
1894, were placed under G. pseudoboletiis and referred to in the
following note : " As it grows in this region, the stipe is always
more or less deformed and often wanting ; the pilei, when sessile,
are sometimes imbricated and connate." In Dec. 1896, it was
found by C. F. Baker, at Auburn, Ala., and in Nov. 1897, Rev.
A. B. Langlois collected it near St. Martinsville, La., on decaying
logs in low woods and marked it " Perhaps a form oi F. liccidus?"
Miss Sadie F. Price has recently sent to the New York Botanical
Garden from Bowling Green, Ky., two specimens of this plant
collected on oak in Nov. 1901, and Feb. 1902. Specimens sent
to Kew by American collectors are included there under G. pscn-
doboletus.

4. Ganoderma parvulum sp. nov.

A very small sessile fungus shining bay above and honey-
yellow below. Pileus woody, nearly circular in outline, attached
by a point, convex above, plane or convex below, thickest behind,
2 X 2.5 X I cm.; surface glabrous, laccate, azonate, slightly tuber-
cular, very lightly marked with a few concentric furrows, margin
acute : context soft-woody, pale ochraceous, o. 5 cm. thick, with
dark horny radiations from the point of attachment : tubes not
stratified, 3 mm. long, 5 to a mm., umbrinous within, mouths
polygonal, honey-yellow, dissepiments entire, obtuse : spores sub-
globose, smooth, pale yellowish-brown, 4 x 5 />«.

Collected by C. L. Smith in Nicaragua during the winter of

606 MuRRiLL : The Polyporaceae of North America

1 891— 1892. This beautiful little plant is closely related in habit
and general appearance to the American species of Ganoderma
occurring farther north, but it is of much smaller size and its
spores are quite distinct in form, size and color. It is possible
that the specimens I have are not quite mature.

5. Ganoderma Oerstedii (Fries)

Polyponis Oerstedii Fries, Nov. Sym. 63. 185 1.

Pileus reniform, ungulate-applanate, gibbous at the base, a foot
in diameter ; surface horny-incrusted, very glabrous, adorned with
shallow furrows, which almost disappear wath age, shining reddish
chestnut becoming almost black ; margin very obtusely truncate
and marked with concentric furrows, the upper annual growths
exceeding the lower ; context partly hard and horny and partly
floccose, umbrinous next to the tubes, more tawny beneath the
cuticle, very thin in older specimens, the tubes forming the prin-
cipal part of the pileus : tubes contiguous and hence indistinctly
stratified, 3—5 cm. long, umbrinous within, mouths lighter in color,
dissepiments entire, obtuse : spores broadly ellipsoid, truncate,
ver>' dark yellowish-brown, abundantly and roughly echinulate,
II X 9/x

Tw^o imperfect specimens of this plant are in the herbarium of
the New York Botanical Garden, one collected by C. L. Smith in
Nicaragua, and the other by C. T. Townsend on an orange tree
in Jamaica. These specimens, while corresponding with the
Friesian types at Upsala, hardly justify any considerable departure
from the Friesian description.

6. Ganoderma zonatum sp. nov.

A soft laccate fungus of medium size marked with numerous
tawny and chestnut-colored zones. Pileus very soft-corky, ses-
sile, dimidiate, applanate or convex above, concave below, gla-
brous, zonate, not sulcate, 5x7x1.5 cm. ; margin velvety, acute,
becoming obtuse and concolorous : context very soft, floccose,
radiate-fibrous, concentrically banded, 0.5 cm. thick, chocolate-
brown : hymenium velvety, not stratose, tubes i cm. long, 3—4 to
a mm., umbrinous within ; mouths white to umbrinous, regular,
polygonal, stuffed at first with whitish material, covered o. 5-2 cm.
from the margin with yellowish or reddish varnish ; dissepiments
entire, obtuse to acute : spores elongated ellipsoid, smooth, pale
yellowish-brown, 8-10 x 4-6//.

MuRRiLL : The Polyporaceae of North America 007

Collected by Professor Underwood in Plorida. Closely re-
lated to G. stdcatiim, but differing in general appearance, size of
pores, and in the extent and arrangement of varnish upon the hy-
meniuni. A similar coating of varnish exists in G. sulcatum and
G. pseudoboletus at times, but in neither of these has it been found
as a broad distinct marginal band. This band is sometimes split
by the growth of the margin and a zone of pores appears between
the two laccate portions. Additional material may reveal the ex-
istence of intermediate forms connecting this species with G. sul-

catuui.

y. Ganoderma sulcatum sp. nov.

A large sessile plant without zones, but marked with a few
conspicuous concentric furrows. Pileus corky, dimidiate, sessile
'or arising from a lateral tubercle, plane or convex abov-e, thickest
behind, 8x11x2 cm.; surface laccate, glabrous, azonate, fulvous
to chestnut, deeply sulcate ; margin rounded, velvety, ochroleu-
cous, at length concolorous : context very soft, floccose, radiate-
fibrous, concentrically banded, i cm. thick, umbrinous-chestnut :
tubes indistinctly stratified, 1.25 cm. long, 4-5 to a mm., umbri-
nous within, mouths whitish or yellowish, at length umbrinous,
dissepiments entire, obtuse : spores ellipsoid, pale yellowish-brown,
smooth, 8-10 X 4-6 iJL.

This plant was collected on soft palmetto logs in Florida by
Mr. C. G. Lloyd, January, 1897 ; type in New York Botanical
Garden. It is very nearly related to G. zonatiim.

Species inquirendae

Fomes incrustatus Fries, collected in Costa Rica by Oersted.
Pileus 5-7 cm. broad, sordid umbrinous, laccate : context scanty,
pallid, tubes concolorous ; stipe central, torulose, unequal. This
species was not transferred by Patouillard to the genus Gano-
dernia.

G. nutatis {FrxQs) Pat., also collected by Oersted in Costa Rica.
Pileus scarcely 2 cm. broad, orbicular, pallid to bay-black, lac-
cate ; context scanty, pallid, pores pallid to brown ; stipe long,
slender, twisted as though climbing, recurved at the apex, giving
the pileus the appearance of nodding ; spores ovoid, brown,
echinulate, 12-15 x lO/i.

Specimens seen in foreign herbaria appear to be very distinct.
Fries' description in Novae Symbolae 61-62 is quite full and can-

608 MuRRiLL : The Polyporaceae of North America

not be improved upon at this time on account of lack of material
for critical study.

G. nitcns (Fries.) Pat. On trunks in tropical America. Pileus
large, sessile, semiglobose, chestnut to black, shining, hanging by a
process behind ; context umbrinous, tubes very long and slender.
It is quite possible that this is not a distinct species.

G. neglectum Pat. Pileus sessile, 8-10 cm. broad, reddish-
black, shining ; context brown, tubes white, very short ; spores
globose, yellow, roughly echinulate, 11 — 12 //. On trunks in;
Nicaragua,

New York City.

I

I

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 30.

STUDIES ON THE ROCKY MOUNTAIN

FLORA-IX

By p. A. RYDBERG

NEW YORK
1902

[Reprinted from the Bulletin of the Torkey Botanical Club, ag : Dec, 1902 ]

Studies on the Rocky Mountain Flora— IX

I'.Y 1'. A. IvYDllKRi;

THE NYCTAGINIACEAE OF THE ROCKY MOUNTAIN REGION

The family Nyctaginiaceac has been sorely neglected by the
systematic botanists in this country. Until lately we have had not
even an attempt at a monographic work since Dr. Gray's notes were
published in the Botany of the United States and Mexican Boun-
dary Survey in 1859. Gray's treatment there as a whole can
scarcely be regarded as an improvement on that by Choisy, pub-
lished ten years earlier in De Candolle's Prodromus, and of course,
both are now out of date. Professor Hcimerl's treatment in Die
ucxturlichcn PJJanzcnfamilioi is as good as could be expected from
a European monographing an almost exclusively American»family ;
but this gives little help beyond the genera. Recently there has
appeared a revision of the family by Marcus E. Jones * as it is
represented on the Great Plateau. As the territory covered by
Jones practically includes that treated in this article, it would seem
superfluous to duplicate the work ; but I have had the advantages
of a large library and the rich collections of Columbia University,
the United States National herbarium, and the New York Botan-
ical Garden. These advantages are, however, somewhat balanced
by Mr. Jones' longer field experiences. Jones' paper is valuable
because it gives fuller descriptions of many poorly known species,
descriptions drawn by a botanist who knows the species in the
field. It is deplorable, however, that this paper in many places
shows a good deal of carelessness, especially in the matter of cit-
ing publications. Under Allionia, it has for instance :
" 5. A. GLABER t (Wats.) Kuntze, Am. Nat. 76," and
"7. A. AGOREGATA (Vahl) Sprcng. Ic. 5 437."
In the first case, one would suppose that Kuntze published the
combination in the American Naturalist, while the fact is that Wat-
son there published Oxybaplius glabcr, on which Allionia glabra is
based.

* Contributions to Western Botany, lo : 34-54- June, 1902.
•)• This should have been A. glabra.

680

681 Rydberg : Studies on the Rocky Mountain Flora

If interpreting the second case in a similar way, one would
come to the conclusion that OxybapJuts aggregates Vahl was pub-
lished in Vahl's Icones, if not in Sprengel's Icones ; but neither is
the case, for no book with that title was ever published by either
Vahl or Sprengel. From Mr. Jones' citation no one could imag-
ine that Ic. 5 437, stands for Cavanilles' Icones, where Mirabilis
aggregata appeared. This is not the earliest appearance of the
name aggregata, however, for this was originally published by
Ortega * as Calyxhymcnia aggregata. As Cavanilles' plant is differ-
ent from Ortega's, the former being Allionia decumbens (Nutt.)
Spreng., the latter A. agregata (Ortega) Spreng. as shown below,
Jones citation becomes not only unintelligible but also incorrect.
It would have been much better to leave out the citation of pubh-
cations altogether, which by the way is advisable for anyone who
does not possess good library facilities.

Key to ttie Genera of tlie Rncky mountain Region

Bracts distinct.

Fruit crested or winged ; bracts in a whorl at the base of the head-like cluster ;

perianth salverform. I. Abronia.

Fruit globular, neither crested nor winged ; bracts attached each to a pedicel of the
umbel-like or corymbose inflorescence ; perianth funnelform. 2. Hertnidium.
Bracts united.

Fruit neither strongly tubercled nor winged.

Fruit not ribbed ; involucre herbaceous, little if any enlarging in fruit, not be-
coming membranous.
Stamens usually 5 ; involucres campanulate, not enlarged in fruit.

3. Quamoclidion.
Stamens 3 ; involucre rotate, somewhat enlarged in fruit in the manner
of the next genus, but not membranous. 4. Allioni4la.

Fruit ribbed ; involucre rotate, in fruit becoming much enlarged and mem-
branous. 5. Allionia.
Fruit with two rows of strong tubercles on the back and surrounded by two toothed
inflexed wings. 6. Wedelia.

I. ABRONIA Juss. Gen. 448. 1789

Tricratus L'Her. ; Willd. Sp. PI. i : 807. 1799.

Cycloptera Nutt.; Gray, Am. Jour. Sc. II. 15: 319. 1853.

In the original publication, no type species was mentioned.
The genus was described from a plant collected on De la Peirouse's
journey in California and cultivated by Mr. Colignon. Hooker in

* Nov. aut Rar. PI. 8 : //. //. 1798 (or 9?).

Rydrerg : Studies on the Rockv Mountain Im.ora 682

his Exotic Mora, />/. igj; & /^y, identifies Colignon's plant as
Abronia uuibcllata. The type of Iricratus is the same, and that of
Cycloptera is A. cycloptcra.

Fruit narrowly winged or crested ; wings or crests not completely encircling the fruit.
Fruit biturbinate, /. e., tapering at both ends, irregularly ridged or crested.
Flowers about 2 cm. long ; limb 5-10 mm. wide.

Hracts broadly obovate, over l cm. long. I. A. /radians.

Hracts ovate-lanceolate, less than i cm. long. 2. A. nui/al.i.

Flowers about I cm. long ; limb 3-5 mm. wide ; bracts ovate to lanceolate-
ovate, about 5 mm. long.
Petioles of the stem-leaves shorter than the very thick blades ; plant low,

about I dm. high. 3. A. ptnnila.

Petioles of the stem-leaves much longer than the moderately thick blades ;
plant slender, 2-4 dm. high. 4. A. ammophtla.

Fruit turbinate or obpyramidal, /. <•., almost truncate above, distinctly winged ; the
wings very broad above.
Plant almost acaulescent ; stem and leaves greatly surpassed by the long

peduncles. 5. A. nana.

Plant with an elongated stem.

Bracts broadly ovate or obovate, acute or obtusish.

Stem distinctly viscid-pubescent ; leaves scabrous-puberulent ; bracts
I-1.5 cm. long.
Blades of the stem-leaves elliptic ; bracts broadly obovate, 12-

15 mm. wide, obtusish. 6. A. salsa.

Blades of the stem-leaves lanceolate; bracts oval, acute, about
6-7 mm. wide. 7. A.fallax.

Stem finely puberulent or glabrous ; leaves glabrous ; bracts 5-8 mm.
long.
Stem puberulent. 8. A. elliptica.

Stem glabrous. 9. A. glabra.

Bracts oblong-lanceolate or lanceolate, attenuate or cuspidate.

Stem glabrous. 10. A. lanceolata.

Stem more or less pubescent.

Stem puberulent ; wings with double lamina, coriaceous.

11. A. Carlefoni.
Stem villous ; wings with single lamina, membranous.

12. A. villosa.
Fruit completely surrounded by the broad netted-veined membranous wings.

Flowers 3 cm. or more long ; limb about I cm. wide ; peduncles longer than the

•eaves. 13. A. cydoptera.

Flowers 1.5-2 cm. long; limb about 5 mm. wide.

Stem glabrous or nearly so ; peduncles often nearly equalling the leaves.

14. A. pedunculata.
Stem densely pubescent ; peduncles at least in flower much shorter than the
leaves. 15. A. viicrantha.

I. Abronia fragrans Nutt.; Hook. Kew Journ. 5: 261, 1853
On the plains from South Dakota to Montana, Idaho, New
Mexico and Kansas.

683 Rydherg : Studies on the Rockv Mountain Flora

2. Abronia nudata sp. no v.

Perennial, stems very long and decumbent, sparingly hispidu-
ious : leaves thick, glabrous; petioles 2-10 mm. long; blades
broadly lanceolate, obtusish, 2-5 cm. long : peduncles 4-6 cm.
long, almost glabrous : bracts ovate-lanceolate, less than i cm.
long: flowers many, about 2 cm. long; limb 5-6 mm. wide:
achenes very irregular, the inner ones of the head bipyramidal,
thickest a little above the middle, crested, 8-10 mm. long, those
of the margin very obliquely ovoid-fusiform, scarcely at all crested.

This species is nearest related to A. fragrans and A. ammo-
_phila. From the former it differs in the small bracts, smaller
leaves and almost glabrous stem ; and from the latter in the much
longer flowers and the scanty or no pubescence.

Montana: Colgate, near Glendive, 1892, Sandbcrg, Mac-
Doiigal &■ Heller, 1016 (type in herb. N. Y. Bot. Garden).

3. Abronia pumila sp. nov.

Perennial, cae.spitose : stems ascending, about i dm. long,
puberulent : leaves very thick, minutely puberulent ; petioles 1-2
cm. long, those of the stem-leaves usually shorter than blades ;
these 1.5-3 cm. long, oval, elliptic or somewhat ovate : peduncles
2-3 cm. long, puberulent : bracts elliptic-ovate or lanceolate, 7-8
mm. long, short-acuminate: flowers about 12 mm. long; limb
about 3 mm. wide : fruit bipyramidal, /. c, tapering towards both
ends but more so downwards, the greatest breadth being about one
third from the apex, merely crested or the inner ones somewhat
winged.

Dr. Heimerl referred this to A. truncala Torr., but the fruit is
very unlike that of that species. A. pumila is nearest related to
A. ammophila, from which it differs mainly in the thicker and short-
petioled leaves and the dwarfed habit.

Utah : Emery, 1 894, M. E. Jones, 5445q (type in U. S. Nat.
herb.); six miles up Salida Caiion, 5416a.

4. Abronia ammophila Greene, Pittonia, 4: 226. 1900
Abronia arenaria Rydb. Mem. N. Y. Bot. Garden, 1:137. 1900.
Not Menz. 1827.

Yellowstone National Park.

5. Abronia nana S. Wats. Proc. Am. Acad. 16 : 294. 1870
Southern Utah to Arizona and southern California.

RvDBEKci : Studies on tiik R()ck\ Mountain Floka 684

6. Abronia salsa sp. nov.

AOroiiin frn^nrfi.s )ri. Wats. King's Rep. 5: 2S4. 1871. Not
Nutt.

Perennial : stem ascending, 3-4 dm. high, densely viscid-pubes-
cent especially above, stout : leaves very thick, puberulent ; petioles
2-3 cm. long ; blades oval or elliptic, obtuse, or the lowest rounded
oval, 3-5 cm. long, 1-3 cm. wide : peduncles 5-7 cm. long, den.sely
viscid-pubescent ; bracts rounded obovate, about i 5 mm. long and
often as broad, pubescent : flowers many, about 2 cm. long with a
limb about 4 mm. wide : fruit about i cm. long, similar to tho.se of
. i. lanccolata and - /. fallax but the wings are rather thicker.

The sheets in the National Herbarium bear the following label-
ing by Dr. Heimerl : ''Abronia fragrans Nutt. {ad fori nam ellip-
ticam (Nelson) accidcns .') ; " but neither A. sa/sa nor A. cUiptica
A. Nelson have the fruit of A. fragrans, and therefore can not be
regarded as forms of that species, A. salsa differs from A. clliptica
in the stouter viscid-pubescent stem, the larger bracts and flowers.
It grows in sandy saline soil at an altitude of 1000- 1300 m.

Utah: Salt Lake City, 1869, 5. Watson, p6j (type in herb.
Columbia University) ; same locality, Capt. Slansdnrj, and 1871,
Harden; Silver Reef, 1894, J/. E. Jones ; Grand Junction, 1900,
.S". G. Stokes.

7. Abronia fallax Heimerl sp. nov.

Perennial ; stem erect, almost shrubb}' below, branched, straw-
color or white, viscid-pubescent above : leaves thick, puberulent,
erect ; petioles 1-2 cm. long ; blades of the lower oval, of the
upper long-lanceolate, 2-5 cm. long, 1-1.5 cm. wide, obtuse or
the upper acute : peduncles 1-2 cm. long, densely viscid-pubes-
cent; bracts 5-6, oval, about 1.5 cm. long, 6-7 mm. wide:
flowers many, about 15 mm. long : fruit with the wings about 6
mm. wide and 8 mm. long, cuneate-obpyramidal.

The type in the U. S. Nat. herbarium bears the following re-
marks from the hand of Dr. Heimerl : " Abronia fallax ni. — / Hy-
brid a ex A. fragrans et A. tnrbinata. — Ab A. fragrans ant hoe ar-
pinis alls apice transverse dilatatis, ab A. turb. bract eis capitulonmi
niagnis, scariosis divcrsa^ * To me it seems that the plant has
little to do with A. fragrans and is less related to . /. tnrbinata

* In a letter lately received from Dr. Heimerl, he has authorized me to use his
name and lujtes.

685 Rydberg : Studies on the Rocky Mountain Flora

than to A. elliptica A. Nelson. From this it differs in the upright
stem, which is decidedly viscid-pubescent instead of merely pu-
berulent.

Utah : Salt Lake City, 1879, M. E. Jones, ijjy (type in U. S.
Nat. herb., also in herb. Columbia University).

8. Abronia elliptica a. Nelson, Bull. Torrey Club, 26: 7.

1899

Wyoming and Colorado.

9. Abronia glabra sp. nov.

Perennial : stem ascending, about half a meter high, glabrous,
straw-colored: leaves thick, glabrous; petioles 1-2 cm. long;
blades oval to oblong, 1-4 cm. long, obtuse : peduncles oval, 2-4
cm. long, glabrous ; bracts obovate or about 5 mm. long, acute :
flowers 12-15 mm. long, numerous: fruit cuneate-obpyramidal,
with the wings 4-5 mm. wide and 7-8 mm. long : wings very
thick, of two lamina, semi-cordate at the apex : wingless tip of the
fruit very short.

This species is intermediate between A. elliptica and the next
species. From the former it differs in the glabrous stem and the
stricter habit, from the latter in the broader and shorter bracts
and the smaller flowers.

Colorado : Grand Junction, 1883, M. E. Jones (type in U. S.
Nat. herb.).

10. Abronia lanceolata sp. nov.

Perennial : stem glabrous, decumbent, several decimeters long:
leaves rather thick, glabrous ; petioles 2-3 cm. long ; blades ob-
long-oval, 2-5 cm. long, 1-1.5 cm. wide: peduncles 5-15 cm.
long; bracts 6-8, lanceolate, acute or acuminate, about i cm.
long: flowers numerous, about 1.5 cm. long, with a limb 4-5
mm. wide : fruit with the thin wings about 8 mm. wide and about
as long, cuneate-obpyramidal in outline, puberulent : wings
strongly reticulate, semicordate at the apex, where the tip of the
achene extends for about 2 mm.

The type was labele(;i A. fragrans, which species it resembles
in general habit, but it is easily distinguished both by the narrow
bracts and the strongly winged fruit. These characters place it
nearer A. Carletoni Coult. & Fisher which is of a different habit and
has a puberulent stem. A. lanceolata grows in drifting sand.

RvDisERG : Studies on the Rocky Mountain Flora 686

Idaho: Klalio Falls, 1901, Merrill & Wilcox, 8yo. (Type in
herb. N. Y. Hotanical Garden.)

II. Ahkonia Carletoxi Coult. &; Fisher, Bot. Gaz. 17; 349.

1892
Colorado.

12. Abronia villosa S. Wats. Am. Nat. 7: 302. 1S73
Southern Utah to Arizona and California.

13. Abronia cvcloptera A. Gray, Am. Journ. Sci. II. 15: 319.

1853
From Wyoming to Texas and Arizona.

14. Abronia pedunculata (M. E. Jones)

Abronia micrantlia pedunculata M. E. Jones, Proc. Cal. Acad.
II. 5: 716. 1895.

In the Navajo Basin of eastern Utah.

15. Abronia micrantha Torr. Frem. Rep. 96. 1845
From South Dakota to Montana and New Mexico.

2. HERMIDIUM S. Wats. King's Rep. 5: 296. 1871
A monotypic genus.

I. Hermidium alipes S. Wats. /. c.
Nevada and western Utah.

3. QUAMOCLIDION Choisy ; DC. Prod. 13': 429. 1849
This genus was based on two species, of which the second was
referred doubtfully to the genus. The first had before been known
as a species oi Mirabilis, viz., -J/, triflora Benth. The type of the
genus Mirabilis L. is J/. Jalapa L. In the latter the filaments are
united at the base, the fruit is not viscid and the corolla is salver-
shaped with a long tube and broad limb. In Quamoclidion the
filaments are distinct, the fruit viscid and the corolla from nearly
cylindrical to bell-shaped but with a small limb. In Mirabilis the
flowers are solitary and in the typical species of Quanwilidion 3—6
in the involucre ; but as the number of flowers are not of value as
a generic character I have here included a species with one-flow-
ered involucres.

687 Rydberg : Studies on the Rocky Mountain Flora

Involucre 5-6-flowered : perianth elongated funnelform. I. Q. niultiflortim .

Involucre l-flowered : perianth opencampanulate. 2. Q. laeve.

I. QuAMOGLiDiON MULTiFLORUM Torr.; Gray, Am. Journ. Sc. II.

15: 321. 1853

Oxybaphns imtltiflorwnToxx. Ann. Lye. N. Y. 2 : 237. 1828.

Nyctagiiiia Torreyana Choisy ; DC. Prod. 13- : 430. " 1849.

Mirabilis umltiflora A. Gray; Torr. Bot. Mex. Bound. 173.
1859.

Choisy, who had not seen any specimens of this species and
who believed that it had distinct bracts, referred it to Nyctaginia ;
but it is evidently congeneric with and closely related to Mirabilis
triflora Benth., the type of QuamoclidioJi. Q. nuiltiflonun ranges
from Colorado and New Mexico to Arizona and California.

2. Quamoclidion laeve (Benth.)

Oxybaplius laevis Benth. Bot. Sulph. 44. 1844.

0. glabrifoliiis var. crassifoliiis C\id\?>y \ DC. Prod. 13^: 431.
1849.

O. glabrifoliiis ToYYO-Y, Pac. R. R. Rep. 4: 131. 1857. Not
Vahl.

Mirabilis Calif oniica A. Gray; Torr. Mex. Bound. Surv. 173.
1859.

O. Calif ornicus Benth. & Hook. Gen. 3: 4. 1880.

This species has quite often been included in Oxybaphns, i. c,
Allionia, and often in Mirabilis. Professor Heimerl, in his treat-
ment of the Nyctaginiaceae in Die iiaturlichen Pflanseufaniilicn,
vcitYgcs Oxybaphus into Mirabilis ; but associates this species with
the one-flowered species of Allionia.

It is evident that if Allionia is to be treated as a distinct genus,
Q. laeve can not be included in the latter genus for it lacks the es-
sential characters, viz., the ribbed fruit and the enlarging and mem-
branous involucre. It is evidently closer related to Quamoclidion
than any other genus. The only important difference between it
and the typical species is the open short perianth and the flowers
solitary within each involucre. The species ranges from Utah
to Arizona and California.

4. Allioniella gen. nov.
Bracts five, united into a gamophyllous viscid rotate involucre,

Rvi)i;i:k(; : Studies on the Rocky Mountain Floka 088

which enlarges somewhat in fruit but does not become membra-
nous : llowers in each invohicre 3 : jjerianth open, short fuiuiel-
form : stamens 3, distinct : fruit cHipsoid, neither anj^led nor ribbed,
very indistinctly tuberclcd, <jlabrous.
Only one species.

I. Allioniella oxybaphoides (A. Cray)
Qitauioclidioii oxybaphoidis A. (jray, Am. Journ. Sc. II. 15:

320. 1853.

Mirabilis oxybaplioiihs A. Gray, Bot. Mex. Bound. Surv. 173.

1859.

OxybapJius jr/vV///// Hemsl. Biol. Cent. Am. 3: 3. 1882.

Allionia oxybaphoides Kuntzc, Rev. Gen. 533. 1891.

From the many synonyms can be seen that this species has
been moved from one genus to another. Most botanists have
regarded it as an Al/io?iia {Oxybaplnis), which it resembles most in
general habit, the perianth and the number of stamens ; but it
lacks the most essential characters of that genus, viz., the ribbed
fruit and the membranous involucre. It is, therefore, more clo.sely
related to the Mirabilis series ; and could be included in genus
Qiiamoclidion had it not the open AUionia-X\V& perianth, only three
stamens and a flat rotate involucre. It is, therefore, better to regard
it as the type of a new genus, intermediate between Allionia and
Qiiamoclidion.

A. oxybaphoides grows from southern Colorado to western
Texas and Arizona ; also in northern Mexico.

5. ALLIONIA Loefl. It. Hisp. 181. 1758

Vitmania Turra; Cav. Ic. 3 : 53. 1794- Not Vitmaunia
Vahl. 1 794.

Oxybaphus L'Her.; Willd. Sp. PI. i : 185. 1797.

Calyxhymcnia Ortega, Nov. aut Rar. PI. Hort. Matr. 5. 1797.

Calymcnia Pers. Syn. i : 36. 1805.

The genus was based on a plant which a year later received
the specific name A. violacca L. Vitmania and OxybapJius were
both based on Mirabilis viscosa Cav.; Calyxhymcnia on C. glabri-
folia and Calymcnia on six species without the type being desig-
nated.

689 Rydberg : Studies on the Rocky Mountain Flora

Key to the Speclen

Leaves from cordate to broadly ovate-lanceolate : all distinctly petioled.

Leaves cordate. I. A. nydaginea.

Leaves ovate, rounded or cuneate at the base. 2. A. floribunda.

Leaves ovate-lanceolate, oblong or linear, sessile or only the lower short- petioled.
Involucres in open terminal cymes.

Stem more or less hirsute as well as viscid.

Leaves ovate or broadly oblong, as well as the stem conspicuously hir-
sute. 3. A. hirsnia.
Leaves linear-lanceolate, almost glabrous ; stem sparingly hirsute or
glabrous except under the nodes. 4. A. pilosa.
Stem glabrous below, not hirsute, viscid-puberulent above.

Flowers solitary in the involucre on short slender pedicels ; fruit nearly

glabrous. 5. A. glabra.

Flowers 2-3 in the involucres, subsessile ; fruit decidedly pubescent.

Leaves of the cymes much reduced and bract-like, upper portion of

the stem densely and finely puberulent. 6. A. bradeata.

Leaves of the cymes neither much reduced nor bract-like.

Leaves erect or ascending ; lobes of the involucre rounded or
broadly triangular-ovate.
Plant prostrate or diffuse ; involucres and branches of the

inflorescence densely viscid hairy. 7. A. diffusa.
Plants more simple, erect or ascending ; branches of the
inflorescence usually merely viscid-puberulent.
Leaves from ovate or obovate to linear-lanceolate,
usually over 5 mm. wide. 8. A. lanceolata.

Leaves narrowly linear, less than 5 mm. wide.

9. A. linearis.
Leaves divergent ; lobes of the involucre elliptic or oval.

10. A. divaricata.
Involucres on solitary axillary peduncles, rarely also in small dense terminal clusters.
Leaves oblong, lanceolate or linear-lanceolate.
Stem hirsute.
Stem glabrous.
Leaves narrowly linear.

II.

A.

aggregafa.

12.

A.

deniinbens.

13-

A.

Bodinii.

I : 100,

■ 1803

334-

1825.

I. Allionia nyctaginea Michx. Fl. Bor. Am.
OxybapJius nyctagiiicus Sweet, Hort. Brit, i :
A well-known and common plant growing in rich soil from
Illinois and Saskatchewan to Wyoming, New Mexico and
Louisiana.*

2. Allionia floribunda (Choisy) Kuntze, Rev. Gen. 533. 1891
A. ovata Pursh, Fl. Am. Sept. i : 97. 18 14. Not Oxybaphus
ovatiis Vahl. 1806.

* Allionia Cervantesii has been reported from Colorado, but the specimens on which
this assertion has been made belong in all cases I know to Allioniella oxybaphoides.

Rydberg : Studies on the Rocky Mountain F'lora 690

O. /on'/fUfu/i/s Chois ; DC. Prod. 13': 433. 1849.

A. nyct(7gmea m'ttta Morong, Mem. Torrey Club, 5 : 146.
1894.

In dry soil from Mi.ssouri and South Dakota to Wyoming,
New Mexico and Texas.

3. Allionia hirsuta Pursh, Fl. Am. Sept. 2: 728. 18 14

Oxybaphus hirsutus Sweet, Hort. Brit, i: 334. 1825.

In sandy soil from Minnesota and South Dakota to Colorado.

4. Allionia pilosa (Nutt.)

Calymcnia pilosa Nutt. Gen. I : 26. 1818.
Oxybnp/uis pilosus SwQ&i, Hort. Brit, i : 334. 1825
This has been confused with the preceding, but I think it
amply distinct. So it was regarded by Nuttall and by Sweet. It is
perhaps nearer related to A. floribimda and Nuttall cites A. <n>ata
Pursh as a synonym ; but it is well to note that Nuttall describes
the stem of his species as pubescent, which does not agree with
Pursh's plant. A. pilosa grows in dry sandy soil from Wisconsin
and North Dakota to Texas and Louisiana.

5. Allionia glabra (S. Wats.) Kuntze, Rev. Gen. 533. 1891

Oxybaphus glaber S. Wats. Am. Nat. 7 : 301. 1873.
A rare plant from the arid districts of southern Utah, Arizona,
and New Mexico.

6. Allionia bracteata sp. nov.

A branched, more or less viscid perennial : stem glabrous
below, densely viscid-pubescent above, 4-12 dm. high, erect or
ascending : leaves subsessile ; blades oblong to linear-lanceolate
or linear, 3-9 cm. long, thick, mostly ciliate, glabrate : leaves of
the inflorescence elliptic or oblong, to lanceolate, reduced, less
than I cm. long, more or less viscid-pubescent : involucres often
numerous in terminal cymes, 10-14 mm. wide; lobes broadly
ovate, obtuse : perianth white or pale, about 10 mm. broad : fruit
oblong-obovate, 5-5.5 mm. long, prominently 8-ribbed, apiculate,
constricted near the base.

This species has been mistaken for A. linearis and A. albida,
but is easy distinguished by the reduced leaves of its inflorescence.
It grows in dry rocky soil.

691 Rydberg : Studies on the Rocky Mountain Flora

Missouri : Maiden, 1894, Btish, 759 (type in herb. Columbia
University); Wayne, 1900, 82^ ; also Courtney, 1891 ; McDonald
county, 1893, j/o.

South Dakota: Spring Basin, 1891, T. A. Williams.

Alabama: Selma, 1888, 7l/(rGr;Y//j'.

7. Allionia diffusa Heller, Minn. Bot. Stud. 2 : 33. 1898

On dry plains, from North Dakota and Wyoming, to Kansas,
New Mexico and Arizona.

8. Allionia lanceolata sp. nov.

A. albida Rydb. Cont. U. S. Nat. Herb. 3: 520. 1896. Not
Walt. 1788.

A branched perennial. Stems erect or ascending, glabrous
below, viscid-pubescent with short hairs, 4-15 dm. high: lower
leaves short-petioled, the upper sessile ; blades lanceolate or ovate-
lanceolate to almost linear, 3— 10 cm. long, very thick, obtuse or
blunt at the apex : involucres numerous, in terminal cymes, i — 1.5
cm. wide ; lobes rounded ovate, sometimes acutish : perianth pink,
about 10 mm. broad : fruit obovoid, 4.5—5 mm. long, with usually
4-5 broad ribs and finely tuberculate faces.

This has usually gone under the name of Allionia albida which it
resembles in habit, but that species has white perianths and sharply
acute or acuminate thin leaves. A. albida is confined to South
Carolina and Georgia and is represented by A. lanceolata in the
West. The latter grows in dry soil on the plains, from Minnesota
and Wyoming to Tennessee and Texas.

Colorado : Estes Park, Larirher county, Qstcrhoiit, ijj6 (type
in herb. N. Y. Botanical Garden).

9. Allionia linearis Pursh, Fl. Am. Sept. 2 : 728. 18 14

Calymenia angiistifolia Nutt. Gen. i : 26. 18 18.
Oxybaphns angusiifolins Sweet, Hort. Brit, i : 334. 1826.
In dry soil on the plains, from Minnesota to Montana, Arizona,
Mexico and Louisiana.

10. Allionia divaricata sp. nov.

A slender perennial. Stems usually solitary, erect, 6-10 dm.
high, glabrous and shining up to the viscid-puberulent inflores-
cence : leaves more or less distinctly petioled, usually spreading ;

RvmjERo : Studies on tiis Rockv Mountain Fi.oka Gt>2

blades c^Iabrous, thickisli, linear to lincar-lanccolatc, acute, 5-10
cm. loii^ : involucres numerous, in terminal C}'mes about 1.5 cm.
wide, cleft below the middle into elliptical or oval obtuse lobes :
perianth pink, about 8 mm. wide : fruit oblong-obovoid, sli<^htly
compressed, bluntly an^^led and not strongly tubercled, striyo.se.

Perhaps closest related to ./. linearis, but distinguished by
the thinner divergent leaves, the form of the involucfe and the
fruit.

Colorado: Durango, 1898, Baker, Earlc c'V- Tian\ ^12 (type
ill herb. N. V. Bot. Garden).

Arizona: Bakers l^uttc, Mogollon Mountains, 1887, Mcanis,

-53

11. Allionia aggregata (Ortega) Spreng. Syst. i: 384. 1825
CalyxliyiJienia aggregata Ortega, Nov. aut Rar. PI. 8 : //. / / .

1798 (or 1799?)-

Oxybaplms aggregatus Vahl, Enum. 2 : 41, in part. 1806.

This is not the plant named 0. aggregatus by Torrey, Watson
and others, from Arizona and northern Mexico, but one that is
closely related to A. Jiirsuta and generally has been confused with
it. It differs mainly in the axillary solitary peduncles and in this
respect is analogous to A. Bodinii. Vahl evidently had two plants
confused, citing as synonyms Calyxliynienia aggregata Ortega and
Mirabilis aggregata Cavanilles, both illustrated and closely related
species of Allionia. Ortega's plant, which was the first published,*
is densely hirsute, while Cavanilles' plant is glabrous. Vahl, in
his diagnosis, describes it as glabrous but remarks in parentheses
" according to Ortega hirsute."

A. aggregata grows in dry soil from Wisconsin to Texas and
New Mexico, also in northern Mexico.

12. Allionia decumbens (Nutt.) Spreng. Syst. i: 384. 1825

Mirabilis aggregata Cav. Ic. 5 : 22. //. ^^y. 1799.
Oxybaplms aggregatus Vahl, /. c. in part.
Calyuienia decumbeus Nutt. Gen. i : 26. 18 18.
Oxybap/ius decumbens SvjQCt, Hort. Brit, i: 334. 1826.

* I have not been able to find the exact date of Ortega's fascicle in which the de-
scription appears. The ten f.iscicles were published in 1 797-1800 ; but Cavanilles cites
Ortega, which indicates that the latter' s description was published first.

693 Rydberg : Studies on the Rocky Mountain FijORA

This species stands in the same relation to A. lanccolata as the
preceding does to A. Jiirsiita and the following to A. linearis, dif-
fering mostly in the mode of blooming. One of the reasons for
holding them as distinct species and not as forms of the species
mentioned is that their ranges are different, A. aggregata and A,
decumbcns extending further south into Mexico and the Gulf states^
while their cymose analogues are northern, and the range of ^.
Bodinii is more limited than that of A. linearis. A. decii7nbens
grows in dry soil on the plains, from Missouri and South Dakota
to Wyoming, Mexico and Mississippi.

13. Allionia Bodinii (Holz.) Morong, Mem. Torrey Club,''5 :

354- 1894
Oxybaphus Bodinii HoXz. Contr. U. S. Nat. Herb, i: 287. 1893.
In dry soil from western Kansas to Utah and western Texas.

6. WEDELIA Loeff. Iter. Hisp. 180. 1758

Allionia L. Syst., Ed. 10, 890, in part. 1759.
The genus was based on a plant which the following year
received the specific name Allionia incarnata L.

I. Wedelia incarnata (L.) Kuntze, Rev. Gen. PI. 533. 1892

Allionia incainiata L. Syst., Ed. 10, 890. 1759.

From western Texas and southern Colorado to California.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 31.

A FOSSIL PETAL AND A FOSSIL FRUIT FROM THE
CRETACEOUS (DAKOTA GROUP) OF KANSAS

By ARTHUR HOLLICK

NEW YORK
1903

[Reprinted from the Bulletin of the Torrey Botanical Club, 30: 102-105. 28 F 1903]

A fossil Petal and a fossil Fruit from the Cretaceous
(Dakota Group) of Kansas

Bv Arthik Hollick

Included in a collection of fossil plant remains from the Cre-
taceous (Dakota Group) of Kansas, recently obtained by the New
York Botanical Garden from Mr. Charles II. Sternberg, of Law-
rence, Kans., are two exceedingly interesting specimens — one rep-
resenting a large petal, the other a fleshy fruit.

Petals, as fossils, are exceedingly rare, and I am not acquainted
with any published figure of anything of the kind which can com-
pare with ours, in regard to either size or satisfactory condition of
preservation. Unfortunately, a portion of the upper part, includ-
ing the apex, is gone, but it is sufficiently perfect to indicate what
was its original shape, and the principal characters of the nervation
are plainly discernible.

Careful examination and comparison has shown that, in all
essentials, it agrees with the petals of some of our large-flowered
magnolias, such as M.foetida Sarg. and M. juacrophylla Michx.,
and as some ten species of Magnolia, founded upon more or less
well-defined leaves, have been described from the Dakota group, I
have decided to refer it to that genus.

The probabilities, of course, are that it belongs to a species to
which one of the fossil leaves belongs, but as it is impossible defi-
nitely to connect the petal with any one of these a distinctive
name is necessary, which should indicate that the fossil was a
petal and not a leaf.

Magnolia palaeopetala sp. nov.

Petal apparently ovate-spatulate in outline, about i 5 cm. long
by 10 cm. broad, convex, laterally constricted and incurved at the
base ; margin entire, wavy or flexuous ; nervation flabellate, dic-
tyodrome, simple and well defined below, forking and thinner
above, anastomosing, the areolae and reticulations becoming suc-
cessively smaller and the nervilles finer, towards the margin.
{Fig. A.)

102

103 HoLLiCK : Fossil Petal and Fruit from Kansas

Locality : Ellsworth Co., Kans.

It is apparent that this petal must have possessed a texture
more or less coriaceous, as otherwise its preservation in such a
relatively coarse medium as the Dakota sandstone would have
been impossible, and this supposition is supported by the fact that
the convexity of the surface, the incurving of the base and margin
and the constriction of the former, all appear to have been char-
acters of the living flower and not to be due to subsequent distor-
tion by fossilization.

Fig. a. Magnolia palaeopetala Hollick.

HoLLicK : Fossil Petal and Fklit from Kansas 104

The fruit is plainly' that of a fig, and although some twenty-
three species oi luciis have been described from the Dakota group
they were based upon leaf-impressions alone. Only two speci-
mens of fruit were known and these were too ill-defined to admit
of specific description. (" Fruits o{ Ficiis," Lesq. Fl. Dak. Group,
85. //. 10. f. 7. 8. See our Fig. B, 2, 3.) They are incidentally
mentioned however as being associated with leaves of Ficiis iuae-
qualis Lesq.

Specimens considerably more satisfactory were described and
figured by Heer in 1874, from the lower Atane beds of Green-
land— representing a geological horizon which is practically the
equivalent of the Dakota group — under the name Ficus protogaea
Heer* (H. Foss. Arct, 3-: 108. pi. jo. f. 5-7. See our Fig. B,
4. 5. 6.)

4 56

Fig. B. I. Fiais neurocarpa Hollick. 2, 3. " Fruits of /jVwj " (after Lesque-
reux). 4-6. Ficus protogaea Heer (after Heer).

*The name Ficus protogaea was previously used by Ettingshausen, in 1867, in de-
scribing a fragmentary fossil leaf from the Cretaceous of Niederschoena in Saxony
(Sitzb. Akad. Wiss. Wien, Math. -Nat. CI. 55 : 249. //. 2. f. j. ) In any revision of
the genus it would therefore be necessary to rename these fruits. In this paper how-
ever it is not thought necessary to do anything more than to call attention to the fact.

105 HoLLicK : Fossil Petal and Fruit from Kansas

These five specimens are the only ones which I have found
recorded as occurring in the same geological horizon as that to
which our specimen belongs, or in any strata which may be re-
garded as its equivalent, and none of them is equal to ours as a
specimen, so far as may be judged from the figures.

Ficus neurocarpa sp. nov.

Fruit broadly obovate-spatulate in outline, 2.8 cm. long by
2.4 cm. maximum width, thick and wrinkled towards the middle,
flattened more or less at or near the margin, minutely punctate and
delicately nerved ; primary nerves convergent but distinct and sepa-
rated from each other at the base, divergent and becoming thinner
and indistinct above, where they anastomose ; secondary nerves
very fine and apparently irregularly disposed. {Fig. B, /.)

Locality : Ellsworth Co., Kans.

This fossil has very much the appearance of many dried her-
barium specimens and it is evident that it must have possessed
considerable consistency in order to retain its original shape, as it
has done to a certain extent, under the pressure to which it must
have been subjected.

I

I

i

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 32.

THE POLYPORACEAE OF NORTH AMERICA
II. THE GENUS PYROPOLYPORUS

By WILLIAM ALPHONSO MURRILL

NEW YORK
1903

[Keprinted from the Bulletin of the Torrky Botanical Club, 30: 109-120. 28 F 1903]

The Polyporaceae of North America. II. The Genus Pyropolyporus*

15y William Aliiionso Murrill

The luiropean species of this genus were first separated into a
distinct generic group by Quelet in his " Enchiridion Fungorum "
published in 1886. His genus PhcUinus established at that time
contained four species, P. igniarius (L.), P. fitlvus (Scop.), P. con-
chatus (Pers.) and /'. saliciuus (Pers.), and was characterized as fol-
lows : " Pileus velvety, persisting ; context corky ; pores small,
fulvous brown ; spores ovoid, fulvous. Plants lignatile." The
name PhcUinus, however, is preoccupied by Phelline assigned in
1826 to a genus of the Ebenaceae. The new name Pyropolyporus
here proposed refers to the use of some species of this group in
ancient times for the purpose of keeping fire.

Synopsis of the ^ortli American Species

1. Pileus thick, ungulate, woody, margin obtuse, tubes several times stratified, the

upper layers not dying before the lower. 2.

Not as above. ll-

2. Context yellowish brown. 3-
Context reddish orange ; plants growing on trunks o{ Juniperus. lO.

3. Spores hyaline. 4-
Spores yellowish brown. 7*

4. Pileus becoming more or less rimose with age. 5*
Pileus covered even in age with a smooth horny crust. 6.

5. Pileus simple, sulcate, sometimes polished, margin usually narrow and rounded ;

not found on species oi Prunus. I. P- igniarius.

Pileus terraced, imbricate or semiresupinate, rarely sulcate, never polished, margin

broad, making an obtuse angle ; found on species of Prttnus. 2. P. fulvus.

6. Context ferruginous, surface dark brown marked with narrow black concentric lines,

tubes thin-walled, 5 to a mm. 3- P- crtislosus,

Conte:a tawny, surface black without lines and more sulcate, tubes smaller, thick-
walled, 7 to a mm. 4- •^- Calkimii.

7. Pileus soon becoming rimose. °'
Pileus not rimose, broadly sulcate, zonate, tubes thin-walled, spores 3/^ in di-
ameter, spines large and abundant ; growing on Quercus. 5. P. Everhartit.

* Read by invitation before the Botanical Society of America at Washington, Jan-
uary I, 1903.

109

110 MURRILL : POLYPORACEAE OF NoRTH AMERICA

8. Tubes long, over 0.5 cm, each season, walls thin, pores large, 3 to a mm. 9.
Tubes very short, 0.1-0.5 cm. long each season, walls equalling tubes in thick-
ness, pores small, 5 to a mm., spores globose, 4-5 Z^, cystidia none; growing
abundantly on Robinia. 6. P. Robiyiiae.

9. Spores globose, 3-4// in diameter, cystidia present ; pileus blackish, very rimose,

hymenium convex, margin truncate, at length hoary. 7. P. praerimosus.

Spores ellipsoidal, 7 X 9i«) cystidia none ; pileus dark brown, hymenium concave,

margin not hoary. 8. P. Underwoodii.

10. Older pores visible in projecting annual layers, tubes 3-4 to a mm., thin-walled;

pileus deeply furrowed, not rimose. 9. P. juniperinus.

Older pores not externally visible, tubes 1-2 to a mm., thicker walled ; surface

very rimose. 10. P. Earlei.

11. Spores hyaline. 12.
Spores yellowish brown. 15-

12. Cystidia abundant, pointed, dark brown; pileus thin, woody, rigid, tubes short,

5 to a mm. II. P- conchatui.

Cystidia none. 13-

13. Context thick, woody, honey-yellow, surface encrusted, margin blunt.

12. P. Haematoxyli.
Context thin, punky, darker brown, surface without a crust, margin sharp. I4.

14. Pileus 10-25 c'l^- broad, covered with narrow shallow furrows, margin undulate

or lobed, pores minute, 8-9 to a mm. 13. P. Langloisii.

Pileus smaller, deeply sulcate, pores larger, 6 to a mm. ; growing on species of

Ribes, very rarely on other shrubs. 14- P- Ribis.

15. Dissepiments thick, hymenium concolorous with context. 16.
Dissepiments thin, hymenium darker than context. . 17.

16. Pileus very rimose, cystidia present ; tubes fulvous, 7 to a mm., spores globose,

3.5-5 II, cystidia 17-35/^. 15- P- Yticataneusis.

Pileus scarcely rimose, encrusted, becoming smooth, cystidia absent ; margin

undulate, tubes 8 to a mm., spores globose. 16. P. sefiex.

17. Pileus hard and heavy, surface sulcate, brown, clothed with lime-white hair,

spores 3.5//, spines 40-60 /z. 17. P. Hnteus.

Upper layers of pileus dead, projecting 1-2 cm. beyond the lower, tube strata

separated by thin layers of context, tubes 7 to a mm., spores 5-7 ,«, cystidia

absent. 18. P. Jamaicensis.

I. Pyropolyporus igniarius (L.)

Boletus igniarius L. Sp. PI. 1176. 1753; Tourn. Inst, i:
562. pi. 330./. A. 1719.

Polyporus igniarius Fries, Syst. i : 375. 1821.

Fomes igniarius Gill. Champ. Fr. i : 687. 1878 ; //. 156.

Phellinus iccniarius Quel. Ench. 172. 1886.

MUKRILL : POLVPORACEAE OF NORTH AMERICA 111

This plant is one of the oldest fungi known, having been used
from very early times for keeping fire, whence the name. It is
also abundant and widely distributed, occurring on birch, willow,
beech, maple, oak, poplar, apple and many other deciduous trees
in various parts of the world. F'ruit-bodies observed on apple,
willow, aspen, birch, etc., in Sweden, agree perfectly with those
growing in the New York forests ; while the variations found in
Europe arc duplicated here, even to the abundance or scarcity of
spines in the hymenium.

Specimens are to be found in all published exsiccati bearing
upon this group. Among the large number of collections exam-
ined, the following may be mentioned : New York (Underwood),
New Hampshire (Miss Minns), Ottawa, Canada (Macoun), New
Jersey (Ellis), Virginia (Murrill), Kansas (Bartholomew), Indiana
(Underwood), Colorado (Earle), Labrador (Waghorne), Maine
(Ricker), New Mexico (Earle), Finland (Karsten), Sweden (Murrill),
Thiiringen (Underwood).

Two forms of P. igniarms are listed as distinct species in Sac-
cardo's Sylloge, /. c. Fames badius Berk, and Fomes nigricans
Fries. The former was collected in Arctic North America by Dr.
Richardson and the type is at Kew. The specimen is 5 X 6 X
2.5 cm. and has three layers of tubes which, in the brief northern
season, were forced to develop quickly and are therefore thinner-
walled than in our forms. Berkeley himself doubted if the plants
were sufficiently distinct from P. igniarius.

Fames nigricans Fries, on the other hand, is decidedly different
from typical P. igniarius, being neat and smooth and shining black,
and rimose in two directions when old. Fries observed it on birch
only. Persoon noticed it also frequently on old willows, and
Underwood has collected fine specimens of the same form on
beech in the New York mountains. While studying living plants
on willov/ trees in Sweden, I found stages on the same tree con-
necting this variety with the typical form and am convinced that
we are not here dealing with two distinct plants but with varia-
tions due to season, food supply, rapidity of growth, or some other
physiological cause yet to be determined. Specimens have been
examined from F"inland (Karsten, Starback), Sweden (Romell,
Murrill), Newfoundland (Waghorne), Greenland (Rostrup), New

112 MURRILL : POLYPORACEAE OF NoRTH AMERICA

Hampshire (Miss Minns) and New York (Underwood). Klotsch's
collections from Scotland are Poly poms foincntarius L. of advanced
age and indurated. This is the conception of F. 7ugricans at
Berlin (see Sydow, Myc. Mar. No. 2604). At Kew, F. 7iigricans
has been considered " rare the world over." A tiny specimen
2X3X2 cm. in the herbarium of Berkeley sent from Upsala
presumably by Fries himself has the external appearance of F.
nigricans (see Fries, Icon. />/. 1S4), but its size would indicate that
it is Povia levigata Fries, in its pileate form, a state of this fungus
not known to Fries and easily confused with depauperate forms of
F. nigricans. The other specimens at Kew are all old hardened
forms of P. fomcntariiis L.

2. Pyropolyporus fulvus (Scop.)

Boletus fulvns Sco]p. Fl. Carn. Ed. 2. 2: 469. 1772.

Polyporiis igniarius b. minor, subperpendicularis Fries, Syst. I :
375. 1821.

Polyporiis fnlvus Fries, Hym. Eur. 559; Icon. pi. 184, f. j.

Fames fulvus Gill. Champ. Fr. i : 687. 1878.

Phellinus fulvus Quel. Ench. Fung. 172. 1886.

Fries at first considered Boletus poniaceus Pers. synonymous
with B. fulvus Scop., but later decided, and correctly so, that it is
only a form of P. igniarius L. In Micheli's no. 9, p. 119, of his
genus Agaricuni, the two forms are confused, though his figure,
pi. 61, ordo II, doubtless represents P. fulvus growing on Primus
and not P. igniarius growing on Mains. This confusion still exists
in European herbaria and it is impossible to determine in them
what is meant by Poly poms fnlvus Fries. In the following descrip-
tion Scopoli very plainly characterizes the plant under discussion :
" Totus fulvus ; poris difformibus, non adeo profundis. Habitat
in cerasis recens detruncatis. Junior integerrimus, adultus vero
margine sinuatus. Fulvus color in margine et subtus evidentior.
Poruli obsoleti, totum fungum subtus non semper occupant."

PI. Crypt. France par Desmazieres no. 5157. " Sur les pru-
niers et les cerisiers." Fung. Selec. Exsic. par Roumeguere No.
6963. " Forma Cerasi, sur un vieux cerisier." Trienti (Bresa-
dola) on Persica, Eisenach and Thuringen (Underwood) on Pru-
nus, Sweden (Murrill) on Primus, and on Pmnus both wild and

MURKILL : POLYPORACEAE OF NORTH AMERICA 113

cultivated, in America in Alabama (Underwood), Colorado (Bethel,
Cameron, Crandall), Kansas (Bartholomew. Cragin), Missouri (?),
Nebraska (Webber), Ohio (Lloyd), Tennessee (Schrenk).

This fungus is very uniform in habit ancF appearance both in
Europe and America, where it occurs on various species of Pritnus
and its near allies. Although confused in literature with Poly-
porus igtiiarius L., it is very distinct and always easily recognized.
In an orchard near Mauritzberg, Sweden, where /'. igniarius was
abundant on apple trees, P. fulviis was confined to the stumps and
dead or dying trunks and branches of plum trees.

3. Pyropolyporus crustosus sp. nov.

An ungulate plant of medium size with brown tubes, ferrugi-
nous substance and smooth encrusted dark brown surface. Pileus
woody, convex above, plane below, somewhat compressed ungu-
late, 7X 12X5 cm.; surface glabrous, horny encrusted, dark brown,
concentrically sulcate, marked with narrow black concentric lines ;
margin rounded, yellowish brown, sterile : context hard, concen-
trically banded, ferruginous, 2 cm. thick behind ; tubes indistinctly
stratified, 0.5-1 cm. long each season, 5-6 to a mm., drab-colored
within, mouths polygonal, concolorous, edges thin, acute, entire :
spores globose to ovoid, smooth, thin-walled, hyaline, 3.5-4X4/^,
hyphae darker, cystidia none.

Collected by Earle on a standing tree trunk at an altitude of
4,000 feet on Rose Hill, Jamaica, October, 1902.

4. Pyropolyporus Calkinsii sp. nov.

A large ungulate fungus glabrous and furrowed above and
uniformly hard and fulvous within. Pileus very hard woody
throughout, ungulate, 10 X 10 X 10 cm. ; surface glabrous, dark
brown to black, marked with rather shallow concentric furrows,
crust thin, horny, never rimose ; margin rounded, concolorous
with the h\-menium : context very hard woody, fulvous, i cm.
thick ; tubes in many indistinct layers, slender, minute, 7 to a
mm., fulvous, mouths nearly circular, obtuse, entire : spores
ovoid, hyaline, with thick, smooth, pale ferruginous wall, 3-5 X
5-7 11., hyphae ferruginous, cystidia none.

This species occurs on living trees of live-oak in Florida, where
it was collected in'considerable quantity by Major W. W. Calkins
during the winters of 1886 and 1887. Several specimens are in
the herbarium of the New York Botanical Garden and a single

114 MURRILL : POLYPORACEAE OF NoRTH AMERICA

specimen in the herbarium of the Division of Vegetable Pathology
and Physiology of the U. S. Department of Agriculture.

5. Pyropolyporus Everhartii (Ell. & Gall.)

Mucronoporus EvcrJuvtii Ell. & Gall. Journ. Myc. 5 : 141-142.
pi. 12. 1889.

XaiithocJirous Everhartii Y2X. Cat. Tun. 51. 1897.

The type of this fungus is in the herbarium of the New York
Botanical Garden. Since its discovery on the living trunk of a
scarlet oak in New Jersey, it has been collected in several localities
on various species of oak and occasionally on beech. Before its
separation as a distinct species it had been confused with P. ignia-
rius, which it outwardly resembles in some particulars.

Collections : New York (Ellis, Mrs. Ellis, Murrill), New Jer-
sey (Ellis, Ely), Canada (Dearness), Indiana (Gentry), Pennsylvania
(Herbst) and Delaware (Commons, Ellis N. A. Fungi no. 3303).

6. Pyropolyporus Robiniae sp. nov.

A large fungus with dark rimose surface and tawny hymenium
very common on Robinia psaidacacia. Pileus hard woody, dimid-
iate, ungulate to applanate, 5-25 X 5-50 X 2-12 cm.; surface
velvety, smooth, soon becoming very rimose and roughened,
fulvous to purplish-black, at length dull black, deeply and broadly
concentrically sulcate ; margin rounded, velvety, fulvous : context
hard woody, concentrically banded, 1-3 cm. thick, fulvous ; tubes
stratose, 0.15-0.5 cm. long, 5 to a mm., fulvous, mouths sub-
circular, dissepiments entire, equalling tubes in thickness : spores
subglobose, smooth, thin-walled, ferruginous, copious, 4—5 fi, cys-
tidia none.

This fungus was one of the first to be noticed by collectors in
this country, but has been unnamed until the present time.
Schweinitz called it Polyporiis igiiiarius and remarked that it was
"frequent especially on Robinia'''; Berkeley confused it with his
P. rinwsus described from Demerara and the Cape of Good Hope,
and Cooke allayed Morgan's anxiety by assigning it most positively
to the same category. To be sure, it resembles P. rifnosus Berk,
from Demerara, but the two plants are entirely distinct in appear-
ance and shape and/*, runosus lacks the decided imbricated-rimose
effect so characteristic of our plant. It was from the African
plant, a different thing from P. riniosiis, that the name riviosjis must

MURRILL : POLYPORACEAE OF NORTII AMERICA 115

have been obtained. It is truly rimosc like /v. Rolnninc, but dif-
fers decidedly from the latter in pore-structure and context, while
the spores are larger and less globose. It would be quite remark-
able if a group of plants from such widely separated localities did
not show specific differences.

P. Rolnniae is abundant in the southern United States on
Robiiiia and extends with it as far north as Connecticut and west
to Missouri and Texas. I have not as yet seen it upon any other
host, but a plant recently collected by Earle in Jamaica on Acacia
cviarginata, nearly related to Robinia, seems not specifically dis-
tinct from it. When Robifiia was introduced into France several
centuries ago from Virginia this fungus must have been introduced
with it, since it was collected there by F. Fautrey in November,
1 89 1, growing on Robinia. The specimen is at Upsala and is
labelled P. igniarius. Among the many collections examined
from America, the following may be mentioned : Massachusetts
(Underwood) ; New York (Underwood, Earle, Murrill) ; New
Jersey (Ellis); Ohio (James, Morgan, Lloyd); West Virginia (Nut-
tall) ; Virginia (Miss V. W. Murrill); Alabama (Underwood, Atkin-
son, Earle).

A good account of the destructive effects of this plant is given
by H. von Schrenk in Ann. Rep. Mo. Bot. Gard. 12 : 21-31. pL
1-3. 1 90 1.

7. Pyropolyporus praerimosus sp. nov.

A large ungulate plant with plane brown hymenium and a very
rimose blackish surface. Pileus woody, rounded ungulate, 8-12
X 7-10 X 8-1 1 cm.; surface exceedingly rimose after the first
year, broadly furrowed, the projecting ridges splitting away in
age, very dark brown to black ; margin obtuse, velvety, rusty to
hoary : context corky to woody, concentrically banded, fulvous,
0.5 cm. or less thick ; tubes indistinctly stratified, 1-2 cm. long
each season, 3 to a mm., concolorous within and without with the
context, mouths rounded to polygonal, ochraceous at first, edges
rather thick, obtuse, becoming thin and often splitting in age :
spores globose, smooth, deep ferruginous, 3-4 fx ; spines ferru-
ginous, 10-17 X 5-10//, largest at the base.

Collected by Earle, July, 1900, on Qiierciis undidata in the El
Capitan Mountains, New Mexico, at an altitude of 7,000 feet.
Related to P. Everhartii (Ell. & Gall.).

116 MURRILL : POLYPORACEAE OF NORTH AMERICA

8. Pyropolyporus Underwoodii sp. nov.

A blackish ungulate plant of large size with furrowed rimose
surface and long brown tubes. Pileus woody, broadly ungulate,
attached by a narrow base, concave below, 7 X 14 X 1 1 cm.; sur-
face many times concentrically furrowed, rimose, uniformly dark-
brown to black ; margin fulvous, acute or somewhat obtuse, vel-
vety, undulate, marked with narrow zones : context hard, fulvous
to dark brown, very thin, less than 0.5 cm.; tubes distinctly strati-
fied, 0.5-1.5 cm. long each season, 3 to a mm., brown within,
mouths darker, circular or polygonal, edges acute, entire : spores
ellipsoidal, smooth, thin-walled, Hght yellowish-brown, 7X9-".
cystidia none.

Collected by Underwood and Griggs during the summer of
1 90 1 near Coamo Springs, Porto Rico. The plant reminds one
of P. Rohmiae, but differs very widely in tube-structure, as well as
in color, density and spore characters.

9. Pyropolyporus juniperinus (Schrenk)
Polyporiis juniperimis Schrenk, U. S. Dept. Agr. Veg. Physiol.

Bull. 21: 9-16./'/. 1-4.. 1900.

So far as I know, there are only two specimens of the fruiting
stage of this plant in existence, one collected by Schrenk in Ten-
nessee and the other by Miss Sadie F. Price in Kentucky. The
latter specimen, the better developed of the two, was sent to the
Underwood herbarium in 1895.

Both specimens were found upon Jiuiipenis Virginiana, the
wood of which is badly affected by its mycelium. The statement
made by the author that the discovery of other specimens might
prove this species to be only a form of P. fomcntarms is entirely
unnecessary, since such connection between two plants so distinct
is beyond the range of possibility.

10. Pyropolyporus Earlei sp. nov.

A broadly ungulate plant with yellow pores, red context and
a dark very rimose surface. Pileus woody, attached by a broad
base, plane below, 6 X 13 X 17 cm.; surface concentrically
sulcate, very rimose in older parts, fulvous to brownish-black,
at length grayish-black from weathering ; margin broad, obtuse,
dark yellowish-orange, clothed with short dense tomentum of the
same color : context woody, dark reddish-orange, concentrically
banded with darker lines, very thin, 0.5 cm., rimose down to the

MURRILL : POLVPOKACEAE OF NoRTH AMERICA 117

tubes; tubes unevenly stratified, 0.5-0.75 cm. long each season,
1-2 to a mm., ochraccous within durint^ the first season, afterwards
latericeous, mouths circular, ochraccous, edges obtuse, rather thin :
spores ellipsoidal, smooth, thick-walled, ferruginous, 7-8 x 9/'.

Collected by Karle in August, 1900, on a standing trunk of
Junifcnis in the El Capitan Mountains, New Mexico, at an attitude
of 7,000 ft. It is closely related to /'. jiinif^cnnns (Schrenk), but
it is much more rimose. has larger pores and lacks the annual
projecting margins of the older tube-layers so characteristic of
that species.

11. Pyropolyporus conchatus (Pers.)

Bo'ctiis salicinns Pers.; Gmel. S\st. 2: 1437. J79> ; S)'n.
543. 1801. Not /). salicuws Hull. Herb. I-^r. //. 4^3./. /. 1789.

Boletus concJiatus Pers. Obs. i: 24. 1796; Syn. 538. iSoi.

Polyporiis salicinns Fries, Syst. Myc. i : 376. 182 i ; Icon. //.
183./. 2.

Polyporns civichatns Fries, Syst. Myc. i : 376. 1821.

FoDhs saliciinis Gill. Champ. Fr. i : 684. 1878 ; Karst. Icon.

4,/ 5- 1883.

Ph : 11 inns sa 'ici?ins Quel. Ench. 172. 1886.

Boletus salicinns Pers. and Indctus conchatus Pers. were differ-
ent forms of the same plant. Most of the old willow stumps in
Sweden are covered with it. On the sides of the stumps it is B.
salicinns and farther up, where the pileus is reflexed, it is B. con-
chatus. This fungus is quite common in Europe and America on
a variety of hosts. The following list includes only a few of the
collections examined : Sydow, Myc. Mar. no. 3423 ; P^llis, N. A.
Fungi no. 918; Romell, P'ungi Suecici no. 12; pjigland (Plow-
right), Sweden (Murrill), New York (Underwood, Ellis, Shear),
Indiana (Underwood), Canada (Macoun), Ohio (Lloyd, Morgan),
Pennsylvania (Rau).

12. Pyropolyporus Haematoxyli sp. no\'.

A smooth applanate plant of considerable size with brownish
tubes and honey-yellow context. Pileus woody, dimidiate, sessile,
thickest behind, 12 X 14 X 4 cm.; surface glabrous, dark brown,
shallowly concentrically sulcate, marked with numerous darker
concentric lines; margin fulvous, thin, rounded, slight!}' undu-
late: context corky to woody, indistinctly concentrically banded.

118 MURRILL : POLYPORACEAE OF NORTH AMERICA

honey-yellow, i cm. thick ; tubes distinctly stratified, longer
behind, 0.5-1 cm. long each season, 6 to a mm., dull brown,
mouths polygonal, concolorous, edges obtuse, becoming thin ;
spores globose, rarely ovoid, thin-walled, smooth, hyaline, 3.5-5
//, hyphae ferruginous, cystidia none.

Collected by Earle at the base of a living logwood tree in
Paradise, Jamaica, November, 1902.

13. Pyropolyporus Langloisii sp. nov.

A large thin expanded fungus with brown hymenium and a
brown or blackish surface. Pileus corky, fan-shaped, attached by
a narrow base, often depressed behind, 8-13 X 10-25 X 0.3-1.5
cm.; surface at first anoderm, soft, clothed with brown tomentum,
many times concentrically sulcate, at length glabrous, rough, in-
durate, black, marked with numerous shallow furrows ; margin
velvety, brown, thin, acute, undulate or slightly lobed : context
soft to corky, indurate in age, deep brown, 0.2-0.5 cm. thick;
tubes reviving, distinctly stratified, 0.2-0.5 cm. long each season,
9-8 to a mm., brown, mouths polygonal, concolorous, edges thin
at maturity : spores globose, smooth, hyaline, 3 fx, cystidia none.

Collected in quantity several seasons in Louisiana by Langlois,
exclusively on dead or dying hawthorn trees near the base of the
trunk. Referred to P. pectmatus Kl., P. senex N. & M., P. con-
chatus Pers., and P. Ribis (Schum.) Fries by European mycologists,
from all of which it is quite distinct, probably approaching nearest
to the last. Several specimens from the various collections made
by Langlois are in the Underwood herbarium and the herbarium
of the New York Botanical Garden.

14. Pyropolyporus Ribis (Schum.)

Bolettis Ribis Schum. Enum. PI. Saell. 2: 336. 1803.

Poly poms Ribis Fries, Syst. i: 375. 1821.

Polyponis ribesius Vers. Myc. Eur. 2: 80. 1825.

Po/jporus Ribis Fries ; Sturm, Deuts. Fl. 3: 137. //. 62. 1837.

Pomes Ribis Gill. Champ. Fr. i : 685. 1878.

According to some authorities P. Lonicerae Weinm. and P.
Euoiiymi Kalchb. are only forms of the above. Rabenhorst-
Winter, Fungi Europaei, no. 2937 ; Roumeguere, Fungi Gallic!,
no. 3304; De Thiimen, Myc. Univ., no. 509; Linhart, Fungi

MURRILL : POLYPORACEAE OF NORTII AMERICA 119

Hung., no. 349 ; Krieger, Fungi Sax., no. 423 ; New York (Flllis) ;
New Jersey (Gcissman, Ell. &. Ev. N. A. Fungi no. 1693);
Kansas (Bartholomew).

The usual host of this fungus is the red currant, but it is also
found on other species of Ridis and occasionally upon other shrubs
growing near by. Specimens from Bartholomew were collected
on living stems of Syuiplioricarpos occidcntalis, July 24, 1895 ; and
it has also been reported as attacking rose-bushes.

15. Pyropolyporus Yucatanensis sp. nov.

A large blackish brown very rimose fungus with tawny tubes
and substance. Pileus woody, dimidiate, applanate, thickest be-
hind, 7-9 X 9-12 X 2-3 cm.; surface clothed at first with tawny
tomentum, becoming very dark brown or black and uniformly
tubercular and broken into small areas by numerous shallow con-
centric furrows and radial cracks ; margin narrow, acute, velvety,
fulvous: context hard, fulvous, 0.5 cm. thick; tubes rather
indistinctly stratified, 0.25 cm. long each season, 7 to a mm.,
fulvous ; mouths circular, punctate, edges thick, obtuse : spores
globose or subglobose, smooth, pale yellowish brown, 3. 5-5 /A
hyphae ferruginous, cystidia thick at the base, pointed, 17-35 fi.

Collected in Yucatan by Millspaugh and in Nicaragua by C.
L. Smith.

16. Pyropolyporus senex (Nees & Mont.)
Polyponis senex Nees & Mont. Ann. Sci. Nat. II. 5: 70-71.

1836.

Fonies senex Cooke, Grevillea, 13 : 118. 1885.

A number of specimens collected by C. L. Smith in Mexico
and Nicaragua are in the New York Botanical Garden herbarium.
Dr. Patouillard says they are typical P. senex. They are larger,
thicker and more woody than plants from Florida and Louisiana
that have been called P. senex.

17. Pyropolyporus linteus (B. & C.)
Polyponis linteus B. &. C. Proc. Am. Acad. Arts & Sci. 4:

122. i860.

Fomes linteus Cooke, Grevillea, 14 : 20. 1885.

The type specimens were collected on dead bark in Nicaragua.

Plants collected in Nicaragua by C. L. Smith agree with the

120 MURRILL : POLYPORACEAE OF NORTH AMERICA

types in all important particulars. The same species from Cuba
is labelled P. pcctinatus Kl. at Kew, though the latter plant does
not, to my knowledge, occur in North America.

1 8. Pyropolyporus Jamaicensis sp. nov.

A rather small fan-shaped plant, the upper layers of which are
dead and much cracked and roughened while the layers added
below are smaller each succeeding year. Pileus woody, applan-
ate, much thicker behind, 8 X 13 X 0.5-5 cm. ; surface uneven,
radiately rimose, dark brown to black ; margin ferruginous, velvety,
acute, becoming black, spreading and lobed, projecting 1—2 cm.
beyond the new layers: context woody, fulvous, 0. 5— i cm.
thick ; tubes stratified, separated by thin annual layers of con-
text, 0.2-0.7 cm. long each season, 7 to a mm., fulvous within,
mouths rounded or polygonal, hoary when young, edges thin,
acute, entire : spores globose or subglobose, thin-walled, smooth,
pale golden-yellow (probably darker in age), 5—7/^.

Collected by Earle, at Port Antonio, Jamaica, November, 1902,
on an old stump of Psidmni. The spores of the present season
appear immature.

Species Inquirendae

In this list is included not only species at present unknown,
but also plants described by foreign authors which have not been
seen or have been as yet only slightly studied by the writer.
Polyporns extensiis Lev., P. elahis Lev., P. Nicaraguensis B. & C,
P. sarcitis Fries, P. sdcrodes Berk., P. scleromyces B. & C, P.
sordidus Lev., P. siibflexibilis B. & C, P. Baccharidis Pat., Gano-
dcruia Mexicanmn Pat., Xanthochroiis igniarioides Pat.

New York City.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 33.

STUDIES IN THE ASCLEPIADACEAE-IV.-Vll.

By anna MURRAY VAIL

NEW YORK
1903

[Keprinted from the Bulletin of the Torrey Botanical Club, a6 : 423-431. 17 Au

1899 ; 38 : 4&5. pi. 30. 27 Au 1901 ; ag ; 662-668. 30 D 1902 ;

30 : 178, 179, pi. 9, 10. Mr 1903.]

[Reprinted from the Bulletin of the Torkev Potanical L'li b, a6 : Aug., i8<^.]

Studies in the Asclepiadaceae.— IV

]iV Anna Mikkay Vail

I. NOTES ON SOME OLD TYPES, WITH DESCRIPTIONS OF NEW OR LITTLE KNOWN SPECIES

AscLEi'iAs scAPOSA Vail, Bull. Torr. l^ot. Club, 25 : 171. 1898
On a specimen of this species in Wright's New Mexican col-
lection, no. 1683 in part, preserved in the Herbarium of the British
Museum, I was fortunate to find a somewhat mature but complete
flower, which through the courtesy of Mr. James Britten I was
permitted to examine closely, so that to my former description I
am able to add the following :

Corolla-segments 5 mm. long, apparently white ; column short
and broad ; hoods of the corona barely exceeding the anthers,
2-2.5 "'"''^- ^"''gJi' ^^'liite or whitish, slightly pendulous or saccate
at base, 5 -dentate at the apex ; central tooth rounded, the inter-
mediate teeth acutish, the two ventral ones infolded, erect, slender,
attenuated, nearly twice as long ; horn arising from about the middle
of the hood, slender, exserted ; anther-wings salient and somewhat
rounded at the base, apparently entire.

In floral structure allied to A. qidnqiicdcntata Gray and to A.
Palmcri Vail, but its low habit and solitary terminal peduncle are,
with few exceptions, rather unusual in the genus. I have seen
only four specimens of it, the one in the Herbarium of Columbia
University (Wright, no. 1684 in part), a fragment in Herb. Gray
from the same collection (no. 1683 in part), the specimen men-
tioned above, and the one referred to below. In the first three
cases it has been distributed with A. longicormi Benth. and A.
parviflora Willd. {A. pcrcnnis parinda Gray). The fourth speci-
men is in the Herbarium of the Missouri Botanic Garden and is on
a sheet with a specimen of (probably) A. longicornu Benth. in fruit
only, the label bearing the following inscription : " No. 7, Ascl.
longiconiu, Wright, 185 i."

Asclepias parvula (A. Gray)

Asdepias parvifolia Willd.? Torr. Mex. Bound. Surv. 164.
1859. Not Ait. 1789.

(423)

424 Vail : Studies in the Asclepiadaceae

Asclcpias pcrennis var. parvitla (A. Gray) Proc. Am. Acad. 12 :
70. 1876.

Asclcpias uivca Hemsl. Biol. Centr. Am. 2: 325 and 4 : 69.
1 88 1-2. Not Linn. 1753.

Pale gray-green throughout. Stem woody at the base, 3 dm.
high, or more, erect, pubescent, often branched above : leav^es
short-petioled ; blades lanceolate or oblong-lanceolate, tapering at
each end, 4-6 cm. long, rather thick and coriaceous, glabrous or
minutely puberulent above, puberulent beneath on the veins and
revolute margins : peduncles 3-9, terminal and lateral, 1-4 cm.
long; umbels io-20-flo\vere'd : corolla white, small; column
slender ; hoods hastately sagittate at the base, not exceeding the
anthers ; horn falcate, thin, exserted, inflexed over the anthers :
anther-wings narrow, entire at the salient base : follicles slender-
fusiform, tapering to each end, 7-9 cm. long, glabrous : seeds 7
mm. long, red-brown, very thin, glabrous ; coma 2-3 cm. long,
abundant.

Mexican Boundary Survey, Head of Rock Creek, Bigelow,
July 7, 1852; '' Nczu Mexico,'' Wright; Texas: Havard, Neally ;
Mexico: Palmer, no. 812.

The Wright specimen [no. 1684 in part] enumerated here is
in the Herbarium of Columbia University. No. 1683, also of
Wright's collection, contains besides A. parvula some specimens
of A. longicornu Benth. (the plant since described as Podostemma
Emoryi Greene, Pitt. 3 : 237), and some specimens of A. scaposa

Vail.

Stelmagonum ? Holtonii

A low perennial herb. Root slender, vertical : stems slender,
twining above, granular-puberulent and thinly hirsute with s.tiff
spreading hairs, the lower portion with small corky-barked ridges :
leaves opposite, on slender, granulose-puberulent and hirsute,
1.5-2.5 cm. long petioles ; blades ovate-cordate, 3-4 cm. long,
1.5-2 cm. wide, acuminate at the apex, basal lobes rounded with
narrow open sinus, granular-puberulent and with a few scat-
tered short stiff hairs above, granular-puberulent and with
more numerous stiff hairs beneath, especially on the veins, mar-
gins ciliate : flowers 4-7 in short-peduncled bracteolate cymes :
peduncles axillary, 3 cm. long ; pedicels 4-5 mm. long, the brac-
teoles subulate, very small, persistent : calyx 5-parted to below
the middle, 2-3 mm. long, granular-puberulent and ciliate ;
segments acuminate, with an erect subulate gland in each sinus :
corolla campanulate, j-^ mm. long ? or more, 5-parted to a

Vail: Studies in the Asclei'iadaleae 425

little below the middle, minutely granular-puberulent on the
outer surface; segments oblong, obtuse, glabrous within; crown
cohering to the corolla and to the raised gynostegium, cup-
shaped, 5-lobed,each lobe abruptly contracted into a slender, linear
I mm. long, erect ligule : stigma rounded, scarcely depressed :
pollinia orbicular, horizontal or ascending on rather broad, winged
caudicles ; corpuscle nearly rhombic. Follicles not seen.

New Grenada: Goudot ; Flora Neogranadina-Magdalena, I. F.
Holton, Opia, no. 461, Dec, 1S52.

Both of these specimens are in the Kew Herbarium and a
duplicate of the Holton number is in the Herbarium of Columbia
University.

Mellichampia ligulata (Bcnth.)

Ensknia ligulata Benth. PI. Hartw. 290. 1848.

Mtilichampia rubcsccns K. Gray ; S. Wats. Proc. Am. Acad. 22 :
437. 1887.

For some time past I have suspected the identity of Mclli-
chainpia with Enslenia ligulata and through the courtesy of the
Director of the Royal Gardens at Kew, I have obtained a tracing
of the type of the latter plant which confirms the suspicion. Be-
sides the type from Aguas Calientes, South Mexico, it has been col-
lected in the State of Jalisco, at Guadalajara, by Dr. Palmer, no.
280, July-October, 1886, and by C. G. Pringle, in copses near
Guadalajara, no. 5432, Aug., 1893. ^^r. Hemsley (Biol. Centr.
Am. 2: 358) also quotes a specimen without locality from Herb.
Pavon as belonging to this species. The specimens distributed as
Enslenia ligulata by Pringle (no. 4494) and Anipdanus ligulata by
A. A. Heller (no. 1899) are species o^ Roulinia.

II. THE TYPES OF GONOLOBUS MICHAUX AND DESCRIPTIONS OF NEW SPECIES IN

VINCETOXICUM WALTER

The types of the three species of Gonolobus (Michx. Fl. Bor.
Am. I : 119) are preserved in the Herbarium of the Museum in
the Jardin des Plantes, Paris, where they are represented by several
specimens each, all in a somev/hat fragmentary condition, though
quite recognizable and agreeing with the descriptions. In the
Richard Herbarium (Herbarium of M. Drake del Castillo, Paris)
the actual specimens owned and described by Richard in the Fl.
Bor. Am. can be seen and they are in every way exact duplicates

426 Vail : Studies in the Asclepiadaceae

of the plants in the Michaux Herbarium proper, but are appar-
ently more carefully labeled and named than those in the Museum.
These last were annotated by Dr. Gray. The first sheet there has as
inscription '' Cynanchuvi inacrophyllinn capsulis angiilosis " and is ap-
parently the plant previously named Vincctoxicum gonocarpos Walt,
and the plant which Dr. Gray (Proc. Am. Acad. I2 : 75) recog-
nized as Gonololms lacvis var. macropliyllns. The specimen in the
Richard Herbarium has the name " G. Jiiacrophylhts" and habitat
" In si/vis Caroliniae " on the sheet and is the same plant as the one
in the Herb. Michaux.

The second species, Gonolobus hirsutiis, is more difficult of de-
termination as of course I was not able to make a dissection of the
flower and the species is not so readily recognized as the first.
There are muricate follicles in both collections and in Herb. Rich-
ard two racemes, one dark purple and the other very faded, dull
greenish and shrunken ; the leaves are large and the plant could
be referred as readily to G. Carolpiensis as to G. hiisntus. It is
apparently identical with the plants that have been named G. hir-
siitiis var. flaviduliis. (See plants collected by Dr. Mellichamp at
Bluffton, S.C, and so named by Dr. Gray). Since seeing these types,
I have examined a long series of the G. Jiirsutiis and var. flavidu-
liis and also G. CaroUncnsis and have come to the conclusion that
it will take very critical study and much more material, especially
fresh material, to determine whether there are really two species
there or only one. The coronal characters are difficult to reach in
the dried plants and in the specimens which have passed through my
hands I have found every form of crenation, both regular and ir-
rep-ular and ereat variation in the thickness of the crown-margin
and also numerous instances in which the thickened alternate cre-
nations have a very pronounced horn-like process within. Some-
times these last characters showed themselves in all stages of
development from a faint obscure ridge near the apex of the crown-
segment to a sharply incurved tooth. These again were to be
found on one or two of the segments and again on every one of
them. So far as I have been able to note, these characters are
constant on the same plant. In some specimens the crown is
uniformly thin, entirely lacking the alternate thickened divisions
of the descriptions, others again have the thin geminate teeth

Vail: Studies in nii': Asclepiadaceae 427

claimed for (i. Carolincnsis and the quadrate alternate divisions
belongini^ to G. hirsiitiis. It seemed entirely impossible with the
material at hand to draw any fast or hard line between them. The
poUinia that have very marked characters in the other species seem
to be about the same in these two species and I could not find any
differences between those of the flowers labeled G. Iiirsiitus and of
those called (j. Carolincnsis.

These two Michauxian species constituted the older genus
Vincctoxicuin of Walter which leaves the third species G. lacvis
as the type of the genus Gonolohns.

The type of Gotiololms lacvis is somewhat of a curiosity.
There is quite a good deal of it, small pieces mostly, but all
of the specimens in the two collections agree and point unmis-
takabl)' to the plant since called Enslcnia albida Nutt. {Am-
pclanus albidaV>x\\X., Bull. Torr. Bot. Club, 2i: 314). In the
Michaux Herbarium there are on a sheet a small specimen of G.
snbci'osns} (the name and query are in Dr. Gray's* handwriting),
one angled follicle and one raceme, noted as G. lacvis, also by
Dr. Gray, and two more leaves and two follicles. The old labels
read ''Cynanclinni capsnlis lacvis" and ''Gonolohns lacvis, Illinois.''
In the Richard Herbarium the specimens, consisting of some
leaves and one angled follicle, are much worm-eaten. They are
all. except the fragment of G. suberosus, Enslcnia albida. This
will clear up the discrepancy of the original description of G.
lacvis with the plants that have passed as such. " Foliis quasi
conoidco-cordatis, scnsiviacuds ncrvis tantnni miniitissime puberidis"'
describes certainly the leaves of Enslcnia albida, but scarcely ac-
curately those of the so-called Goiiolobus lacvis. The synonymy
of this plant should therefore be as follows :

GoNOLOBUS LAEVis Michx. Fl. Bor. Am. i: 119. 1803

Enslcnia albida Nutt. Gen. Am. i: 164. 18 18.
Ampclanus albidns Britton, Bull. Torr. Bot. Club, 2i: 314.
1894.

* Gonolobus suberosus and G. viacrophyllus are verj' close in general appearance,
though the former has commonly yellowish-green leaves, with truncate base, and the
latter darker green leaves that have notably large rounded basal and often overlapping
lobes.

428 Vail : Studies in the Asclepiad.aceae

The type of this species is from lUinois, where it is not infre-
quently met with on river-banks and in thickets. It has a wide
range, eastward and southward. The specimen quoted from
Washington, D. C, as Vincetoxiciun gonocarpos laevism. Britton &
Brown, Ilkistrated Flora, 3: 18, is in fruit and is the true Gonolo-
bus lacvis Michx. (Herb. Columbia University).

A second species Gonolobus volubilis {Nematiiris vohibilis
Turcz. Bull. Soc. Nat. Imp. Mosc. 21 : 254. 1848. Enslciiia
volubilis Karst. Fl. Columb. 2: 117. PI. 162. 1866) occurs
near Pt. Cabello, U. S. of Colombia.

ViNCETOXicuM SUBEROSUM (L.) BHtt., Mem. Torn Bot. Club, 5:

266. 1894.

Cynanchiuii sitberosiim L. Sp. PI. 212. 1753. Dill. Hort. Elth.
300. pi. 22g. f. 2g6, excluding Hort. Cliff, and Gron. 27.

Gonolobus suberosus R. Br. Ait. Hort. Kew. ed. 2, 82. 18 11.

Dr. Gray * has written the history of this species and in addition
to his notes it may be of interest to point out that the plant
" Apocynuiji scandens fruticosiun fungoso corticc Brasiliamimy
Herm. Parad. 53 is probably the plant since referred to Ibatia, and
is not the specimen of the Hortus Cliffortianus which resembles it
only as to the corky, ridged bark of the stems. The Cliffort plant
can be seen in the Herbarium of the British Museum and is diffi-
cult to identify as it is a mere fragment, but the character of the
stem should make it easily recognizable were there more available
material of the same species for comparison.

I venture to describe the following species as new.

Vincetoxicum Floridanum.

Puberulent throughout. Stems very slender, hirsute with few
short, scattered hairs : leaves opposite ; petioles 5-20 mm. long,
angled ; blades ovate, 2-5 cm. long, cordate, tapering to the acumi-
nate apex, the basal lobes rounded, with open sinus, about equally
puberulent with a fine soft pubescence on both surfaces; midvein ob-
scurely bi-glandulose at the base above : racemes about the length
of the petioles ; pedicels 1 2 mm. long : calyx very small ; seg-
ments 2 mm. long, linear-lanceolate, with a subulate gland in each
sinus: corolla dull greenish-purple, 5 -parted to a little below the

* Proc. Am. Acad. 12: 75. 1876.

Vail : Studies in the Asclepiadaceae 429

middle ; segments linear-lanceolate, acute, 3-5 m. long, somewhat
fleshy, minutely puberulent, on the outer surface, glabrous within ;
crown red-purple, saucer-shaped, with 5 broad undulations each
bidentate at the callous thickened apex, and a smaller tooth on
each side below the middle : stigma depressed, not 5-angled : pol-
linia oblong, the caudicles broad and apparently not twisted ? Fol-
licles not seen.

East Florida : Dr. Leavenworth.

The specimen from which this description has been drawn is
in the Herbarium of Columbia University and was labeled by Dr.
Torrey, Gonolobus " jiia:rop/ijllus." It was seen by Dr. Graj'
when he was working on the Asclepiadaceae for the Synoptical
Flora and bears his note to the effect that it has "short denticu-
late lobes" and on the Synoptical Flora label the doubtful one of
"seemingly Gonolobns Carolincnsisy It differs however from
that species, as elsewhere ticketed by Dr. Gray in its much smaller,
greenish flowers, and the crown which has not the regular divisions
of that of Vincetoxiciiin Carolinensis, but especially in the inflores-
cence which is irregularly racemose, whereas that of V.Carolinensis
is more cymose and much longer pedicellcd.

Vincetoxicum crenatum

A twining vine. Stems somewhat angled, papillose-puberulent
and retrorsely hirsute with scattered short hairs : lea\es opposite,
on slender, striate, 3-5 cm. long petioles ; blades yellowish-green,
6-9 cm. long, ovate, rather long-acuminate at the apex, cordate,
basal lobes rounded with open or closed sinus, papillose-puberu-
lent on both surfaces, paler beneath : inflorescence sub-corymbose ;
peduncles 5-8 cm. long, 4-10-flowered ; pedicels slender, 8-15
mm. long, 1-2 bracteolate at base : calyx-segments ovate, acutish,
puberulent outside, ciliate, glabrous, and with a subulate gland in
each sinus within: corolla 13-14 mm. long, dull yellowish pur-
ple, rotately spreading, ovate-conic in bud ; segments linear-ob-
long, acutish, puberulent outside, glabrate within and vertically
reticulated, sparingly and minutely hirsute below the sinus, trans-
versely wrinkled and glabrous in the short tube : crown shallow,
saucer-shaped, 5-crenate ; lobes rounded, not exceeding the an-
thers, each with a short, barely free internal process or appendage
at about the middle ; process truncate at the apex : stigma 5-an-
gled, not depressed : anther-appendages small, fleshy ; pollinia
obliquely semi-orbicular, saccate and broader at the summit,
dented at the angled base, caudicles and corpuscles short. Follicles
not seen.

430 Vail : Studies in the Asclepiadaceae

Mexico : Barranca near Cuernavaca, State of Morelos, Pringle,
no. 6388, July 27, 1895.

Distributed as Gono/odus p'/osns Benth. Resembling in habit
Goiiolobus angustilobns Rob. & Greenm. (Proc. Am. Acad. 29 :
388. 1894) from the State of Jalisco, but differing in its yellow-
ish-green foliage, the more numerous flowers and the curiously
reticulated character of the corolla. The leaves of G. angustilobns
are grayish-green, the flower appears to be solitary and the corolla
is not at all reticulated. Detailed floral characters of this last spe-
cies, owing to lack of material, are not accurately known.

In regard to the true Gonolobus pilosus Benth. (PI. Hartw.
289. 1848) it is perhaps worth noting that a specimen collected
by Dr. Coulter, in Mexico, no. 975, is preserved in the Herbarium
of Columbia University. This number is quoted by Hemsley
(Biol. Centr. Am. 2 : 333) as belonging to Gonolobus pilosus Benth.
It coincides in every respect with the description of that species.
The flowers are at least 3.5 cm. in diameter when open, and are
of a dark, dull reddish-purple. The calyx-segments and bracte-
oles are over 13 mm. long, ovate, acutish and foliaceous. The
crown is barely 5 -parted, lacerate-denticulate along the whole
margin, each division with an adnate, thickened internal appendage
which is laciniate and barely free at the broadly quadrate summit.
The pollinia are remarkably large, obliquely oblong, rounded at
the base and somewhat saccate, slightly tapering to the curved
caudicles ; corpuscle broadly obovate at the apex, abruptly con-
tracted to the much narrower rounded base. The hyaline anther-
tips are conspicuously large.

Vincetoxicum Greggii

Gonolobus productusToxx.Vie^.ViO\xxvd.^wx\.\6'^. 1859. In

part.

A slender, twining vine. Stems minutely puberulent and hir-
sute with short scattered hairs : leaves opposite ; petioles 5-1 5 mm.
long; blades 1.5-3 cm. long, or more, ovate-hastate, long-
acuminate at the apex, the basal lobes rounded with open sinus,
rather thick, papillose-puberulent on both surfaces : inflorescence
subcorymbose ; peduncles 1.2 cm. long, 5-8-flowered : calyx 2 (?)
mm. long, minutely hirsute ; segments ovate-triangular, acute, with
a subulate gland in each sinus : corolla subrotate, 6-^ mm. long ;.

Vail: Studies in the Asclepiauaceae 431

crown 5 -parted, to below the middle ; segments thick and fleshy,
broadly rounded at the apex, each with an internal ligulate horn
or process arising from near the base, exceeding the anthers and
incurved over them: stigma 5 -angled : pollinia quadrate on
slender, winged caudicles, corpuscle narrowly oblong. Follicles
not seen.

" Slender vine-like plant, flower purplish," Cadena, Mexico,

Dr. Gregg, May 8, 1847.

The specimen described here is in the Herbarium of Columbia
University and was included by Dr. Torrey in his description of
Gonolobus prodiictns and was ticketed under that name by Dr.
Gray when he revised the Torrey Collection for the Synoptical
Flora. It is closely allied to Vincetoxicum acuminatum {Gonolo-
bus aanninatns A. Gray, Proc. Am. Acad. 21 : 399. 1 888-6) as well
as to V. productum {Gonolobus prodiictus Torr.) All three species
belong in Dr. Gray's section Chthamalia and without careful dis-
section could scarcely be distinguished one from the other, except
that V. Grcggii has a more rotately spreading corolla than its con-
geners.

There are marked differences in the pollinia of the three species.
In V. prodiictnm the pollinia are spreading and perhaps subpendu-
lous, obliquely oblong, twisted (?) and somewhat bent or dented
at the middle ; the caudicles are apparently winged and the cor-
puscle is oblong, also appearing as if winged. The pollinia of V.
acuminatiim are more truly horizontal, subovoid, rounded on one
side and straight on the other, on broad caudicles, with a broad
corpuscle. V. Grcggii has ovoid-quadrate pollinia, with very
slender, somewhat twisted caudicles and a slender oblong cor-
puscle. In this last species the leaves are smaller than in the two
others, the corolla is more rotately spreading, the crown is more
deeply parted and the internal ligules are free and incurved over
the anthers. The pollinia and crown characters of V. Grcggii are
also nearly those of Gonolobns bifidiis Hemsley, but in that species
the corolla is much smaller and truly rotate. In V. acinnina-
tiim the crown -segments are quadrate at the apex and the internal
processes are short, barely free at the apex.

(Reprinted from the Huli.btin of tuf Tokrky Botanicai. Ci.i'ii, 28: Anpiisi, 1901.]

Studies in the Asclepiadaceae.— V. A New Species of Vincetoxicum

from Chihuahua

By Anna Murray Vail
(With Plate 30)

Vincetoxicum Wootonii

{Gonolobus A. Gray, non Michaux)

A low branching perennial herb with small axillary sub-
umbellate clusters of dull greenish-purple flowers : stems angled,
puberulent with short, apparently twisted hairs : leaves opposite,
on petioles 3-5 mm. long ; blades 7 mm.-i cm. long, pale gray-
ish-green, round-ovate, obtuse at the apex, cordate, the basal lobes
rounded, obtuse, thick and coriaceous, glabrous above, puberulent
on the veins beneath, ciliate : peduncles 5-12 or 15 mm. long,
3—6- or /-flowered, puberulent: pedicels very slender, 4-6 mm.
long, puberulent, bracteolate at base : calyx-segments ovate-lanceo-
late, less than 2 mm. long, acute, puberulent, purplish, eglandu-
lose at the base within : corolla-segments rotately spreading, dull
yellowish-green, thickish, less than 4 mm. long, oblong-ovate,
obtuse, glabrous, vertically reticulated, subglobose ? in bud:
corona saucer-shaped, 5-lobed, somewhat higher than the anthers,
fleshy ; lobes broadly rounded, thickened at the center on the inner
side, appendaged with 5 ligulate Asclepias-\\kQ incurved horns which
adhere to the lobe to near the middle : stigma flat, barely 5-angled :
pollinia obliquely semi-orbicular ; caudicles broadly winged, very
slender at the attachment to the narrowly oblong corpuscle.

Nearly related to Vincctoxiann biflonim (Raf.) Heller and
remarkable for the small leaves and AscUpias-Vik^ horns.

Mexico: State of Chihuahua, Colonia Garcia, Sierra Madre,
altitude about 7200 ft., collected by Elmer O. Wooton, no. 13,
June, 1899.

Explanation of Plate 30

Fig. I. Vincetoxicum Wootonii, slightly reduced.

Fig. 2. Flower diagrammatically enlarged , X 5-

Fig. 3. Flower enlarged, X 2^.

Fig. 4. Corona diagrammatically enlarged.

Fig. 5. Pollinia enlarged, X 50-

485

[Reprinted from ihe Bullktin op thi; Tokrev Botanilal Cli'B, 19 : December, 1903. |

Studies in the Asclepiadaceae— VI. Notes on the Genus Rouliniella

By Anna Murray Vaii.

A close examination of herbarium specimens variously labelled
Roulinia Jacqiiiiiii, R. unifaria and R. raccmosa show that there
are several species which if not new are at least deserving of rec-
ognition ; and this is an attempt to straighten out some of the
species of a most interesting little genus.

About twenty-one or twenty-two species can be enumerated,
ranging from Texas to Argentina ; but as the genus is but poorly
represented, even in foreign herbaria, it is difficult withoi^t careful
dissection to determine whether all the species credited to Roulinia
are correctly placed.

Rouliniella

Roulinia Decne.; D. C. Prod. 8: 516. 1844; Schumann,
Die nat. Pflanzenfam. 4-: 255. 1895. Not Brongniart, 1840.

The seven species enumerated here all bear small flowers and
are closely related. The available material of the South American
species is at present too scanty to study with any kind of satisfac-
tion.*

Ke>- to tlie Species

Crown-segments not exceeding the stigma. i. R. unifaria.

Crown-segments barely exceeding the stigma, conspicuously 3-lobed.

2. R. Coltinibiana.
Crown-segments exceeding the stigma, long-ligulate.
Flower-buds sub-globose.

Central lobe of the crown-segments obscurely 3- toothed at the apex ; floweT-

buds obtuse. 3. R. Palmeri.

Central lobe of the crown-segments entire ; flower-buds acutish.

4. R. racemosa.
Flower-buds ovate, acute or sub-acuminate.

Central lobe of the crown -segments truncate at the apex. 5. R. lignosa.
Central lobe of the crown-segments acute at the apex. 6. R. foetida.
Central lobe of the crown-segments very long and slender, undulate.

7. R. Jaiiscana.

*The accompanying figures represent : ^, the bud ; c, the crown (diagrammaticaliy
enlarged) ; /, the expanded flower, and are all magnified four diameters ; /, the pollinia,
are magnified twenty diameters.

(562

663

Vail : Studies in the Asclepiadaceae

I. Rouliniella unifaria (Scheele)
Gonolobus miifarius Scheele, Linnaea, 21 : 760.

1848.

Roidinia unifaria Engelm. Mex. Bound. Surv. 160. 1850.

A slender climbing and twining vine. Stems pale gray-green
when fresh, striate, glabrous or minutely pubescent in lines on the
upper portion and somewhat sulcate : leaves opposite, on petioles
1-4 cm. long ; blades 3-7 cm. long, ovate, somewhat hastately cor-
date, gradually acuminate, glabrous, thin, the auriculate basal
lobes rounded, with a broad open sinus and a few glands at the
base of the midvein above : peduncles longer than the petioles
and as long as or exceeding the leaves : racemes as long or longer
than the leaves, 7-12-flowered : rhachis minutely pubescent or
puberulent in lines : flowers globose in bud : pedicels stout, i cm.
long or^iess, pubescent : calyx-.segments lanceolate-oblong, 4 mm.
long, with a gland in each
sinus : corolla 4-6 mm. long,
white ; segments 3 mm. long,
linear-oblong, 2.5 mm. wide,
broadly obtuse or nearly trun-
cate, thin, 3 -veined, the callous
thickened margins revolute ;
crown attached to the base of
the column, 5 -parted to near
the base ; segments erect, not
exceeding the anthers, min-
utely 3-lobed, the middle one
slightly the longest, the lateral
lobes with thin infolded margins at the base within : anther-
wings small ; anther-tips semi-orbicular, inflexed on the scarcely
conic stigma: pollinia elliptical -ovoid ; caudicles broad at the
attachment, orange-red on the upper margin ; corpuscle obtuse at
the apex : follicles 6 cm. long, ovate, acuminate with an obtuse
apex, coriaceous, glabrous : seeds 5 mm. long : coma i cm. long.
{Fig. /.)

Type locality: On the Upper Guadeloupe River, near New

Braunfels, Texas. Thickets, Kerr county, Texas, to " Mts. of

Muerte," New Mexico. Wright ; western Texas to El Paso,

Wright, 5-1-6 ; near Kerrville, Heller, iSgg (distributed as Am-

pelanus ligulata) ; Mexican Boundary Survey, no. lo^g.

The specimens cited by Hemsley, for Roidinia imifaria from
" North of Monclova, Coahuila," Palmer, 82g are Roidiniella Pal-

FlG. I.

V.\ir, : Studies in the Asclepiadaceae

(i(;4

meri. I have not seen any true R. nnifana as yet from Mexico.
It closely resembles R. raccniosa in general appearance, but in the
latter species the leaves are commonly larger, the segments of the
crown are 3-lobed, the central lobe long and acutely ligulate and
the pollinia are oblanceolate in outline on very long slender cau-
dicles. The crown-segments of R. toiifaria are not long-ligulate
and the pollinia are merely oblong in outline, the caudicles being
very short and broad.

2. Rouliniella Columbiana sp. nov.

A twining vine up to 7 m. long. Stems puberulent or gla-
brate, slender : i)etioles channelled, 2-4 cm. long or somewhat
shorter than the blades : leaves opposite; blades 3.5-8 cm. long,
ovate-oblong, abruptly acuminate at the ape.v, the cordate basal
lobes rounded, spreading, yellowish-green on both surfaces, spar-
ingly pubescent with scattered hairs : racemes about the length of
the petioles, io-20-nowered : peduncles and pedicels puberulent,

channelled : calyx-segments
2.5 mm. long, lanceolate,
minutely puberulent : flow-
ers sub-globose in bud,
about 5 mm. in diameter
when mature : corolla- seg-
ments 2 mm. long, ap-
parently pale yellow, ob-
long, obtuse or acutish
when dry, the margins revo-
lute; crown 5 -parted ; seg-
ments white. 3-lqbed, hood-
like, the central lobe longer and incurved over the stigma, the
lateral lobes with a narrow callous incurving ridge or crest on the
inner side : antlu'r-tips scarious, inflexcd, conspicuous : pollmia
oblong on broad caudicles. Follicles not seen, {^''g- 2.)

U. S. Colombia : Bohda, thickets on the lowlands near the river,
1898-99, rare: Herbert H. Smith, 166S. Flowers Oct.-Nov.
Type in the herb, of the New York Botanical Garden.

Fig. 2.

3. Rouliniella Palmeri (S. Wats.)
Roidinia Paluicri S. Wats. Proc. Am. Acad. 18

I I

> '883.

A slender twining nearly glabrous vine. Stems with a narrow
puberulent line : leaves on slender 3—4 cm. long petioles ; blades
cordate-hastate, with a wide open basal sinus, acuminate at the

665

Vail : Studies in the Asclepiadaceae

Fig.

apex, 3-8 cm. long, grayish-green, thin : inflorescence sub-corym-
bose : peduncles 3-5 cm. long, 5-7-flowered : flowers sub-globose
in bud : calyx-segments slender, linear-lanceolate, 4-5 mm. long,
puberulent : corolla greenish-white, obtuse in the bud ; segments
somewhat crisped and revolute on the margin, a little longer than
the calyx : crown -segments
united at the base, quadrate,
3-lobed at the apex ; middle
lobe triangular, acutish or
obtuse, obscurely 3-toothed
at the apex, a little longer
than the infolded lateral
lobes : stigma nearly flat :
pollinia oblong ; caudicles
broad, with an orange-colored
spot along the upper margin :
follicles 10 cm. long, some-
what spongy when dry,
wrinkled, glabrous : seeds 7

mm. long, glabrous, crenate along the apex: coma 2.5-3 ^^n-
long. {Fig. J.)

Mexico : Mountains 24 miles northeast of Monclova, State of
Coahuila, September 1880; Palmer, 82^.

4. Rouliniella racemosa (Jacq.)
Cynanchuin raceniosum Jacq. Select. Am. 'i\. pi. 5^. 1788.
Roiilinia racemosa Kuntze, Rev. Gen. PI. 422. 1891.

A much branched twining vine. Stems angled, slender, gla-
rous except for a narrow pubescent line : leaves on petioles 2-4

cm. long ; blades ovate-cordate,
6—9 cm. long, long-acuminate,
with broadly rounded basal
lobes and open sinus : inflor-
escence racemose, shorter than
the leaves : peduncles 4—7 cm.
long, 10— 20-flowered : flowers
5 mm. in diameter, sub-globose
in bud : calyx-segments lanceo-
late, acuminate, with a gland in
each sinus : corolla greenish-
white ; margins of the lobes
crisped : crown -segments united
at base, obtusely 3-lobed at the
twice or more longer than the

Fig. 4.

summit

central lobe ligulate.

Vail : Studies in the Asclepiadaceae

666

infolded lateral lobes : sticjma rounded at the apex, scarcely
conic : pollinia oblonj^ ; caiidicles broad and short, with a deep
orange-yellow spot alon<r the upper margin ; corpuscle large,
bright red. Fruit not seen. (/'4''- 4-)

Type locality, Cartagena, U. S. Colombia.

Guatemala : Patulul, Depart. Solola, alt. 3000 pp., Heyde et
Lux, <5j./p, January, 1894.

Venezuela : near Tovar, Fcndlcr, lojj (in herb. Gray).

5. Rouliniella lignosa sp. nov.

A woody vine. Stems pale grayish-brown, glabrous : branches
2—3 dm. long or more, somewhat angled, minutely tomentulose in
hnes, becoming glabrate : leaves opposite, on petioles 1.5-4 cm.
long ; blades ovate or sub-lanceolate, cordately hastate, 3—6 cm.
long, gradually acuminate, rather thick, yellowish-green, glabrous
above, minutely pubescent on the veins beneath ; basal lobes
rounded, short, with an open sinus^ glaiidulose at the base of the

midvein : peduncles longer
than the petioles, pubescent :
racemes slender, 5—7 cm. long,
exceeding the leaves, 9-15-
flowered : buds ovate : bract-
eoles very small, setaceous :
pedicels 7—9 mm. long, slender,
pubescent: calyx -segments
ovate-lanceolate, i mm. long,
puberulent with a minute gland
in each sinus : corolla 5 -parted
to near the base, dull yellowish-
purple, acute in the bud ; seg-
ments 6 mm. long, acute.

Fig. 5.

minutely hispidulous on the inner surface, with a callous ridge
within the membranous whitish revolute margins : crown
attached to the base of the column, 5-partcd to the base; seg-
ments very thin, with somewhat involute lateral margins, pro-
longed into a long-ligulate obtuse or truncate apex, which is nearly
as long a."^ the corolla-segments : anther-wings very narrow and
slender, not salient at the base : anther-tips ovate, inflexed over
the barely conic, obscurely 5-rayed stigma : pollinia broadly ob-
long-ovoid ; caudicles broad ; corpuscle obtuse at the apex. Fol-
licles not seen. {Fig. 5.)

Mexico: State of Jalisco, Rio Blanco, i2)^6, Palincr, ji^ ;
(herb. Columbia University): ''Cynaiic/unii parvijlonun" Pavon,

667

Vail : Studies in the Asclepiadaceae

in herb. Kevv : Oaxaca, Gliiesbreght (also in herb. Kew). The
last two specimens are cited in Hemsley, Biol. Centr. Am. 2 : 329
under Roulinia Jacqjiini.

Differing from R. nnifaria and R. racemosa in its more woody
stem and lighter green thicker leaves. It has the crown-segments
of R. racemosa but the pollinia of R. nnifaria.

i 6. Rouliniella foetida (Cav.)

Asclcpias foetida Cav. Ic. 2 : 45. //. ij8. 1793.
Cynanchuni foetidnni H.B.K. Nov. Gen. Sp. 3 : 202. 181 8 ?
Roulinia Jacqidfiii T)Q. Prodr. 8: 517. 1844. Excl. syn.

A slender twining vine. Stems glabrous or with a narrow
puberulent line : petioles slender, 3—7 cm. long : leaves opposite ;
blades ovate or some of the smaller ones ovate-lanceolate, 5— ii
cm. long, acuminate at the slender apex, cordate at the base with
open sinus and rounded basal lobes, puberulent on the veins be-
neath : flowers 8-12 mm. in diameter in small sub-umbellate
clusters ; buds ovate : peduncles
minutely puberulent, 2-3 cm.
long, commonly shorter than
the petioles: calyx 1.5 mm.
long; segments linear-oblong,
acutish at the apex, minutely
ciliate, with a large gland in
each sinus : corolla 6-7 mm.
long, dull- greenish, fleshy,
glabrous on the outer surface,
whitish or cream color ? with
two raised puberulent callous
ridges or crests along the
upper portion within ; crown -

segments white, very thin, 5, the lateral margins infolded and
adnate to the inner surface, the long ligulate apex very slender, as
long as the corolla segments or nearly so : anther-wings long and
narrow : pollinia oblong, on short broad pale yellow caudicles ;
corpuscle reddish, rounded at the apex, narrowet at the bifid base :
stigma barely conic, 5 -rayed : follicles angled and winged. Mature
follicles not seen, [Fig- ^O

Mexico : Oaxaca, Ga/cotti, ijj8 (in herb. Kew) ; near
Oaxaca, alt. 1600 m., 1896, C. Conzaiti, 2iy ; Valley de Etla,
Sept. 1895, C. Ahmrcz, j68 ; Huitzo, alt. 1700 m., 13 Aug.
1897, C. Conzatti & V. Go)Lzalcs, jj6 (in herb. Gray).

\'a1L: SjUDUi.S IN IlIE ASCLEI'IADACKAE

668

The plant from which the Cavanilles figure was drawn is said to
have been grown at the Royal Botanic Garden, Madrid ; and was
native of Mexico. It agrees well with the specimens just cited,
though in detail the central lobe of the crown-segments seem shorter.

7. Rouliniella Jaliscana sp. nov.

A twining vine. Stems glabrous, angled : branches glabrous
or with a narrow puberulent line above : petioles channeled ? 2—7
cm. long or more, glabrous or mjnutely puberulent : leaves

opposite; blades 4-12 cm.
long, ovate-oblong, abruptly
acuminate at the apex, trun-
cate at the base, dark green
or glabrous above, paler,
glabrous or minutely puber-
ulent on the veins beneath :
r acemes shorter than the
leaves, 8-1 2 -flowered : buds
ovate : peduncles minutely
puberulent in lines, chan-
neled : calyx-segments ob-
long, obtuse, 2 mm. long,
minutely ciliate, oblong-
linear : corolla -segments
4-5 mm. long, broadly
obtuse or truncate at the
obtuse somewhat ligulate
revolute apex, with a narrow callus incurving ridge or crest on
each side near the middle within : crown-segments white, 3-lobed,
the lateral lobes infolded within ; the central lobe long-ligulate,
very slender, undulate, nearly as long as the corolla-segments :
anther wings very short, corneous : stigma 5-rayed, flattish : anther
tips semi-orbicular, inflexed : pollinia oblong, pyriform on long
slender caudicles : corpuscle lozenge-shaped. Follicles not seen.

Mexico : State of Jalisco, Palmer, j8^ ; State of Jalisco,
Pringle, 44^4 ; Orizaba, Bottcri. (Type in herb, of the New York
Botanical Garden.)

New VukK Botanical C/xrden.

Fig. 7.

Studies in the Asclepiadaceae — VII. A new Species of Vincetoxicum

from Alabama*

By Anna Murray Vam.
(With Plates 9 and 10)

Vincetoxicum Alabamense

High climbing, sparingly hirsute with long rather weak hairs
and a minute glandular pubescence : petioles 3-8 cm. long, chan-
nelled ; leaf-blades 6-15 cm. long, 5-10 cm. broad, rather thin,
ovate or oblong-cordate, acuminate at the apex, the basal auricles
rounded and never Ov'erlapping, the sinus narrow at the base,
sparingly puberulent and dark green above, lighter and more
closely puberulent beneath : peduncles shorter than the petioles,
3-6-flowered : calyx bi-glandulose, the lobes lanceolate, acute :
corolla dull greenish -yellow, 2 cm. in diameter, 5 -parted to near
the base ; segments oval, obtuse, puberulent and minutel)' papillose
on the outer surface, conspicuously reticulated within ; gynostegium
raised on a low broad column ; crown orange-yellow, consisting of
a low saucer-shaped ring attached to the base of the corolla, 5-
parted, each division minutely and irregularly toothed or entire,
appendaged within by 5 tooth-like erect crests attached to the ba.se
of the column ; stigma white, flat, obtusely 5 -angled : anther-tips
white, scarious ; pollinia obovate : ovaries minutely puberulent :
folHcles 8-9 cm. long, glandular-puberulent, closely and strongly
muricate : seeds 8-9 mm. long, obovate, granuiose, tipped by a
tuft of long silky coma.

" Sandy hillsides on edge of woods," Dale County, Ala-
bama, collected by T. G. Harbison. June 3 and September 5,
1902.

Differing from Vincetoxicum reticulatimi (Engelm.) Heller in
the greater size of its leaves and flowers, the more strongly muri-
cate follicles and the horn-like appendages to the crown.

In Vincetoxicum reticulatum the leaf-blades are 3-6 cm. long,
the basal sinus broad with often overlapping auricles ; the racemes
are mostly the length of the petioles ; the flowers are scarcely
more than 1.4 cm. in diameter jA^hen expanded ; the crown is very

* Presented by invitation before the Botanical Society of .\merica, Washington.
D. C, January i, 1903.

178

1 79 Vail : Studies in the Asclepiadaceae

obscure, and the horn-like projections of the column are also ob-
scure or lacking.

New York Botanical Garden.

Hxplaiiation of HIates

Plate 9, Vincetoxicum Alobamense

a. Flower enlarged, X 3-

b. Calyx (without the corolla) enlarged, X 8.

c. Sepals (showing the two glands) enlarged, X 16.

d. Ovaries enlarged, X ^^•

e. Pollinia enlarged, X 20.

f. Diagrammatic transverse section of gynostegium enlarged, X 16.

,^, h. Different aspects of the outer and inner crown (the gynostegium cut off)
enlarged, X 1 6.

Plate id. Vincetoxicum reticulalum

a. Flower enlarged, X 3-

b. Calyx (without the corolla) enlarged, X ^•

c. Sepal (showing the solitary gland) enlarged, X 16.

d. Ovaries enlarged, X '6.

e. Pollinia enlarged, X 20.

f. Diagrammatic transverse section of gynostegium enlarged, X l^-

g. h. Different 'aspects of outer and inner crown (the gynostegium cut off) en-
larged, X 16.

Bull. Torr. Bot. Club, 28.

PL. 30.

n

H

O

O

I— I

o

G

O
O
H
O
55

j^'^^\yM

VINCETOXICUM ALABAMENSE Vail.

Bull. Torrey Club, 30.

Pi.. 10

VINCETOXICUM RETICULATUM (Engelm.) Heller.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 34.

THE GENUS RIELLA, WITH DESCRIPTIONS OF

NEW SPECIES FROM NORTH AMERICA

AND THE CANARY ISLANDS

By M. a. HOWE and L. M. UNDERWOOD

NEW YORK
1903

[Reprinted from the Bulletin of the Tokrey Botanical Club, 30: 214-224. Ap 1903]

' The Genus Riella, with Descripticns of new Species trom North
America and the Canary Islands *

Itv M. A. Howe and L. M. Underwood
(With Plates ii and 12)

The genus Riella occupies a unique position among the Hepa-
ticae. The striking pecuHarities of its gametophytic phase have
attracted the attention of such niorphologists as Hofmeister, Leit-
geb, and Goebel, in addition to the interest excited among those
who have devoted themselves more exclusively to a study of the
systematic relations of the Hepaticae. Riella helicophylla, an Al-
gerian species, is alluded to in some of the standard botanical
text-books as being peculiar among the liverworts in having a
leaf-like lamina or wing disposed spirally in relation to the axis or
stem. Later studies of this species, however.f indicate that the
supposed helicoid spiral arrangement was exaggerated in the
original figures and description and that the spiral appearance is
due to the slight torsion of a stem bearing a strongly undulate
lamina. Nevertheless, the species of Riella m general are peculiar
enough in that the lamina or wing appears at first sight to be
attached to one side of the stem ; but the position of the sexual
organs, of the root-hairs and of the scale-like appendages shows
that the plant is bilaterally symmetrical in the plane of the wing
and the conviction is now general that the wing is dorsal in relation
to the stem. Goebel % has expressed the opinion that the chief
difference between Riella and the other liverworts is that in Riella
the development of the thallus is in the vertical instead of in the hori-
zontal plane. The species of Riella are all aquatic, commonly
growing entirely submerged, and it is doubtless this condition of
growth which makes possible the leading peculiarity in form.

The growing point of a young plant or of a young branch of

* Read in abstract before the Botanical Society of America, Ninth Annual Meet-
ing, Washington, D. C., December 31, 1902.

fTrabut, Rev. Gen. Bot. 3 : 451. 1891. Stephani, Bull. Herb. Boiss. 7: 659.
1899. •

+ Flora, 77 : 107. 1893.

214

215 Howe and Underwood: The Genus Riella

Riella is not apical in the ordinary sense of the word, but inter-
calary. The whole tissue of the young shoot appears to be meri-
stematic at first and of one layer of cells. Later, the growth ac-
tivities are concentrated at either one or two points on the margin of
the shoot intermediate between its distal and proximal extremities.
If at two points, these are situated on opposite margins of the shoot.
The new cells formed on the upper or distal side of the growing
point now contribute to the growth of the unistratose wing, which
is dorsal in position. The new cells formed on the lower or prox-
imal side of the growing point go to constitute the multistratose
stem, which is ventral in position. If two growing points on oppo-
site margins of a young shoot persist, a double or twin plant is
formed, the two branches of the axis bearing a single continuous
dorsal wing. If, however, only one growing point persists, the
plant or branch is apparently one-sided, with stem on one side and
wing on the other, though in reality the wing is dorsal and the
stem ventral.

On account of the absence of elaters, Riella was at first placed
with the Ricciaceae, together with SpJiacrocarpiis, in which, like-
wise, elaters are not developed. A better understanding of the
morphology of these two genera has led to placing them in the
order Jungermanniales, of which, together with the exclusively
American genus Thallocarpiis, they constitute the simplest mem-
bers. The genus Riella forms the subfamily Rielloideae and in
the usual arrangement stands between the Sphaerocarpoideae and
Metzgerioideae in the family Metzgeriaceae (the Jungermanniaceae
anakrogynae of Leitgeb and Schiffner).

The geographical distribution of this strongly characterized
-genus Riella is of interest. Up to within a few months ago, the
genus was supposed, as far as the literature on the subject is con-
cerned, to be confined to the Mediterranean drainage basin of
Africa and Europe, with seven species, as commonly recognized.
One of these, however, Riella Gallic a, was reduced by M. Cor-
Ibiere in the last number of the Revue Bryologique for 1902 to
forma Gallica of the Algerian R. Battandieri. To these seven
known species, or six, as now conceived, another from a region
far removed was added by Morten P. Porsild in a recent number
of the Botanisk Tidsskrift, where R. Paulsenii from Turkestan was

Howe and Underwood: The Genus Riella 21G

described and figured. References to the principal literature dcal-
in^"- with the genus, together with the names and distribution of
the species hitherto described, are given below :

Riella Mont. Ann. Sci. Nat. III. i8 : i i. 1852. Leitgeb, Un-
tersuch. Leberm. 4 : 7A-^7- P^-7< S. f.iS. 1879. Trabut,
Rev. Gen. Bot. 3 : 449. 1891. Goebel, Flora, 77: 104-
108. pi. 2. f. /-?. 1893. Schiffn.; K. & W Nat. Pflanzcn-
fam. i': 51. 1893. Stephani, Bull. Herb. Boiss. 7:
658. 1899.

Duriaca Bory & Mont. Compt. Rend. Hebd. Acad. Sci. 16 :
I I 15. 1843 ; Ann. Sci. Nat. HI. i : 228. 1844. G. L. &
N. Syn. Hep. 593. 1846. Not Diiricita Merat, Mem.
Soc. Roy. Sci. Lille, 1827-28 : 432. 1829. Not Z)//;7-
eua Boiss. & Reut. Diag. PI. Nov. Hisp. 14. 1842.

Maisonnnivea Trevis.* Mem. R. 1st. Lomb. HI. 4: 442. 1877.

Diiridla Clauson & Billot, /<^r Schiffn., /. r.

Riella Battandieki Trab. Rev. Bryol. 13: ^5. pi. 3. 1886.

Schiffn. Bot. Centralb. 27: 240. //. /./. 5. 1886.
Algeria: Maison-Blanche near Algiers, Battandkr, Trabut.
Exsicc: Husnot, Hep. Gall. no. 173.
Forma Gallica (Bal.) Corbiere, Rev. Bryol. 29: 113. 1902.
Riella Gallica Bal.; Trab. Rev. Gen. Bot. 3: 450. //. iS. f.

G. 1891.
France: Roquehaute, near Beziers, Herault, Balansa, 1866,

Crosals, 1902 ; mare de Rigaud near Agde, Herault, Croz-

als, 1902.

Riella Cossoniana Trab.; Batt. & Trab. Atlas Fl. Alg. i : 6. //.
2. f. 1-8. 1886; Rev. Bryol. 14: 12.//. 1887.
Algeria f: El Kreider, Province Oran, Cosson, 1854; Tra-
but, 1885.
Exsicc: Husnot, Hep. Gall. no. 174.

*Trevisan rejected Riella Mont, on account of Riella Raf.; L6v.; Orb. Diet.
Univ. Hist. Nat. 8 : 488. 1849. This, however, is a nomen rtiidum and is probably
a misprint for Rimella Raf. Jour. Phys. 89 : 106. 18 19.

t Allusions to the occurrence of Riella Cossoniana in the Canaries have recently ap-^
peared in print (Bot. Tidsskrift, 24 : 327. 1902 ; Rev. Bryol. 29 : no. 1902), but
it is possible that these refer to Riella affinis, described below as new.

217 Howe and Underwood: The Genus Riella

RiELLA HELicoPHYLLA (Bory & Moiit.) Mont. Ann. Sci. Nat. III.

i8: 12. 1852; Syll. Gen. Sp. Crypt. 94. 1856.
Duriaca hclicophylla Bory & Mont. Compt. Rend. Hebd.

Acad. Sci. 16 : 1 1 16. 1843 ; Ann. Sci. Nat. III. i : 229.

1844. G. L. & N. Syn. Hep. 593. 1846. Mont.; Bory

& Durieu, Expl. Sci., Algerie, Bot. //. J^. 1846-49.
Algeria : Province Oran, Ditrku, Trabiit.
Exsicc: Husnot, Hep. Gall. no. 172.

Riella Notarisii (Mont.) Mont. Ann. Sci. Nat. III. 18: 12.

1852 ; Syll. Gen. Sp. Crypt. 94. 1856.
Sphaerocarpiis Notarisii Mont. ; De Not. Mem. R. Accad.

Torino. II. i : 343. /. d. 1-8. 1839.
Diniaea Notarisii Bory & Mont. Compt. Rend. Hebd. Acad.

Sci. 16: II 16. 1843; Ann. Sci. Nat. III. i: 229.

1844. G. L. & N. Syn. Hep. 593. 1846.
Sardinia: near Pula, De Notaris, 1834.
Greece : Phalerus, Chaboisseau, fide Stephani.

Riella Parish Gottsche ; G. & R. Hep. Eur. no. 375. 1867.
[Illust.] Leitgeb, Untersuch. Hep. 4: 76. //. 7. /. 17.
1879. Stephani, Bull. Herb. Boiss. 7: 661. 1899.
Riella ClaHSoni\^s'] Letourneux ; Husnot, Hepaticologia Gal-
lica, 87. pi. 12. f. 151. 1 88 1. Batt. & Trab. Atlas Fl.
Alg. 1:13. //. S. 1886. Trab. Rev. Gen. Bot. 3 : 452.
1891.
Algeria : Maison-Carree, etc., near Algiers, Clatison, Paris,

Trabut.
Exsicc. : G. & R. Hep. Eur. no. 375 ; Husnot, Hep. Gall,
no. 121.

Riella Paulsenii Porsild, Bot. Tidsskrift, 24: 323-327- /• ^-?-
1902.
Turkestan: Bokhara, P^w/j-^w, 1898.

Riella Reuteri Mont. Ann. Sci. Nat. III. 18 : 12. 1852. Hof-
meister, Bericht. Verh. Kgl. Sachs. Gesellsch. Wiss.
1854: 92-95. //. 4..
Switzerland: near Geneva, Renter, 1851.

Howe and Underwood: The Genus Riella 218

In describing Riella Paulscnii, the first species of the genus
known to occur outside of the Mediterranean region, Porsild ex-
presses the surmise that the distribution of the genus may prove
to be still more widely extended, which is well verified by the
known existence at the present time of three specimens of Riella
collected within the boundaries of the United States. One of these,
strangely enough, was collected by Schott as long ago as 1855,
but has remained unstudied and unnoted in literature. In April,
1902, western Texas, the region from which Schott's specimens
came, was visited by Professors F. S. Earle and S. M. Tracy, who
secured further specimens which we consider the type of the well-
marked species described below :

Riella Americana sp. no v.

Erect or ascending, 10—30 mm. high, simple or more com-
monly 1-4 times furcate : axis elliptical in section, o.2-o.(S mm.
wide, mostly 6—10 cells thick, root-hairs borne only on the basal
parts and usually few : wing 2-5 mm. broad, rounded-falciform at
apex, slightly undulate-crisped, subentire or erose, tapering toward
the base and commonly deficient below the first dichotomy ; cells
near the axis about 60 fi in greatest diameter, those near the
margin about 40/>«: scales few, small, 0.2-0.6 mm. long, lingui-
form and obtuse or irregularly lanceolate and subacute, those near
the growing apex usually intermingled with multicellular gemmae :
gemmae trichomic in origin, soon oblong or orbicular-oblong in
outline, showing later a median constriction and becoming finally
panduriform and subspatulate : dioicous : antheridia about 0.36
X 0.16 mm., sometimes as many as 75 (including empty loculi)
in a single elongated marginal series : 9 gametophyte, or each of
its branches, maturing for the most part 3-12 sporogonia in acro-
petal order : involucres smooth, ellipsoidal-ovoid or at full maturity
subglobose-ovoid, 1.4- 1.8 mm. x 0.8-1.2 mm., narrowed rather
gradually to the truncate or slightly pointed subpapillose orifice :
capsule globose, 0.8-1 mm. in diameter, seta about 0.2 mm. long,
mostly a trifle shorter than the ovoid-conic foot : spores dark-
brown, 100-130/^ in maximum diameter (spines included) ; outer
face bearing numerous sometimes curved spines 10— 24/i long,
with dilated apices, these spines more or less connected by radiat-
ing basal membranes forming irregular reticulations ; inner faces
bearing conical, non-capitate spines, 3-6// long, with basal mem-
branes obsolescent or entirely wanting. (77. // ; pi. 12, f. 21, 22).

Limpia Canon, Texas, F. S. Earle and S. M. Tracy, April 25,

1902, no. 251 ; this, the type specimen, is deposited in the her-

219 Howe and Underwood : The Genus Riella

barium of the New York Botanical Garden. The plants are said
by Professor Earle to have been found in a pool headed by a small
waterfall, forming mats on rocks and stones which were covered
by 10-60 cm. of water in the dry season. The same species was
collected by Schott in 1855, his specimen originating from
" Limpia, near its head, Western Texas." His plants, though
otherwise agreeing perfectly with those of the more recent col-
lection, are considerably smaller and are mostly unbranched. A
Riella (in herb. A. W. Evans) was collected by Mr. De Alton
Saunders at Brookings, South Dakota, in 1898. The material is
scanty and the sporogonia are so immature that the spore-char-
acters cannot be determined. The involucres and scales are
similar to those of R. Americana and it is quite probable that the
species is the same. This South Dakota specimen marks the
northern limit in the known distribution of the genus.

Riella Americana is easily distinguished from any of the species
hitherto known. Its nearest ally is probably Riella Battandieri f.
Gallica, of southern France, which differs in being monoicous,"
in the narrower wing, the smaller spores, with shorter non-capi-
tate spines, etc. Gemmae have not before, to our knowledge, been
definitely described in any species of Riella, though Goebel (Flora,
77 : 105. 1893) in figuring three young stages of Riella Battan-
dieri (?) remarks in a footnote that one of them sprang from a
" Zellkorper " while the others came from " Zellfaden." By anal-
ogy with what we have observed in Riella Americana, it seems very
probable that his plant from the " Zellkorper " came from a brood-
body of some kind, while those from the " Zellfaden " were de-
rived from spores. The young plants oi Riella Cossoniana, repre-
sented by Trabut in his figures i and 2 (/. r.),also, with little doubt,
originated from gemmae. The gemmiform appendages figured by
Trabut on the axis of his R. Gallica are of a more doubtful nature.
These were not found by Corbiere in his recent studies of what he
believes to be the same species ; possibly they were designed to
represent remains of old involucres and sporogonia.

The gemmae of R. Americana originate on the axis as trichomic
outgrowths, each of about three cells in a single series (/. 10,
11) of which the terminal cell is usually the largest. The two

* Corbiere, Rev. Bryol. 29: ill, 113. 1902.

Howe and Underwood : The Genus Riei.la '2*20

lower cells then divide in the sagittal plane (/. //) and this
this is followed by a division of the terminal cell. The basal cells
afterward grow out and divide in such a way as to give the body
a free margin throughout its periphery, though it remains attached
to the axis for a considerable time by a single cell on its proximal
surface. F. 13-16 show stages in the subsequent development,
/./ being a view of its proximal surface, and /j, /j and 16 of its
distal, the point of attachment in each case being at b. The part
which is derived from the base of the original trichome con.sists
soon of cells which are smaller, richer in protoplasm, and capable
of more rapid division than those of the part which is derived
from the apex of the original trichome. This part of basal origin
is more regularly suborbicular in outline than the other. In
neither part can an apical cell be pointed out with any certainty.
In the earlier stages the whole body seems to be mcristcmatic ;
later a constriction appears near its middle and the formation of
new cells is then the most active in the zone of this constriction.
The larger-celled part, derived from the apex or distal end of the
initial trichome, becomes at about this time more than one cell
thick in its middle and terminal regions and shows papilliform out-
growths, the beginnings of the root-hairs. The smaller-celled
portion derived from the base or proximal end of the initial tri-
chome remains only one cell thick except in the isthmus of con-
striction where it finally, in part at least, becomes two or more
cells thick. F. 16 shows a well-developed gemma inverted as
regards its original relations to the axis of the gametophytc. The
edges of the two parts here overlap slightly in the zone of con-
striction. Later, the isthmus elongates, giving the body a some-
what panduriform or at length subspatulate outline. F. ij shows
the outline of a gemma in an older stage, and f. 18 a still later
development.

Attempts to germinate the spores of the plants collected by
Earle and Tracy were made in the autumn and winter of 1902-03.
Though the spores were to all appearances mature, the attempts
were successful in the case of only a very few spores, which were
carried a little beyond the stage represented in our /. 21 and /.
22. The germ-tubes in every case observed came out from near
the middle of the outer or more spiny face. This tube is first

221 Howe and Underwood : The Genus Riella

divided by transverse walls into a single row of several cells, three
or four of the terminal of which are then divided by median lon-
gitudinal walls. This divided terminal portion by growth and
multiplication of its cells becomes ovate or spatulate in outline.
In the latest stage that we have been able to see, one somewhat
more advanced than that represented in our f. 21, the whole body
still consists of only one layer of cells. F. 22 shows in edgewise
view the same plant that is shown in /. 21. A root-hair is sent
out from the spore soon after the emergence of the germ-tubes.
A second species of Riella, apparently new, from the Canary
Islands, has come to our attention ; this we would describe as

follows :

Riella affinis sp. nov.

Apparently erect or ascending, becoming prostrate on the sub-
sidence of the water, 6-1 5 mm. high, densely caespitose, simple or
sparingly furcate : axis flattened, 0.1-0.4 mm. wide, mostly thin
and flaccid, radicelliferous at base : wing 1-3 mm. broad, rounded
or obtusely pointed at apex and falciform, deeply lobed ; cells
near the axis 45-60 /i in diameter, those near the margin about
30// : scales usually inconspicuous, 0.2-0.5 mm. long, linguiform,
lanceolate, or linear, obtuse or acute: monoicous : antheridia
solitary or in groups of 2-7 in more or less profound marginal
sinuses: sporogonia mostly 3-10 : involucres ovoid, 8-winged,
1.4-2 mm. long, 1-1.5 mm. wide (inch wings), much contracted
and often subacute at mouth; wings 0.1-O.2 mm. broad, with
undulate-sinuate or subentire margin: capsule subglobose, 0.8-1
mm. in diameter, seta 0.1-0.2 mm. long: spores brown, 85—
\20 fi in maximum diameter (spines included); outer face densely
covered with spines 6-15// long, their apices mostly truncate, often
slightly dilated, occasionally emarginate-bifid, rarely acute, the
supporting basal membranes sometimes almost deficient but com-
monly forming a few imperfect areolae ; inner faces with mostly
truncate or obtuse spines or warts 3-6^ high, basal membranes
wanting. {PI. 12. f. 23-36)

On the bank of a reservoir, Tafira, Grand Canary, June, 1897,
O. F. Cook (no. 729). We are informed by the collector that the
specimens were found partly submerged and partly exposed on a
sloping bank.

Riella affinis is allied to the Algerian R. Cossoniana Trabut and
the central Asian R. Paulsenii Porsild, but differs from both in
being monoicous and in the much longer spines of the larger

Howe and Underwood: The Genus Riella 222

spores. The involucres are less globose and more pointed at the
apex than those of R. Cossoniana. Judging from two specimens
of R. Cossoniana collected by Professor Trabut, R. affinis has the
appearance of being a larger species with broader lamina, though
this might not appear from a comparison of the measurements
given above with those given by authors for A'. Cossoniana. Our
specimens of R. affinis are prostrate and more or less entangled
with mud and are so delicate and fragile that it is quite possible
that the measurements in the above description may fail to do full
justice to the height of the plant. It may be remarked that in
occasional capsules the spores though showing a well-developed
brown coloration have short spines or papillae much like those
of R. Cossoniana ; such spores are always smaller than is normal
for the species and are probably immature or else have ripened
under abnormal conditions.

Our experiments in germinating the spores of Riella affinis
have been more successful than those with the spores of Riella
Americana, though they were not begun until December, 1902,
five and a half years from the date of collection of the specimens.
More than half of the spores experimented with germinated in a
few days by being placed on a piece of wet filter-paper in a glass
dish kept in a moist chamber at ordinary living-room tempera-
tures. The germ-tube in practically all cases emerges, as in R.
Americana, from the outer or more spiny face of the spore, usually
near its middle. The root-hair follows a little later, its lumen re-
maining continuous with that of the germ-tube. The length of the
germ-tube varies exceedingly. Finally, there appears in it a some-
what curved transverse wall with its convexity turned toward the
spore. The part above contains most of the starch grains and in
the course of time begins to show chlorophyll, cell-divisions mean-
while taking place as described above for R. Americana. The
length of the germ-tube from the spore-wall to the curved septum
has been observed to vary in different cases from 0.02-0.7 mm.
One or two root-hairs, in addition to the one which comes from
the base of the germ-tube, may spring out later from some part of
the filamentous stalk of the young gametophyte. The forms as-
sumed by the young gametophytes are extremely varied and are
doubtless determined to a considerable extent by the conditions of

223 Howe and Underwood: The Genus Riella

illumination, though no exact investigations were undertaken to
demonstrate the influence of light in this matter. The prevailing
form, however, while still unistratose, is spatulate, though flabel-
liform and linear outlines are not uncommon and grotesquely
lobed and branched stages are often met with. Linear or ribbon-
shaped forms, like that shown in /. j6, are especially common
when the young plants are from the first well covered with water.
Just below the chief growing point c, in the stage illustrated by
/. J4, there are cell divisions in the plane of the paper, which
possibly foreshadow the axis or stem, though they may have
arisen in this case through changed relations to the light. We
have not yet been able to follow the development much beyond
this point, but we have grounds for believing that the subsequent
history is essentially as described by Goebel (/. c.) for young plants
of A. Battandicri with the exception that the wing is probably
more lobed than in that species. In most cases, the development
is confined to practically a single plane, though occasionally, as in
the case represented by /. jj, the wing-lobes show an irregu-
larly spiral arrangement, due perhaps to changed positions in re-
gard to the light.

Kxplaualiou of Plates

[The drawings have been prepared by M. A. Howe, chiefly with the aid of a
camera lucida.]

Plate ii. Riella Americana Howe & Underw.

1 and 4. $ gametophytes, natural size.

2 and 3. 9 gametophytes, natural size.

5. Terminal portion of a branch of the $ gametophyte, with a young branch at

the apex, X H-

6. Terminal portion of a branch of the 9 gametophyte, X l*-

7. An involucre, with mature sporogonium, X ^6.

8. A scale, X 55-

9. Portion of axis near the apex, showing gemmae and scales, X 40- The
gemmae are shown at a and below ; the scales are, in this case, nearer the wing.

10-18. Stages in the development of the gemmae. See text, pp. 219 and 220.

10. A trichome destined to become a gemma, in distal (outer) aspect, X 193-

11. A similar stage in lateral view, X I93-

12-15. Later stages ; 12, 13, and 15 in distal aspect, 14 in proximal, X ^93 5 ^^
point of attachment.

16. A more advanced stage, inverted as regards its original relations to the axis

and as regards the preceding figures, X 193-

17. Outline of a gemma in a later stage, X 55- •

18. A young gametophyte derived from a gemma, X 28.

Howe and Underwood: The Genus Riella 224

19 and 20. Spores, outer and inner faces, X 305-

Figures 4, 18-20, are drawn from material collected by Schott in 1855 ; the others
from material collected by Earle and Tracy, April, 1902.

I'l.AlF. 12

Figs. 21 and 22, Kiclla Aincricaiia Howe & Underw. ; Figs. 23-36, Kiclla afj'niis
Howe & Underw.

21. Germ plant, X 245.

22. Same in edgewise view, X245.

23. Gametophyte of AV<7/(/ ^^;/«, X II-

24. Apical portion of another, with involucres and antheridial locuU, /^ 11.
25 and 26. Involucres, X ^^•

27 and 28. Spores, outer and inner faces, respectively, X ZIP-

29 and 30. Germinating spores, X 55-

31. Germ plant, showing the earlier cell divisions, X 245.

32-36. Various forms assumed by the germ plants, X 55- The principal vegeta-
tive point is indicated by c. The elongate form represented in Fig. 36 is common
when the young plants are well covered with water.

Bull. Torrey CLun, 30.

Pl.ii

>'' //

RIELLA AMERICANA Howe & Underw.

thin,, 't'tmimf i'f*m, i^.

rtrtt

/

%%

,'/

ffC

K1

/

¥4

v

■ O f

«•

/•■

.'•'//A

/^■••v.

Z1

w^

JV,/'. KIKI.f.A AKKIN'IH H«w« At L'ntterw,

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 35.

THE POLYPORACEAE OF NORTH AMERICA -
III. THE GENUS FOMES.

By WILLIAM ALPHONSO MURRILL

NEW YORK
1903

[Keprinted from the Bulletin of the Torkey Botanical Club, 30: 225-232. Ap 1903]

The Polyporaceae of North America.— III. The Genus Pomes

r>V Wll.I.IAM AI.IIIdNSO MURRll.L

The generic name Fames is usually accredited to Fries, but this
author never employed it to designate a genus. In Novae Sym-
bolae, 46, 185 1, the genus Polyporus is subdivided into Rupoly-
poriis and Fomes, the latter division being described as follows :

" Pileus primitus lignoso-induratus (raro molles, plorantes), contextu floccoso inter-
texto, crusta rigida obductus, azonus, sed demum concentrice sulcatus. Fungi peren-
nes {vulgo straiose), reviviscetites, seJ stratum annuttm vegeliiin tatttum est.'^

In listing species Fries wrote the word Fomcs in parenthesis
after Polyporus.

Gillet (Champ. France, I : 682. 1878) raised Fomes to generic
rank and described the group as follows :

*' Hymenium poreux ; pores non separables entr'eux, distincts du chapeau, dis-
poses par couches stratifiees. Esp^ces ligneuses-subereuses, stipit^es, dimidiees ou re-
supinees, recouvertes d'une croute dure, epaisse et st. marquee de zones annuelles
separees par des sillons concentriques. Champignons vivaces a accroissement in-
defini."

Under this genus Gillet lists thirteen species, the first being F.
marginatus, which is therefore considered its type. Most of the
remaining species of Fomes, as the term is used in Saccardo's Syl-
loge, were assigned to this category by Cooke in Grevillea (v. 1 3
and 14, 1884-85). Karsten was the first to divide generically this
large group of perennials, which now comprises a half dozen or
more genera, the name Fomes being connected with the natural
subdivision to which F. margimiUis belongs. Karsten's name
Fomitopsis, applied to this subdivision in 1881, unfortunately be-
comes a synonym of Fomes. His names Ganoderma and Elfvin-
gia, however, hold for other important subdivisions.

Synopsis of tbe :Kortli American Species

1. Context white or yellowish. 2.
Context flesh-colored, pileus often effused or effused-reflexed, flesh-colored, soon

blackening, tubes round, 3-4 to a mm. I. /'. roseus.

2. Pileus over 3 cm. broad. 3*
Pileus less than 3 cm. broad. 8-

3. Pileus encrusted, surface darker than the context. 4-
Pileus rarely encrusted, surface concolorous with the context. 9.

225

226 MuRRiLL : Polyporaceae of North America

4. Pileus thick, sulcate, ungulate, rarely applanate. 5-
Pileus at first thin, triangular or lobed, shining reddish-chestnut above, becoming

dimidiate, ungulate at the center, and dull brown: spores ovoid, 3.5 X 5 i"-

Pileus attached by a lateral stipe which soon disappears. 7. F. stipitatiis.

Pileus thin, distinctly zonate, irregular or applanate, crust brown to black : spores

ovoid, hyaline, 6 X 4("- 2. F. amiosus.

5. Surface not soon rimose, older pores not visible. 6.
Surface soon becoming rimose, deeply sulcate, older pores visible in the upper pro-
jecting annual layers: pileus exactly ungulate, found only on Shepheniia.

4. F. Ellisianus.

6. Pores 4-5 to a mm. 7-
Pores 2-3 to a mm. : pileus subtriangular, gray to black, context white to pale

cinnamon: spores ellipsoidal, 7-8X6- 7/^ ; abundant on Fraxinus.

5. F. fraxinophilus.

7. Pileus applanate, many times sulcate, horny encrusted, extremely hard in sub-

stance ; tropical. 6. F. lignens.

Pileus ungulate, applanate when very large, deeply annually sulcate, surface often

resinous, bay or black in color ; abundant on coniferous trees in temperate regions.

3. F. nngidatus.

8. Pileus ungulate, becoming black only at the base, zonate and concentrically sulcate

in age, tubes over 0.2 cm. long. 8. F. Ohiensis.

Pileus scutellate, uniformly black when quite young, tubes less than 0.2 cm.

long, context thinner than tube-layer. 9. F. scutellatus.

9. Pileus cylindrical, tubes long, visible at edges of older strata, context friable, be-

coming bitter ; growing on conifers. 10. F. Laricis.

Not as above. lO-

10. Pores large, 2 to a mm., becoming reddish when bruised, annual strata separated

by thick layers of context ; tropical. 13. F. riibritinctus.

Pores small, 5-7 to a mm., not changing to reddish when bruised. II.

11. Tubes less than 2 mm. long eixch season, context punky, hymenium glistening,

not becoming dark in color. II. F. populimis.

Tubes more than 2 mm. long each season, context hard and rather friable, hyme-
nium becoming smoky or brownish, cracking in age. 12. F. Mehae.

I. Fomes roseus (Alb. & Schw.) Cooke

Boletus roseus Alb. & Schw. Consp. Fung. 251. 1805.

Poly poms roseus Fr. Syst. i : 372. 1821.

Polyporus carneus Nees, Nov. Act. Nat. Curios. 13 : //. j.
1827.

Fomitopsis rosea Karst. Rev. Myc. 3: 18. 1881.

Fojues 7'oseus Cooke, Grevillea, 14: 19. 1885.

Fojues carneus Cooke, Grevillea, 14: 21. 1885.

This fungus is very widely distributed and does serious injury
to various coniferous trees both in Europe and America. Its
form varies greatly with the character of the host plant. Besides
the names mentioned above, P. rufo-pallidus Trog (Flora, 15 : 556.

Murrii.l: Polypokaceae of North Amkkica 227

1832. Fries, Icon. //. 1S6. f. i) doubtless refers to the same
plant and I have a suspicion that Polystictiis arcticus Fr. is also a
synonym, since plants at Kew collected near Carlton in May,

1858, are only grayish zonate forms of F. roseiis. The actual
types of P. arcticus I have not seen.

A few of the collections examined are here listed : Finland
(Karsten), Canada (Macoun, Dearness), Maine (Miss White), New-
foundland (Waghorne ?), New Hampshire (Minns, Underwood),
Ohio (James), New Jersey (Flllis), Colorado (liethel, Pammell),
West Virginia (Nuttall), Washington (Piper), Florida (Calkins),
Nebraska (Kennedy), South Carolina (Ravenel). Alabama (Peters),
California (McClatchie), Connecticut (Underwood), Georgia (Un-
derwood).

2. FoMES ANNOSUS (Fr.) Cooke

Polyporus annusus Fr. Syst. i : iJi. 1821. Icon. //. 186.

f.2.

Polyporus scrpcutarius Pers, Myc. Fur. 2 : 82. 1825.

Polyporus subpilcatus Wcmm. Fl. Ross. 332. 1836.

Polyporus resinosus Rostk.; Sturm, Deutsch, P^l. 4: 61-62. //.
2g. 1838.

Trametes radicipcrda R. Hartig, W'icht. Krankh. Waldb. 62.
1874.

Foviitopsis annosaY^2.x^t. Rev. Myc. 3: 18. 1881.

Foines annosus QookQ, Grevillea, 14: 20. 1885.

Hcterobasidion annosuni Bref Untersuch. 8: 154. //. p. 1889.

Polyporus irregularis Underw. Bull. Torrey Club, 24: 85.
1897.

Kreiger 121, Romell 13, Sydow 3107, 1108, 504, Allescher
& Schnabl 137, 432, Roumeguere 31 11, 7248, Rabenhorst 405,
de Thiimen 106, Kunze i, Briosi 2, Cavara 324. Sweden (Mur-
rill), Tyrol (Bresadola), California (Harkness), Oregon (McBride),
W^est Virginia (Nuttall), Pennsylvania (Gentry), Alabama (Earle,
Baker, Woodward), Florida (Gory).

Many other specimens have been examined, but the above-
mentioned are sufficient to show that the plant is of wide distribu-
tion. It is particularly injurious to various species of coniferous
trees.

228 MuRRiLL : Polyporaceae of North America

^ 3. Fomes ungulatus (Schaeff.) Sacc.

Boletus igniar'ms Scop. Fl. Carn. ed. 2. 2 : 469. 1772. Not B.
igniarius L. 1753.
J Boletus ungulatus Schaeff. Fung. Bav. 4 : 88. //. 137, 138.

^774-

Boletus fulvus Schaeff. Fung. Bav. 4: 89. pi. 262. 1774.

Boletus semiovatus SchdiQ.^. Fung. Bav. 4: g2. pi. 270. 1774.

Boletus marginatus Pers. Obs. 2 : 6-y. 1 799.

Boletus pinicola Sw. Vet. Akad. Handl. 88. 18 10.

Polyporus marghiatusYx. Syst. i: 372. 1821.

Polyporus pinicola Fr. Syst. i: 372. 1821.

Fames inarginalus Gill. Champ. Fr. I : 683. 1878.

Fountopsis pinicola Karst. Rev. Myc. 3: 18. 188 1.

/^^/;zf.y /'zV/w/rt Cooke, Grevillea, 14 : 17. 1885.

Fames ungulatus Sacc. Sylloge, 6: 167. 1888.

A large and widespread species growing abundantly on conif-
erous trees, such as the hemlock and pine, and found more rarely
upon deciduous trees standing near its usual hosts. Beech, elm,
maple and birch are known to have been attacked by it in Ameri-
can and European forests where conifers abound. Fresh speci-
mens collected on fir trees in Sweden correspond in all respects
with Schaeffer's B. ungulatus, so I do not hesitate to use his name.
Collections showing the range of this plant are as follows : Saxony
(Krieger 13), Sweden (Romell 116, Murrill), Siberia (de Thumen
1906), France (Roumeguere 2205), Bohemia (de Thumen 814),
Hungary (Linhart 250, 446), Canada (Macoun), New Hampshire
(Minns), New York (Atkinson, Underwood, Murrill), Colorado
(Crandall, Cockerell), West Virginia (Nuttall), Michigan (Wood),
Maine (Ricker), Virginia (Murrill).

4. Fomes Ellisianus Anders.
Fomes Ellisianus Anderson, Bot. Gaz. 16 : 113. //. 12. 1891.
Fomes circumstans Morgan, Jour. Cincinnati Soc. Nat. Hist.

18: 37. pi. i.f. 4. 1895.

This species was described from plants collected by Anderson
in Montana during the summer of 1889. Its only host thus far
reported is Sliepherdia argentca Nutt., on the living trunks and
branches of which it is very abundant in the Teton Valley,

Murrill: Polyporaceae of North America 229

Morgan's description is based on a collection in Dakota by T. A.
Williams from the same host. His name refers to the half-en-
circling habit of the plant when it grows upon a branch. The
fundus has also been collected in Colorado.

5. Fomes fraxinophilus (Peck) Sacc.
Polvponis fmxinopliilus Peck, N. Y. State Mus. Rept. 35 : 136.

1882.

FoiNCS fraxinop/iil lis Sdicc. SyWogQ, d : 172. 1888.

Ohio (Morgan, Gentry), Kentucky (Kellerman, Price), Mis-
souri (Uemetrio), Kansas (Bartholomew), Nebraska (Bates), Penn-
sylvania (Martindale), Iowa (McBride), Indiana (Underwood), New
York (Peck).

This species is confined to the ash, on which it is very com-
mon. It has been confused with P. fraxincns (Bull.) Fr. by Cooke
and some other mycologists. Anderson's no. 537 in the Parasitic
Fungi of Montana was distributed under the name oi P. fraxino-
pliilus before it was described as Fomcs Ellisianiis.

6. P'^omes ligneus (Berk.) Cooke
Poly poms ligneus Berk. Ann. Nat. Hist. 3: 387. 1839.
Fomes ligneus Cooke, Grevillea, 13: 119. 1884.
Excellent specimens were collected in Nicaragua by C. L.
Smith and in St. Kitts by Britton and Cowell. Plants from Mau-
ritius at Kew bearing this name are specifically distinct from our
fungus. Fomes sulcatus Cooke from Demarara is nearly allied to

F. lis'neus.

7. Fomes stipitatus sp. nov,

A plant of medium size with short evanescent lateral stipe,
shining reddish-chestnut crust and broad white thin margin.
Pileus exceedingly hard and horny, flat and triangular or some-
what circular when young, becoming dimidiate with ungulate
center and thin spreading margin, which is usually lobed and often
deeply cleft ; hymcnium normally plane, but frequently convex
because of the upturned margin, 5 x 10x3 cm.; surface glabrous,
varnished, reddish-chestnut, with numerous shallow concentric
furrows, becoming shining black and at length dull smoky-brown ;
margin broad, flat, glabrous, obtuse, 3 mm. thick, crenate to cleft,
sterile portion 0.5 cm. wide, white: context hard, woody, 0. 5-1
cm. thick, pale ochraceous, with strands of dark-brown shining

230 MURKILL : POLYPORACEAE OF NORTH AMERICA

horny substance extending from the stipe to the margin in a tree-
like fashion : tubes plainly though unevenly stratified, 2-4 mm.
long each season, 4-5 to a mm., grayish-brown within, mouths
circular, white or very pale yellow, dissepiments thick, obtuse,
entire : spores broadly ovoid with attenuate base, light yellowish-
brown, smooth, 3.5 X ^iJ.: stipe lateral, i x i cm., cylindrical,
equal, glabrous, shining black to dull brown, with substance like
the context, apparently absorbed or overgrown as the pileus
enlarges.

This very distinct and easily recognized species was collected
by C. L. Smith and B. Shimek in Nicaragua during the winter of
1891-92. Its conspicuous varnished surface and lateral stipe sug-
gest at once the genus Ganoderma, but more important characters
connect it with Fonics and I have with some hesitation assigned it
to this genus, recalling the fact that Fomes ungulatus often has
considerable varnish on its surface and that an umbo often has its
beginning in a reduced stipe.

8. Fomes Ohiensis (Berk.)
Trametcs Ohiensis Berk. Grevillea, i : 66. 1872.
This plant is unusually small for the genus Fomes, making
with F. sciitellatiis a rather distinct subgroup. It grows upon rails
and dead branches of various deciduous trees. Among collections
examined are those from Canada (Dearness), Kansas (Cragin),
Michigan (Dennen), Ohio (James), New York (Peck).

9. Fomes scutellatus (Schw.) Cooke

Polyponis scutellatus Schw. Trans. Am. Phil. Soc. 4: 157.
1832.

Fomes sc2itellatus Cookt, Grevillea, 14: 19. 1885.

This minute species was described by Schweinitz from material
collected on Syringa in Pennsylvania. It has since been found
on alder, witch hazel, sweet gum and other deciduous shrubs and
trees. Some of the collections examined are as follows : Penn-
sylvania (Everhart), Florida (Calkins), Maine (Harvey, Blake),
Canada (Macoun), Delaware (Commons), Ohio (James), Alabama
(Mell).

10. Fomes Laricis (Jacq.)

Boletus Laricis Jacq. Miscel. i : 164-203. pi. 20-21. \J'J^.
Bull. Herb. France, pi. 2g6. 1786.

Murrill: Polyporaceae of North America 231

B. officinalis \\\\a.YS, Delph. 1041. 1786.

B. purgans Pens. Syn. 531. 1801.

Polyponis officinalis Fr. Syst. I : 365. 1821.

This fungus has becMi known from ancient times on account of
its medicinal properties and is still collected in considerable quan-
tities in the larch forests of Kurope and Asia for use in medicine.
Single specimens are sometimes found weighing as much as fifteen
pounds in the dry state. Many of the older botanists mention this
plant. Hauhin (Pinax, 375. 1623) gives a good general descrip-
tion of it under the name "Agancion, sive fungtts laricis." Micheli
figures it (Gen. />/. 61. f. i. 1729). Even Dioscorides knew its
value. According to Miss Southworth, the substance of the fun-
gus consists mostly of resin-granules about knots of mycelium,
containing at times one or more curiously shaped bodies resem-
bling branching bast cells, which grow out from the mycelium.
These resin-granules contain the medicinal properties.

The European host of this fungus is the living larch. In
America, it has been found also on pine and spruce. Specimens
found in Michigan in 1886 on living white pine were used by
Calkins and others instead of quinine. MacDougal collected a
very handsome specimen on dead spruce in Montana in July, 1901.
Macoun also found it on spruce in British Columbia in April,
1889. A recent note in Science from Professor Bessey refers to
specimens received by him from Montana and the Yellowstone
Park, collected on undetermined species of conifers. Calkins' re-
port of this species in America a dozen or more years ago evi-
dently escaped Professor Bessey's attention.

II. Pomes populinus (Schum.) Cooke

Boletus popidiiius Schum. Enum. PI. Saell. 2: 384. 1803.

Poria obdiicens Pers. Myc. Eur. 2: 104. 1825.

Polfponis connatiis \Wt\nm. Fl. Ross. 332. 1836. En. Stirp.
Petrop. 208. 1837.

Polvponts connat2is Fr. Epicr. 472. 1836-1838. Icon. //.
185./. 2.

Fomes connatus Gill. Champ. P^rance, i : 684. 1878.

Fomcs popiilitius Cooke, Grevillea, 14: 20. 1885.

Finland (Karsten), Bavaria (de Thiimen), Germany (Sydow),
England (Massee), Sweden (Murrill), Canada (Dearness, Macoun),

232 Murrill: Polyporaceae of North America

Georgia (Ellis), Ohio (James, Morgan), Delaware (Commons),
New York (Earle, Underwood), Massachusetts (Underwood),
Pennsylvania (Herbst), Florida (Martin), New Jersey (Ellis), Mis-
souri (Glatfelter).

The favorite host of this species is the maple, though it is also
found on poplar, linden and a few other deciduous trees. Living
Norway maples in Sweden were found to be seriously injured by
its attacks, and I have noticed it on this and other species of
maple used as shade-trees in American cities.

12. Fomes Meliae (Underw.)

Polyporits Meliae Underw. Bull. Torrey Club, 24: 85. 1897.

Described from specimens collected by Underwood on branches

of Melia Asederach in 1895. It also occurs on Gleditsia and

Fraxiniis, as is indicated by the recent Alabama collections of

Earle and Baker.

13. Fomes rubritinctus sp. nov.

A large perennial plant of many layers, the upper partly dead
and discolored, the lower smooth, light-colored and anoderm with
large pores, which become dark red when bruised. Pileus corky
to woody, dimidiate, convex, 4-12 x 5-15 X 3-10 cm.; surface
anoderm, smooth, velvety to the touch, white or pallid, becoming
dark brown and roughened with age ; margin obtuse, sterile, con-
colorous : context corky, becoming woody, pallid, 1.5 cm. thick :
tubes unevenly stratified, the layers separated by unusually thick
cushions of context, 0.5-1.0 cm. long each season, 2 to a mm.,
isabelline within, becoming umbrinous in the older layers ; mouths
subcircular or polygonal, edges thin, isabelline, becoming dark
red when bruised : spores ovoid to globose, 5-7 fi long, smooth,
hyaline, wall of medium thickness, hyphae hyaline.

Collected by C. L. Smith in Nicaragua and distributed as no.
138 of his Central American Fungi.

Species Inquirendae
Fomes geotropus Cooke, Grevillea, 13: 119. 1^2,4. {Polyponis
geotroptcs Cooke, GrQvxWe^i, 1^: 32. 1884). This species is allied
to Fomes ulmarius Fr. It was described from collections in Dem-
erara, but seems to range northward to Mexico, Florida and Ala-
bama. Other doubtful or unknown species are : Fomes pcrpusilliis
(Pers.) Cooke, F. extensiis (Lev.) Cooke, F. aegerita (Fr.) Cooke,

F. microporus (Sw.) Cooke, and F. albogriseus Peck,
New York City.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 30.

STUDIES ON THE ROCKY MOUNTAIN FLORA.-X.

By per axel RYDBERG

NEW YORK
1903

[Reprinted from the Bulletin of the Torkey Botanical Club, 30: 247-262. Ap 1903]

Studies on the Rocky Mountain Flora.— X

By Pkr Axel Rydherg

Corispermum marginale sp. nov.

Annual : stem glabrous, much branched, 2-5 dm. high : leaves
narrowly linear, 2-5 cm. long, 1.5-2 mm. wide: .spike 3-8 cm.
long, rather dense ; bracts usually overlapping each other, the
lower lanceolate, about i cm. long, the upper ovate, 5 mm. long ;
all more or less acuminate, with a strong midrib and conspicuous
scarious margins, slightly pubescent when young or glabrate :
fruit about 4 mm. long and 2.5 mm. wide, with a broad wing
margin.

This is perhaps most nearly related to C. Jiyssopifoliiini of the
salt marshes of the coast regions, but is taller, more slender, with
less fleshy leaves, smaller fruit and more conspicuously scari-
ous-marfrined bracts. From C. nitidinn it differs in the broader
bracts, denser spikes and larger fruit. It grows in similar localities,
viz., in sandy soil, in draws and ravines.

New Mexico: Albuquerque, 1884, C. L. Herrick (type in
herb. N. Y. Bot. Garden).

Colorado : Huerfano Valley, near Gardner, 1900, F. K. Vree-
land, 6^"/ ; Rocky Ford, 1893, Crandall ; Denver, 1887, 5. M.
Tracy.

Chenopodium Watsoni a. Nelson, Bot. Gaz. 34 : 362. 1902
Chcnopodiiun olidiim S. Wats. Proc. Am. Acad. 9 : 95- 1874,
mainly. Not Curt.

Annual, rather stout, sparingly farinose, 2-6 dm. high : stem
branched, striate and sometimes tinged with red : leaves rather
thick ; petioles 2- 10 mm. long ; blades 1-2 cm. long, oval, ovate
or oblong, obtuse and occasionally cuspidate-mucronate, dark
green and sparingly mealy : inflorescence narrow, dense : flowers
I mm. in diameter: seed 1.2 5- 1.5 mm. in diameter: pericarp
mealy and closely adherent.

Watson's name being preoccupied, I apply Nelson's substitute
to this species as it is Watson's C. olidinn as to the description
and most specimens cited. Several specimens given by him be-
long, however, to the next species, which has been confused with

247

248 Rydberg : Studies on the

it. Mr. Powell sent some seeds to Elihu Hall. A specimen raised
from these seeds and preserved in the herbarium of Columbia Uni-
versity I take as typical C. Watsoni.

Chenopodium Wolfii sp. nov.

Chenopodmm olidum S. Wats. Proc. Am. Acad. 9: 95- 1874,
partly.

Annual, sparingly inealy : stem erect, 3-4 dm. high, often
branched, striate : leaves moderately thick, dark green, only
slightly mealy ; petioles about i cm. long ; blades 1-3 cm long,
oblong to ovate-lanceolate, entire or rarely slightly hastately
toothed, obtuse : inflorescence dense, narrow : flowers small :
seeds less than i mm., usually .75 mm. in diameter : pericarp thin
and easily separated from the seed.

Closely resembling the preceding in habit, but is easily distin-
guished by the smaller flowers and seed and the separating
pericarp. These characters place it nearest to C. oblongifolium
Nutt., from which it differs in the less mealiness and the smaller
seeds.

Colorado: Twin Lakes, \77l, JoJin Wolf, 233, apparently

also 258 and 263.

Wyoming: Wamsutter, 1897, Aven Nelson, j6yi.

Endolepis ovata sp. nov.

Low, annual, usually less than i dm. high: stem branched,
straw-colored or pinkish, almost glabrous : leaves subsessile,
usually less than i cm. long, ovate or lance-ovate, 3 -nerved at the
base, sparingly mealy : pistillate flowers solitary in the axils of the
middle leaves : staminate flowers in small clusters in the axils of
the upper leaves or at the end of the branches, otherwise as in E.
StLckleyana Torr.

In my opinion Dr. Torrey was correct in separating Endolepis
from Atriplcx. It is interesting to find a second species of the
former genus, which hitherto has been known as monotypic. The
original Endolepis has narrowly lanceolate leaves, which are thin-
iner and without lateral ribs.

Wyoming: Buffalo, 1900, Frank 1 weedy, 3 2 go (type in herb.
N. Y. Bot. Garden); Wallace Creek, 1898, ^/zVw Nelson, 4ggg\
Rock Creek Station, 1881, Z. F. Ward.

Montana: Glendive, 1892,/. H. Sandberg.

RocKV Mountain Flora '24U

Cleomella cornuta sp. nov.

Annual : stem straw-colored, 2-3 tlm. his^li, branched below
with ascending branches, j^labrous tiirou<jhout : leaves ternate ;
petioles 1-1.5 cm. long ; leaflets 1-2 cm. long, oblong or oblong-
oblanceolate, obtuse, mucronate : inflorescence short : pedicels
very slender, almost I cm. long : sepals yellowish, ovate, cu.spi-
date, less than i mm. long : petals light yellow, narrowly oval,
clawless. about 3 mm. long : filaments about twice as long : fruit
broadly rhombic, broader than long, about 3 mm. long and 4 mm.
broad ; the corners often produced into short processes : stipe 6-
8 mm. long; beak over 1 mm.: seeds about 2.5 mm. long, 1.75
mm. wide, smooth and unmarked.

This is related to C. oocarpa and C. plocaspcnua, but has
broader leaflets. From the former it also differs in the strongly
rhombic pod and from the latter in the broader and unmarked
seeds. The type grew at an altitude of 1350 m.

Utah : Cainsville, 1894, Marcus E. Jones, ^^5^ (*^ypc •" U. S.
Nat. Herb.).

Cerastium Earlei sp. nov.

Perennial with a slender branched and stoloniferous root-
stock : stems weak, ascending, 2-3 dm. high, viscid-puberulent,
branched : leaves oblong or oblanceolate, obtuse, viscid-puber-
ulent, 1.5-3 cm. long, 3-8 mm. wide : inflorescence open ;
bracts ovate or ovate-lanceolate, not scarious ; pedicels 2-2.5 cm.
long : sepals lanceolate, acute, sparingly pubescent, scarious on
the margins, about 5 mm. long : petals about i cm. long or more,
fully twice as long as the sepals, not very deeply cleft.

The type number was determined by M. E. Jones as Cirastuim
alpinuui near var. glabraUini, and the other two numbers of the

Baker, Earle and Tracy collection are labeled Cerastium ,

and Cerastium ai-vense ohlongifolium, respectivel)\ The six sheets
(two of each number) in the N. Y. Botanical Garden herbarium
(except one of number 621 , which represents a luxuriant state)
are so alike that it is impossible to refer them to different species.
In the size of the flowers, form of the leaves and general habit,
the species resembles most C. alpinum, but the pubescence is
different : in C. alpinum long-villous and less viscid, in C. harlei
very short and very viscid. The latter grows at an altitude of
2700-3600 m.

Colorado : Near La Plata P. O., 1898, Baker, Earle & Trarr,

250 Rydberg : Studies on the

^7^ (type in Herb. N. Y. Bot. Garden) ; Little Kate Basin, La
Plata Mts., §68 ; Cumberland Basin, 621.

Cerastium Leibergii sp. nov.

Perennial with a slender, branched rootstock : stems simple,
4-5 dm. high, sparingly villous and viscid-puberulent ; uppermost
internode of the stem about i dm. long : leaves few and scarcely
half as long as the internodes, linear-oblong, 2.5-3 cm. long, 6-7
mm. wide, the lower obtuse, the upper acutish, i -nerved, thin,
puberulent when young, ciliate on the margins : inflorescence vis-
cid-puberulent ; bracts lanceolate, 4-8 mm. long, not scarious-
margined ; pedicels in fruit 2.5-3 cm. long : sepals about 5 mm.
long, lanceolate, viscid-pubescent, scarious-tipped and margined :
petals scarcely twice as long as the sepals.

This is perhaps most nearly related to the preceding, but differs
mainly in the broader, obtuse leaves and smaller flowers. It
grows in open Pinus ponderosa forests at an altitude of 970 m.

Idaho : Upper St. Mary's River, 1895, /. B. Leiberg, iioj
(type in U. S. Nat. Herb.).

Cerastium graminifolium sp. nov.

Perennial with slender branched rootstock : stems 3-5 dm.
high, densely short-villous and somewhat viscid, erect ; upper-
most node of the stem below the inflorescence elongated, 5-10
cm. long: leaves linear-lanceolate, 2-3 cm. long, 3-5 mm. wide,
1 -nerved, short viscid-villous, mostly spreading: bracts lanceolate
or ovate-lanceolate : pedicels in fruit 2-4 cm. long : sepals nar-
rowly lanceolate, densely viscid-pubescent, only slightly scarious-
margined, acute, 5-6 mm. long: petals 12-14 mm. long, cleft
%-% their length into oblong, obtusish lobes : capsule almost
straight, about ^ longer than the sepals.

This has gone under the name of C. arvensc, but differs from
the European plant, in the larger flowers, the more villous pubes-
cence and the more acute leaves, which resemble much those of
Alsinc graniinca. It grows in rich bottom lands.

Washington : Pullman, 1897, Elmer, lyj (type in herb. N. Y.
Bot. Garden).

Idaho: near Lewiston, 1896, A. A. & E. Gertrude Heller,
301 J ; Upper Ferry, Clearwater Ferry, 1892, Sandberg, Mac-
Dougal & Heller, ^g.

RocKV Mountain Flora 251

Draba sobolifera sp. nov,

A low cespitose perennial, but the leafy stolons often some-
what elongated : basal leaves oblanceolatc, obtuse or spatulate,
1-2 cm. long and 3-4 mm. wide, sparingly stellate-villous, thick
with indistinct midrib : peduncles 3-6 cm. long, stout, rather
many-llowered : petals )'ellow, obovate, about 4 mm. long, more
than twice as long as the rounded sepals : pedicels in fruit 7-8
mm. long: pod ovate, about 6 mm. long, 3—4 mm. wide, finely
stellate : style about 5 mm. long.

This species is related to l\ vciitrosa and D. alpiiia. From
the former it differs in the longer leaves and the finer and sparser
pubescence ; from the latter in the more elongated stems, the
shorter and stellate pod and the shorter styles. It grows at an
altitude of about 2500 m.

Utah : Marysvale, above timber line, 1894, Marcus E. Jones,
jgj6 (type), j8gj;c and jSpja.:; (all in U. S. Nat. Herb.).

Draba argyrea sp. nov.

A cespitose tufted perennial, with underground horizontal
stems ; basal leaves obovate, ^-y mm. long, in dense rosettes,
densely and finely stellate, almost silvery ; midvein indistinct : pe-
duncles slender, 3—5 cm. long, naked, few-flowered : petals yellow,
3 mm. long, obovate, about twice as long as the ovate sepals : pedi-
cels in fruit 5—8 mm. long: pods lanceolate, 6-y mm. long, 2.5—
3 mm. wide, minutely stellate : style about i mm. long.

This species resembles somewhat D. oligospcrnia in general
habit and flowers, but the leaves are shorter and broader, more
decidedly stellate and without the strong midrib, and the pod is
longer. It also reminds one of D. vetitrosa but is a more delicate
plant with finer pubescence and longer pods. The type grew at an
altitude of about 3000 m.

Idaho : Rock crevices. Sawtooth Mountains, head of Pettit

Lake, July, 1895, L. F. Henderson, 353S (type in U. S. Nat.

Herb.).

Draba uncinalis sp. nov.

Small cespitose and pulvinate perennial : flowering stems scapi-
form, 2-3 cm. high, sparingly stellate, leafless ; basal leaves in
dense tufts, thick, rounded obovate, 3—4 mm. long, sparingly stellate,
with indistinct midrib : inflorescence corymbiform, few-flowered ;
pedicels in fruit 4-5 mm. long : pod ovate, 3-4 mm. long and 3
mm. wide, glabrous or nearly so : style about 0.5 mm. long.

252 Rydberg : Studies on the

A small plant characterized by the short pod, found other-
wise only in two North American species, viz., D. andina and D.
oligosperma ; but in both of those species the leaves have strong
midribs and are less stellate. In habit it resembles mostZ?. nivalis,
but that species has a different pod. It grows at an altitude of
3500 m.

Utah : Marysvale, Tate Mine, 1 894, M. E. Jones, ^g^oam
(type in U. S. Nat. Herb.).

Lesquerella Utahensis sp. nov.

A tufted perennial : stems ascending, 5-10 cm. (seldom 15 cm.)
high ; basal leaves 3-4 cm. long, petioled ; blades oval or obovate,
densely and finely stellate ; stem-leaves smaller, oblanceolate or
spatulate : petals yellow, about 8 mm. long, with cuneate blades,
J4, longer than the oblong sepals : pedicels in fruit ascending, 5-
10 mm. long : pod almost spherical or slightly transversely flat-
tened, 3-4 mm. wide, finely stellate : style 4-5 mm. long, much
longer than the pod.

This species is related to L. Wardii, i. c, it has the same habit
and pubescence and the pod is as in that species somewhat thicker
than broad ; but the latter is shorter, rounder, neither elongated
nor acute at the apex. L. Utahensis grows at an altitude of 2000-
3500 m.

Utah : American Fork Canon, 1880, Marcus E. Jones, ijj4-
(type in herb. N. Y. Bot. Garden); Friser, 1880, 1810 ; Marys-
vale, 1884, 5J75^, 5p5c?;/ ; Irelands Ranch, Salina Canon, 5^^/;
Canon above Tropic, jji2d.

Stanleya runcinata sp. nov.

Stem stout, glabrous : basal leaves about 2 dm. long, lanceo-
late in outline, runcinately divided, softly pubescent with long white
hairs ; lobes broadly triangular or the lowest ones oblong : stem-
leaves ovate to oblanceolate, more glabrous : raceme 1-3 dm. long,
many-flowered : sepals linear, about i cm. long and i mm. wide,
light yellow, glabrous : petals about _i^ longer ; blades oblanceo-
late, about ^ as long as the glabrous claws : filaments glabrous,
about twice as long as the sepals : pods almost erect and straight,
6-7 cm. long, 1.5 mm. wide, on spreading pedicels about i cm.
long.

This species is perhaps most closely related to 6". pi)inata, but
easily distinguished by the lobing of the basal leaves, the straight

RocKv Mountain Flora 253

pods, the oblanceolatc blades of the j^etals and the tj;Ial)rou.s fila-
ments.

Idaho (?): 1S96, Ilouicrson (t\'pc in U. S. Xat. Ikrb.).

Thelypodium ovalifolium sp. nov.

Perennial, perfectly (glabrous : stems terete, 3-5 dm. hipjh,
ascending: earliest basal leaves about 5 cm. lonrj, petioled with
broadly oval blades, <^labrous, entire, obtuse ; the later ones ob-
lanceolatc ; stem-leaves sai^ittate, 2 cm. lon<^, thick, sessile and
claspinf^, with short rounded basal lobes : intlorescence eloni^ated :
sepals oblon<j, 3 mm. long : petals about 5 mm. long with spatu-
late blades and slender claw : pedicels in fruit 4-5 mm. long,
spreading: pods ascending, 1—3 cm. long and about i mm. wide,
somewhat torulose.

This species is perhaf)s most nearly related to T. pauiculcilimi^
but differs in the short pods and the broad basal leaves.

Utah: Panguitch Lake, 1894, M. PI Jones, (5o/5r (type in
U. S. Nat. Herb.).

Lepidium Georginum sp. nov.

Apparently biennial, cespitose from a slender taproot, 1-2 dm.
high : stems hirsute with short spreading hairs : basal leaves 1—5
cm. long, deeply pinnately divided with ovate lobes ; stem-leaves
similar or the upper oblanceolatc and entire ; all more or less
pubescent : inflorescence short : petals spatulate, slightly exceed-
ing the sepals : pods pubescent, rounded oval, 4 mm. long,
broadly winged above, forming at the apex two broadly ovate
obtuse lobes, over 0.5 mm. long.

This species is in habit nearest L. WrigJitii \\\\\\ which it has
been confused ; but it is less hairy and the pods are those of L.
dictyotu7)i.

Utah : " Southern Utah," 1874, C. C. Parry, ig (type in herb.
Columbia University); 1877, E. Palmer, 40 ; St. George, 1880,
M. E. Jones, 16 1 4; Milford, tSji in part; 1894,5/70/.

Thysanocarpus trichocarpus sp. nov.

Annual, perfectly glabrous, except the fruit, 1-3 dm. high :
stem terete, branched : lower leaves oblanceolatc or oblong, sin-
uately dentate, thick and somew^hat glaucous ; uppermost leaves
linear or linear-lanceolate, entire : racemes often i dm. long :
petals slightly over i mm. long ; blades broadly spatulate :
pedicels in fruit about 5 mm. long, recurved : pod nearly orbicular,

254 Rydberg : Studies on the

about 4 mm. wide, short-pubescent: wing-margins crenate or
lobed, not fenestrate : style scarcely exceeding the wing-margin.
Utah : Silver Reef, 1 894, M. E. Jones, 5163b, in part (type
in U. S. Nat. Herb.), 5i4gd and J/Jp^/.

Trifolium lividum sp. no v.

Perennial, densely cespitose and tufted : branches of the caudex
short and covered by the remnants of the leaf-bases and scarious
stipules ; free portion of the latter subulate-filiform : petioles 4-1 5
cm. long, slender, glabrous ; leaflets linear-lanceolate or oblanceo-
late, very acute at both ends or cuspidate at the apex, 3-4 cm.
long, glabrous above, sparingly appressed hairy beneath, green,
entire-margined: peduncles 1-1.5 dm. high, sparingly appressed
hairy: heads nearly 2 cm. in diameter; bracts linear-subulate to
almost filiform, longer than the calyx-tube, glabrous, slightly if
at all scarious-margined : calyx-tube 2.5-3 cm. long ; teeth subu-
late-filiform, 4-5 mm. long : corolla about 12 mm. long, light
purple with a darker keel.

This species is nearest to T. dasyphylhim, but is a larger, more
glabrous and greener plant, and with much larger and narrower
bracts. It grows at an altitude of about 3000 m.

Colorado: Graymount, 1895, P. A. Rydberg, 2442 {\.y^& in
herb. N. Y. Hot. Garden); Alpine Tunnel, 1897, C. L. Shear.

Trifolium pedunculatum sp. nov.

Perennial, with a somewhat cespitose base : stems 2-4 dm.
high, rather slender, striate, glabrous or with a few appressed
hairs : stipules 2-3 cm. long, lanceolate, i^ or ^ of their length
adnate to the petioles, denticulate : petioles 4-6 cm. long, slender ;
leaflets 3, oblanceolate to Hnear, acute, finely denticulate, 2-3.5
cm. long: peduncles i dm. or more long: heads 15-18 mm. hi
diameter, not bracteate : flowers subsessile, 10-12 mm. long, in
fruit not reflexed : calyx-tube 2 mm. long, glabrous below, pubes-
cent towards the throat ; calyx-teeth subulate, the lower 3-3.5 mm.
long : corolla purple.

This species is nearest to T. longipes, but differs in the smaller
purple flowers and the calyx which is glabrous below.

Idaho: Long Valley, 1895, L. F. Henderson, jogd (type in
U. S. Nat. Herb.); near Sawtooth, 1896, B. IV. Evermann, 3^2.

Lotus longebracteatus sp. nov.

Perennial, cespitose, decumbent : stems 2-3 dm. long, diffusely
branched, grayish strigose : leaves thick, grayish strigose with a

RocKV Mountain Flora 2o">

distinct rachis ; leaHets 4, linear to oblong or of the lowest leaves
obovate, 1-1.5 cm. long or the lower shorter: peduncles 4-5 cm.
long, i-2-flowered ; bracts linear, about equalling the calyx : calyx
strigose, its tube and teeth each about 4 mm. long; the latter
subulate: corolla about 14 mm. long, strongly curved upwards,
the banner much e.xceeding the wings and keel : pod linear,
straight. 2.5-3 ^ni. long, pubescent.

This is in some respects intermediate between L. rigidiis and
L. W'riirhtii, differing from the former in the narrower leaflets, in
the shorter peduncles, longer bracts and narrower calyx-teeth and
from the latter in decumbent habit, leaves with a distinct rachis and
elongated peduncles.

Utah : "South Utah," 1877, Dr. E. Palmer, g4 (type in U.
S. Nat. Herb.).

Lupinus depressus sp. nov.

Perennial, cespitose : stems decumbent, 1-2 dm. long, ap-
pressed silvery silky pubescent : petioles slender, 5-7 cm. long,
also silver}' silky; leaflets about 7, oblanceolate, usually obtusish,
1.5-2 cm. long, densely appressed silvery canescent ; stipules sub-
ulate : racemes short, slightly exceeding the leaves ; bracts lanceo-
late, short, deciduous : flowers 8-9 mm. long : calyx long-villous ;
lips ovate-lanceolate ; the lower about )i longer than the upper :
corolla light bluish-purple ; banner glabrous, slightly shorter than
the wings.

Being perhaps most nearly related to L. iiiininms, it differs
from that in the decumbent or spreading leafy stem and more silky
pubescence. It grows in gravel on dry mountain slopes, at an alti-
tude of about 1750 m.

Idaho: Divide between St. Joe and Clearwater River, 1895.
John B. Leiberg, 1201 (type in U. S. Nat. Herb.) ; Lo-Lo Trail,
1880, S. Watson, 80.

Lupinus Evermannii sp. nov.

Perennial and densely cespitose : the numerous stems about
1.5 dm. high, slender, erect or ascending, leafy, appressed hairy:
petioles 2—4 cm. long, also appres.sed hairy ; leaflets oblanceolate
to linear-oblanceolate, 1-2.5 cm. long, densely canescent on both
sides with long appressed hairs ; stipules lanceolate, acuminate :
raceme very short-peduncled, dense and short ; bracts lanceolate,
short, deciduous: flowers 6—7 mm. long; pedicels and cah'.x
densely pubescent with spreading hairs ; lips of the latter ovate-

256 Rydberg : Studies on the

lanceolate, the lower about >^ longer than the upper : corolla
purplish-blue ; banner with a light spot, glabrous.

This species is nearest related to L. holoscriceus, but the corolla
is lighter and the pubescence of the calyx is spreading. It is also
lower and more cespitose, resembling more L. candicans in habit,
but has much smaller flowers.

Idaho: Near Sawtooth, 1896, B. W. Evermann, 533 (type in

U. S. Nat. Herb.).

Lupinus Jonesii sp. nov.

Perennial, very stout : whole plant yellowish-green : stem over
I mm. high, densely pubescent with two kinds of hairs, viz., a
short pubescence and long spreading silky hairs ; petioles about 5
cm. long, also pubescent with spreading hairs ; leaflets 7-9, ob-
lanceolate, densely appressed, yellowish hairy on both sides, 4-7
cm. long ; stipules setaceous, deciduous : raceme long and dense,
many-flowered ; bracts subulate-filiform, much exceeding the buds ;
pedicels and calyx densely hairy with short spreading hairs : flow-
ers about 16 mm. long : lips of calyx ovate-lanceolate ; the lower
about i^ longer than the upper : corolla white ; the banner with
a faint brownish spot ; keel purple-tipped.

This species is nearest related to L. lencophylliis, but is less
densely pubescent, the flowers are larger and in the type sheet
whitish and the bracts are much longer. It grows at an altitude
of about 4200 ft.

Utah : Silver Reef, May 3, 1894, Marcus E. Jones, 31^3 (type
in U. S. Nat. Herb.).

Lupinus adscendens sp. nov.

Perennial : stem 4-6 dm. high, appressed silky, leafy : petioles
5-10 cm. long, ascending, also appressed silky ; leaflets about 7,
linear-oblanceolate, 4-5 cm. long, acute, appressed silky on both
sides but green ; stipules subulate : racemes elongated, narrow ;
pedicels strongly ascending, rather long hairy : flowers about 1 2
mm. long : calyx grayish silky ; the lower lip slightly longer than
the upper : bract linear-subulate, much longer than the buds,
almost equalling the calyx in length : corolla dark bluish-purple ;
the banner glabrous with a lighter spot.

This species is related to L. laxiflorus, but the calyx is not
spurred and the bracts are much longer.

Wyoming : Headwaters of the Tongue River, 1898, F. Tzveedy,
I2g (type in herb. N. Y. Bot. Garden); Head of Big Goose
Creek, 1893, F. Tivcedy, 13.

Rocky Mountain Flora 257

Lupinus argentinus sp. nov.

Perennial, densely appressed white-silky throuf^'hout : stem
4-6 dm. hi^di : petioles 5-7 cm. lon^^ ; stipules lanceolate-subu-
late ; leallets about 7. oblanceolate. densely .silky on both sides,
3-5 cm. lonL,^ conduplicate and with strongly curved back : in-
florescence rather short and den.sc ; bracts ovate-lanceolate, short,
deciduous : flowers about i cm. long : calyx and pedicels densely
white-silky, .spurred at the base ; lower lip lanceolate, about
one third longer than the upper : corolla bluish-j)urple ; banner
pubescent on the back, about as long as the wings : pods oblong,
densely pubescent, about 3 cm. long and 5-6-seeded.

This species is nearest to L. argophyllns, but differs in the denser
and whiter pubescence, the broader curved leaflets and the more
pubescent banner. It grows at an altitude of 1200-2200 m.

Utah : Near Reno, 1900. 5. G. Stokes (type in herb. N. Y.
Bot. Garden); Provo River, 1881, M. H. Jones, 2i6q ; Salt Lake
City, 1869, S. Watson, 22 j.

Lupinus comatus sp. nov.

Perennial, somewhat cespitose : .stem about 3 dm. high, pubes-
cent with long spreading silky hairs : petioles of the basal leaves
about I dm. long, also pubescent with .spreading hairs, leaflets
narrowly oblanceolate, acute, 2.5-4 cm. long, green, pubescent on
both sides ; stem-leaves similar, but with shorter petioles ; stipules
linear lanceolate, long-acuminate : raceme short and dense, short-
peduncled ; bracts lanceolate-subulate, equalling or exceeding the
buds ; pedicels and calyx densely pubescent with spreading hairs ;
lips almost equal, ovate-lanceolate: flowers 8-10 mm. long:
corolla bluish-purple, rarely white ; banner slightly shorter than
the broad wings.

The species is somewhat related to L. laxiflonis but the calyx
is not spurred, the raceme short and dense, the flowers larger and
the pubescence of the stem and petioles long and .spreading.

Colorado: Lake City, 1878, F. N. Pm.Jt' (type in herb. Co-
lumbia University); Gunnison Co., 1895, C. L. Pol/ard {}) (U. S.

Nat. Herb.).

Lupinus maculatus sp. nov.

Perennial, strict : stem 4-6 dm. high, finely appressed puberu-
lent : petioles short, 2-4 cm. long, almost glabrous ; leaflets 5-7,
those of the lower leaves obovate-oblanceolate, mucronate ; those
of the upper oblanceolate and acute, sparingly appressed hairy
beneath, glabrous above, 3-5 cm. long : inflorescence short-pedun-

e

258 RvDBERG : Studies on the

cled and dense ; bracts ovate-lanceolate, acuminate, deciduous ;
the short pedicels and the calyx densely pubescent with short
spreading hairs ; lips of the calyx ovate-lanceolate, the lower
slightly longer than the upper: flowers 10-12 mm. long; banner
light purple with a dark spot, glabrous, much shorter than the
white or purple-tinged wings ; keel white with purple tip : pods
oblong, densely pubescent, 2.5-3 cm. long, 4-5-seeded.

The type was labeled Luphms parvifloj^iis, to which it is some-
what related, but the raceme is much denser, the flowers larger,
the corolla of a different color and with a very dark spot.

Utah : P. V. Junction, Wasatch Mts., 1883, M. E. Jones (typ
in herb. N. Y. Bot. Garden and duplicates in U. S. Nat. Herb.);
Woods Class, 1882, M. E. Jones.

Lupinus pulcherrimus sp. nov.

Perennial, more or less cespitose : stems 3-6 dm. high, spar-
ingly appressed hairy, leafy : petioles 3-6 cm. long, also appressed
hairy or glabrate ; leaflets 7-9, linear-oblanceolate, acute, appressed
silky on both sides, but green ; stipules lanceolate, subulate : raceme
more or less elongated and rather dense, short-peduncled ; bracts
lanceolate, short-acuminate, shorter than the buds, deciduous :
pedicels and calyx grayish silky ; lips of the latter lanceolate, the
upper somewhat shorter : flower about i cm. long : corolla dark
purple ; banner with a light spot, pubescent on the back : pods
broadly oblong, 3-3.5 cm. long, 5-6-seeded.

This species is also related to L. laxiflonis, but the raceme is
denser, the calyx scarcely spurred, and the flowers are larger. It
also resembles L. pseudopai'viflorus, but in that species the leaves are
glabrous above and the calyx also spurred. From L. adsccndens
it differs in the spreading flowers and the short bracts.

Wyoming: Battle, Continental Divide, 1901, F. Ttveedy, 4215
(type in herb. N. Y. Bot. Garden) ; Headwaters of Tongue River,
1898, Tweedy, 130 ; Big Horn Mountains, 1899, Tweedy, 2363 ;
La Barge, Uintah County, 1894, E. Stevenson, 158.

Montana : Divide between McDonald and Comas Lakes, 1901,
F. K. Vrccland, gg6 ; Little Belt Pass, 1896, P. A. Rydberg, 3318 ;
J. H. Flodnian, 620 ; Baltic, 1900, E. V. Wileox, 38.

Lupinus laxus sp. nov.

Perennial and somewhat cespitose : stem 4-6 dm. high, slender,
sparingly appressed, silky: petioles 3-12 cm. long, appressed

Rocky Mountain Flora 259

silky ; leaflets 7-9, linear or linear-oblanceolatc, 3-4 cm. long,
green, but appresscd silky on both sides ; stipules subulate :
racemes slender and lax ; bracts lanceolate, short-acuminate, not
longer than the buds : pedicels and calyx rather long, hairy :
flowers about 8 mm. long: calyx not spurred; lips ovate, the
upper slightly shorter : corolla light bluish-purple ; banner broad,
somewhat shorter than the wings.

This species is closely related to L. laxijlonis, but the flowers
are smaller and lighter and the calyx is not spurred. It grows at
an altitude of about 2000 m.

Montana: Forks of the Madison, 1897, Rydbcrg & Bcssey,
4442 (type in herb. N. Y. Bot. Garden), also 4443a.

Wvo.ming: La Barge, Uintah Co., 1894, E. Stivetison, ijj
(somewhat more hairy than the type).

Lupinus leucanthus sp. nov.

Perennial, about 6 dm. high : stem minutely strigose, leafy :
petioles about 6 cm. long, almost glabrous ; leaflets 5-7, narrowly
oblanccolate, 4-5 cm. long, acute or short-acuminate, glabrous
above, sparingly appressed hairy beneath ; stipules subulate :
raceme lax ; bracts lanceolate, acuminate, about equalling the
pedicels, 6-8 mm. long : pedicels and calyx appressed short-
haiiy ; lower lip of the latter linear-lanceolate, about ^ longer
than the ovate upper one : corolla white with a brownish spot on
the banner ; the latter glabrous and almost equalling the wings.

This species is nearest related to L. Sclienbcrac, but differs in
the larger white flowers and the longer upper lip of the calyx.
The type sheet bears the name Lupinus barbigcr Wats., but the
plant is not at all related to that species.

Utah: Springdale, 1894, Marcus E. Jones, j24ge (type in
U. S. Nat. Herb.).

Pachylophus macroglottis sp. nov.

Acaulescent or nearly so: leaves 8-15 cm. long, tapering
into a margined petiole ; blades broadly oblanccolate, acute, sinu-
ate-dentate, sparingly puberulent, and villous ciliate on the mar-
gins and the veins, very thin : hypanthium very long and slender,
12-14 cm. long, sparingly villous or nearly glabrous, gradually
widening into a funnelform throat, which at the insertion of the
sepals, petals and stamens, is about i cm. in diameter : sepals
lanceolate, gradually long-acuminate, about 5 cm. long and much
exceeding the broadly obcordate petals, which are about 3.5 cm.

260 Rydberg : Studies on the

long and fully as wide, very thin, at first white but turning pink in
age : filaments about 2.5 cm. long : anthers versatile, 1.5 cm. long
and 2 mm. wide : style exserted ; lobes of the stigma about 5 mm.
long : capsule with rather strong sinuately lobed ridges.

This species differs from P. cacspitosum (Nutt.) Raimann in the
villous ciliate leaves, the longer hypanthium which is much broader
at the throat than in any of the other species, the longer sepals
which much exceed the petals in length, and the less tubercled
ridges of the pod. It grows in cairions at an altitude of 1500-
2800 m.

Colorado : Tributaries of Turkey Creek, 1900, Rydberg &
Vreeland, ^^57 (type).

Pachylophus exiguus (A. Gray) Rydb. nom. nov.

Oenothera exigua A. Gray, PI. Fendl. 1849.

This species was merged into P. caespitosus by Dr. Watson ;
but it is perfectly distinct. The plant is nearly always caulescent
with a stem 0.5-2 dm. high, the leaves and hypanthium are villous
as in the preceding species, but still more so, the hypanthium is
fully as long, but less widened at the throat, and the angles of the
pod with more prominent sinuately lobed crests than in P. caespi-
tosus. This species is represented in the New York herbaria by
the following specimens :

Colorado: Rocky Ford, 1891, C, S. Crandall ; Mesas near
Pueblo, 1900, Rydberg & Vreeland, 58^8 ; Rocky Ford, 1900,
G. Oste7'ho2it, 2085.

New Mexico : White Mountains, 1897, E. 0. Wooton, 6§g.

Arizona: Walnut Canon, i^g^, MacDougal, j^i.

Androsace puberulenta sp. nov.

A puberulent rosulate annual : leaves oblanceolate, 1—3 cm.
long, acute, entire or sinuately denticulate, densely puberulent : pe-
duncles several, 3-10 cm. long : bracts narrowly lanceolate, 3-4
mm. long: pedicels 1-5 cm. long, spreading, densely puberulent :
calyx 3-4 mm. long, cleft to the middle ; tube glabrous, turbi-
nate ; lobes lanceolate, strongly keeled, densely puberulent, in fruit
much exceeding the short capsule : corolla white, about equalling
the calyx.

This species is in habit intermediate between A. siibinnbellata
and A. diffusa, but differs from both in the densely puberulent

RocKv Mountain Floka JCl

pedicels and calyx-lobes. In both specimens mentioned the calyx-
lobes are of the length and shape of the valves of the maturinj^
capsules, while in A. puberulcnta they are much lonj^er. The
range of the species extends from Manitoba and the Mackenzie
River south to New Mexico. The type was collected in south-
ern Colorado near Veta Pass, 1 900, Rydbcrg & Vrccland, 5772
(N. Y. Bot. Garden).

Gilia subacaulis sp. nov.

Apparently biennial with a slender taproot : stem less than i dm.
high, almost leafless, diffuscl}- branched, glandular-puberulent :
leaves almost all basal, 1.5-3 cm. long, pinnatifid with triangular to
oblong divisions, glandular-puberulent, or slightly white-villous
when young : inflorescence open ; pedicels 2-4 mm. long : calyx
2-2.5 "^"^- ^o^g. puberulent, scarious between the green ribs;
lobes lanceolate, about half as long as the tube, acute: corolla 4-
6 mm. long, funnelform, fully twice as long as the calyx, its tube
distinctly exserted : stamens included : capsule rounded ovoid,
acute, 4-5 mm. long.

This is related to 6. iiiconspuita, but differs in the almost leaf-
less stem, and the smaller corollas which have a comparatively
longer tube. The following specimens belong here.

Wyoming : Point of Rocks, 1901, Merrill & Wilcox, 607 (type
in Herb. N. Y. Bot. Garden).

Utah: St. George, 1877, Palmer, ^28 ; 1874, Parry, igg.

Colorado: Gunnison River, 1894.

Mertensia subpubescens sp. nov.

Mertensia Sibirica Rydb. Mem. N. Y. Bot. Gard. i : 335.
1901 ; mainly.

Perennial with a thick rootstock : stem simple, stout, 6-10
dm. high, glabrous throughout: lower leaves petioled, the upper
sessile; blades broadly lanceolate, 5-10 cm. long, tapering at
both ends, thin, glabrous or nearly so above, short-pubescent
beneath and ciliate on the margin : inflorescence a rather many-
flowered short panicle ; pedicels slightly strigose : calyx-lobes
oblong-linear, obtuse, ciliate on the margin, 2-2.5 rn^n- ^ong. about
y^-Yl as long as the corolla-tube: corolla about 12 mm. long;
the tube only slightly longer than the limb : filaments broad and
dilated, broader than the anthers and about half as long.

This is perhaps nearest related to M. eiliata, but is easiest dis-
tinguished by the pubescent lower surface of the leaves. It

262 Rydberg : Studies on the

grows along streams at an altitude of 2000-2200 m. The follow-
ing specimens from Montana belong here :

Montana: Spanish Basin, 1897, Rydberg & Bessey, 48^6
(type in herb. N. Y. Bot. Garden); near Indian Creek, ^6*7.? /
Bridger Mountains, 487^ ; Mystic Lake, 1895, C. L. Shear, joyd ;
Deer Lodge County, 1901, Mrs. Eiiniia W. Se/uuber, 42 ; Basin,

1902, Kilscy.

Heliotropium spathulatum sp. no v.

Hcliotropiujii eiirassai'ieitin Wook. Fl. Bor. Am. 2: 81. 1840.

Not L.

? Heliotropium curassavicum obovatiim DC. Prod. 9:538. 1 845.
Not H. obovahiin D. Don.

A glabrous, more or less glaucous, fleshy perennial with as-
cending more or less branched stems, 3—5 dm. high : leaves spat-
ulate, fleshy, indistinctly nerved, 2-5 cm. long, obtuse or rounded
at the apex, tapering at the base and the lowermost petioled : inflor-
escence branched into 2—5 racemes ; these often starting from one
point at the end of the common peduncle : calyx about 3 mm.
long, cleft to the middle ; its lobes ovate-lanceolate, acute : corolla
white or slightly tinged with blue, 6-8 mm. long ; the limb of
about the same width : nutlets 2.5-3 '^n^- long, scarcely rugose.

This is the common plant of the Rocky Mountain region that
has been known as H. curassaviciini L., but has much broader
leaves, more than twice as large flowers and larger fruit than the
cosmopolitan plant found on our eastern seaboard and in the south-
ern states. It is probably the same as the plant collected by
Douglas in the Blue Mountains of Oregon and briefly described
in the places cited above. Hooker states that in the Blue Moun-
tains were the only places where Douglas found H. ciirassavicnin.
In the Columbia herbarium there is a specimen, labelled : " Cali-
fornia, Douglas^ If this specimen is one of those collected in
the Blue Mountains and described by Hooker and DeCandoUe,
the synonyms cited above should be excluded ; for the specimens
evidently belong to the short-leaved and smaller-flowered form of
H. citrassavicuni common on the Pacific coast and described by
Willdenow under the name H. clienopodioidcs.

H. spatludatnm is fairly common in meadows throughout the
Rocky Mountain region and the great plains from Assiniboia and
Washington to Iowa, New Mexico, Chihuahua and California.
The type was collected at Great Falls, Montana, 1890, R. S.
Williams, ^42 (in herb. N. Y. Bot. Garden).

New York Botanical Garden.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 37.

SOME GENERIC SEGREGATIONS

By p. a. KYDBERG

NEW YORK
1903

[Reprinted from the Bulletin of the Tokrey Botanical Club, 30: 271-281. My 1903]

Some Generic Segregations

BV P. A. RVDliERG

(With Plates 13 and 14)

The concept of a genus may vary among botanists as well as
among zoologists. To some a genus is an aggregate of all forms of
organisms which can be included within a certain artificial descrip-
tion or diagnosis ; to others it is a natural group of closely related
species or forms and may be modified, extended or limited as soon
as these forms become better known. It is unnecessary to state
that the former view nowadays has very few advocates, for nature
was never made to follow the narrow concepts of man. A system-
atist who holds the second view, tries first to find what species are
closely related, and when these are brought together, he tries to
draw a diagnosis of the genus. The limitation of a genus depends
naturally upon the personal views of the taxonomist. What is a
single genus to one botanist may to another constitute half a
dozen or more good genera, /. c, groups of related species. It
matters little how broad or narrow concepts we have of a genus,
if only we are consistent and in the same family or tribe desig-
nate as genera equivalent natural groups of related species ; /. c, not
making in one case the limits of a genus too large and in another
too narrow.

I. An inconsistency of this kind exists, I think, in the usual
treatments of the family Mclanthaccae. Chrosperma and Stenan-
tJiium, Mclantliinni and Vcratruin, are separated by rather trifling
characters, while in Zygadcnus are included species of no closer
relationship. If we keep as distinct all of the first four genera,

271

272 Rydberg : Some Generic Segregations

we must, if consistent, divide Zygadenus into at least three genera.
Briefly these may be distinguished as follows :

Plant with a rootstock ; each petal and sepal with two glands. Zygadenus.
Plant with a bulb; each petal and sepal with a single gland.

Ovary wholly superior ; gland obovate or semi -orbicular. Toxicoscordion.

Ovary partly inferior ; gland obcordate. Antidea.

Zygadenus Michx. Fl. Bor. Am. i: 213. 1803
This genus was established on a single species Z. glabcrrimiis
and the generic name will go with it. This species differs from all
that later have been included in the genus, in having a thick elon-
gated rootstock instead of a bulb and two distinct round glands
on each of the petals and sepals. As far as I know the genus
consists of only one species.

Toxicoscordion gen. nov.

Glabrous, more or less poisonous perennials with narrow linear
conduplicate leaves, yellowish racemose or paniculate flowers and
membrane-coated bulbs. Flowers perfect (in one species described
as polygamous, but as far as I know erroneously so). Petals and
sepals bearing a .single obovate or semiorbicular gland at or above
the base of the blade, in most species more or less clawed. Ovary
wholly superior. Stamens (except in one species) more or less
adnate to the base of the petals and sepals ; anthers subreniform,
confluently one-celled. Capsule 3-celled, 3-beaked, septicidal.
Seeds numerous.

The first species known belonging to this group was published
under the name Amianthhun Nuttallii A. Gray ; but as this is
somewhat aberrant from the general type of the genus, I have
selected one of the most common species, Zygadenus intermcdms,
as the type. The following species belong here :

1. T. intermedium; Z. intermedins Rydb.*

2. T. venenosum (S. Wats.) ; Z. venenosns S. Wats.

3. T. gramineum ; Z. gmminens Rydb.

4. T. acutum ; Z. acntus Rydb.

5. T. f alcatum ; Z.falcatus Rydb.

6. T. paniculatum (Nutt.) ; Hclonias panicjdata Nutt.

7. T. Nuttallii (A. Gray) ; AmiantJdnm Nuttallii A. Gray.

*For citations and synonyms, see Bull. Torrey Club, 27 : 534-538. 1900.

Ryduerg : Some Generic Segregations 273

8. T. Fremontii (Torr.) ; Anticka Frcmontii Torr. Pac, R. R. Rep.
4: 144. 1856; Zygadcnus Fremontii Torr.; S. Wats. King's
Exped. 5 : 343. 1871.

Anticlea Kunth.

This genus was established on A. Sibirica and A. glaiica with
A. Mcxicana added as a doubtful species. A. Sibirica, the first
one mentioned, must be regarded as the type. It differs somewhat
from our American species, especially in the narrow somewhat
recurved petals and sepals and the narrow glands, but these differ-
ences are rather unimportant and do not warrant any generic
separation. The genus is characterized by a membranous coated
bulb, linear glabrous leaves, greenish or yellowish-white flowers,
withering persistent petals and sepals, which are adnate to the base
of the ovary and bear a single obcordate gland ; free stamens and
a partly inferior ovary. The species are :

1. A. Sibirica (L.) Kunth, Enum. 4: 191. 1843; Mclantlmnn

Sibiricnvi L. Sp. PI. 339. 1753; Zygadcnus Sibiricus Y^nvX\v,
Enum. 4: 192, as synonym.

2. A. chlorantha (Richards.) ; Z. cJdorantlnis Richardson, Frank!.

Journ. 736. 1821 ; Z. conimutatns Schultes, Syst. 7: 1560.
1830; Z. glaiicns Nutt. Jour. Acad. Phila. II. 7: 56. 1834;
Anticlea glauca Kunth, Enum. 4: 192. 1843.

3. A. virescens (H.B.K.) ; Hclonias viresccns H.B.K. Nov. Gen.
•& Sp. I: 267. 1816; A. Mcxicana Kunth, Enum. 4: 193,

1843 ; ^- Mexicaniis Hemsl. Biol. Cent. Am. 3: 382. 1885.

4. A. elegans (Pursh) ; Z. clegans Pursh, Y\. Am. Sept. 241.

1814.

5. A. Coloradensis ; Z.Coloradensis Rydb. Bull. Torrey Club, 27 :

534. 1900.

6. A. porrifolia (Greene) ; Z. porrifoliiis Greene, Bull. Torrey

Club, 8: 123. 1881.

II. Another segregation needed, I have found in the genus
Rubles. If Dalibarda is to be held distinct from that genus and
DucJicsnca from Fragaria, etc., certain species have to be removed
from Rubus as now understood, as both in general habit and the

274 Rydberg : Some Generic Segregations

structure of the fruit they are so unlike the rest of the species.
The segregated genera would be distinguished as follows :

Styles club-shaped ; stigmas slightly 2-lobed ; receptacle flat ; shrubs, unarmed, with
shreddy bark and digitately ribbed maple-like leaves.
Drupelets capped by hard hairy cushions ; styles glabrous ; erect shrubs. Rubacer.
Drupelets without cushions ; styles hairy ; prostrate or reclining vines. Oreohatus.
Styles filiform, glabtous ; stigmas usually capitate ; receptacle hemispherical, conical or
nipple-shaped ; drupelets without cushions ; leaves in most species pinnate and
stem in most prickly. Rzibtts.

Rubacer gen. nov.

Unarmed shrubs with more or less shreddy bark, more or less
glandular above. Leaves petioled, simple, digitately ribbed and
lobed, with acute or acuminate lobes and cordate bases, resem-
bling the leaves of certain maples. Flowers showy, in small
panicles : sepals ovate with long slender acuminations : petals pink
or white, oval or rounded : pistils numerous : receptacle flat or
nearly so : drupelets numerous, capped with a dry, rather hard,
finely and densely pubescent cushion with the club-shaped style
attached under the margin.

The type is the first species cited below.

1. Rubacer odoratum (L.) ; Rnbus odomtiis L. Sp. PI. 494- I753-

2. Rubacer parviflorum (Nutt.); Rubus parvifloms Nutt. Genera

i: 308. 1818; Rubiis Niitkanus Moc; Seringe, DC. Prodr.
2 : 566. 1825.

3. Rubacer tomentosum ; Rubus vdutinus Hook. & Arn. Bot.

Beechey Voy. 140. 1841 ; Not Vest. 1823.

4. Rubacer Columbianum (Millsp.) ; Riibus odomtiis var. Cohim-

bianus Millsp. Bull. W. Va. Agr. Exp. Sta. 2: 355. 1892.

Oreobatus gen. nov.

Prostrate or reclining shrubby vines with light colored, shreddy
bark. Leaves petioled, simple, maple- or gooseberry-like, digi-
tately ribbed and lobed with rounded serrate lobes. Flowers
showy, usually solitary : sepals ovate or ovate-lanceolate, acumi-
nate or the tips foliaceous, in fruit erect: petals white, 1.5-3 cm.
long, broadly obovate : stamens very numerous, hiding the fewer
pistils : receptacle flat or nearly so : styles club-shaped, curved ;
stigma somewhat 2-lobed ; young fruit apparently very pubescent,
owing to the hairy styles : drupelets not very fleshy, without hairy
cushions.

The first of the following two species is regarded as the type :

RvDBERG ; Some Generic Segregations 275

1. 0. deliciosus (James); Rubus dcliciosns James; Torr. Ann,

Lye. N. V. 2: 196. 1828.

2. 0. Neo-Mexicanus (A. Gray) ; Rubus Nco-Mcxicanus A. Gray,

PI. VVri^rht. 2: 55. 1S53.

III. A similar inconsistency we find in the Loasaceae, where the
genus Mcntzelia as it is represented in United States is composed
of at least four distinct groups deserving generic rank * with just as
good right as several other recognized genera in the family.
The principal distinguishing characters are :

Placentae with horizontal lamellae between the seeds ; these in 2 rows.

Filaments 3-cleft at the apex, the middle tooth bearing the anthers, the lateral

teeth cuspidate ; seeds angled or with folds, not winged ; annuals.

Bicuspidaria.

Filaments not cleft ; seeds very flat, more or less winged ; perennials.

Touterea.
Placentae without lamellae ; seeds usually prismatic.

Placentae slender, filiform ; ovules in one row, 10-40 ; seeds minutely muricate,
not striate ; filaments free or nearly so. Acrolasia.

Placentae broad, band-like ; ovules in 1-2 rows, few ; seeds distinctly striate,
often rugose ; filaments at the base united with the petals into a ring.

Afenlzelia.

Bicuspidaria (S. Wats.) gen. nov.

Mentzclia § Bicuspidaria S. Wats. Proc. Am. Acad. 20 : ^dj.

1885.

Annuals with rather large flowers. Petals 5, free; stamens
80-130, dilated, 3-cleft at the apex, the middle tooth bearing
the anthers ; placentae with a vertical wing and horizontal la-
mellae between the ovules and seeds ; these i 5-40 in 2 rows on
each placenta ; capsule cylindric, inflated, sessile, thin-walled; seeds
irregular, angled or with folds, not winged.

The first of the following species is taken as the type.

1. B. tricuspis (A. Gray); Mcntzelia tricuspis A. Gray, Am.

Nat. 9 : 271. 1875.

2. B. involucrata (S. Wats.) ; Mcntzelia involucrata S. Wats.

Proc. Am. Acad. 20: 367. 1885.

3. B. hirsutissima (S. Wats.); Mcntzelia hirsiitissiuia S. Wats.

Proc. Am. Acad. I2 : 252. 1877.

* It may be that the other three sections of Urban and Gilg, viz. , Micromentzelia,
Octopetaleia and Dendromenlzelia should also be treated as genera ; but better knowl-
edge is needed.

276 Rydberg : Some Generic Segregations

TouTEREA Eaton & Wright, N. Am. Bot. 454. 1840

Bartonia Sims, Bot. Mag.//. 148J. 1804. Not Muhl. 1801.

lorrcya Eat. Man. Ed. 7. 560. 1836. Not Raf. 18 18.

Hesperaster Qoz\s.&xqW., Torreya, i: 142. 1901.

This genus contains nearly all the large-flowered species in-
cluded in Mentzelia. The outer filaments are more or less dilated ;
in most they are without anthers and in some as large as and
resembling the petals. These species are often described as hav-
ing ten petals. The genus Toiiterea was based on the first two
species mentioned below ; Bartonia Sims on Bartonia decapetala
and the type of Hesperaster is the same. As far as I know, the
following species belong here :

1. T. decapetala (Pursh) ; Bartonia decapetala Pursh, Bot. Mag.
pi. i^Sy. 1812; B. ornata Pursh, Fl. Am. Sept. 327. 1814;
lonterea ornata Eat. & Wright, N. Am. Bot. 454. 1840; M.
ornata Torr. & Gray, Fl. N. Am. i: 534. 1840.

2. T. NUDA (Pursh) Eat. & Wright, N. Am. Bot. 454. 1840;
Bartonia nnda Pursh, Fl. Am. Sept. 328. 1814 ; M. nuda Torr.
& Gray, 1. c.

3. T. laevicaulis (Dougl.) ; Bartonia laevicanlis Dougl. ; Hook.
Fl. Bor. Am. i: 221. 1833; M. laevicanlis Torr. & Gray,
Fl. N. Am. I : 535. 1840.

4. T. Brandegei (S. Wats.) ; Af. Brandegei S. Wats. Proc. Am.
Acad. 20 : 367. 1885.

5. T. parviflora (Dougl.) ; B. parviflora Dougl. ; Hook. Fl. Bor.
Am. I : 221. 1833.

6. T. pterosperma (Eastw.); M. pterosperma Eastw. Proc. Cal.
Acad. II. 6 : 290. 1896.

7. T. stricta Osterhout Mss. ; Hesperaster strictiis Osterhout,
Bull. Torrey Club, 29: 174, 1902.

8. T. Rusbyi (Wooton) ; Mentzelia Rusbyi Wooton, Bull. Torrey
Club, 25 : 261. 1898.

9. T. Wrightii (A. Gray) ; M. Wrightii A. Gray, PI. Fendl. 48.
1848.

10. T. speciosa Osterhout Mss.; M. speciosa Osterh. Bull. Torrey
Club, 28: 689. 1901 ; M. anrea Osterh. Bull. Torrey Club,
28:644. 1901. Not Nutt. 1818.

RvDBERG : Some Generic Segregations 277

I I. T. densa (Greene) ; M. dcnsa Greene, Pittonia, 3 : 99. 1896.

12. T. pumila (Nutt.) ; M. piimila Nutt. ; Terr. & Gray, Fl. N.
Am. I : 535. 1840.

13. T. chrysantha (Engelm.) ; M. chrysautha luigelm. ; Brand.
Bull. U. S. Geol. Surv. Terr. 2 : 237. 1876 ; M. lutca Greene,
Pittonia, 3 : 99. 1896.

14. T. humilis (A. Gray) ; M. iiiit/ii/IoraxSiY. hiniii/is A. Gray, PI.
Wright. I : 74. 1852.

15. T. multiflora (Nutt.) ; Bartonia viultijiora Nutt. Jour. Acad.
Phila. II. I : 180. 1847.

16. T. perennis (Wooton) ; M. perennis Wooton, Bull. Torrey
Club, 25 : 260. 1898.

17.* T. MULTic.\ULis Osterhout, Bull. Torrey Club, 30 : 236. 1903.

AcROLASiA Presl. Rel. Haenk. 2: 39. 1831

TradiypJiytum Nutt. ; Torr. & Gray, Fl. N. Am. i : 533. 1840.

This genus was established on a Chilian species, A. bartoni-
oidcs Presl, which has fewer ovules (12) and fewer stamens than
its North American allies, but the structure of the ovary and seeds
is the same. All the species are annuals and characterized by the
long cylindric capsules, the filiform placentae and the muricate
more or less prismatic seeds. Most of the species are small-
flowered, with linear, lanceolate or ovate-oblong, laciniate, dentate
or more rarely entire sessile leaves. The following North Ameri-
can species belong here :

1. A. congesta (Nutt.); M. congesta Nutt; Torr. & Gray, Fl.
N. Am. I : 534. 1840; Trachypliyttim congestuvi Nutt. ; Torr.
& Gray, 1. c.

2. A. ctenophora ; M. ctenophora Rydb. Bull. Torrey Club, 28 :
11- 1901-

3. A. Tweed5ri ; M. Tioccdyi Rydb. Mem. N. Y. Bot. Gard. i :
271. 1900.

4. A. albicaulis (Dougl.) ; M. albicanlis Dougl. ; Hook. Fl. Bor.
Am. I : 222. 1833 ; Bartonia albicaulis Hook. 1. c.

5. A. tenerrima ; M. tenerrima Rydb. Mem. N. Y. Bot. Gard.
I : 271. 1900.

278 Rydberg : Some Generic Segregations

6. A. integrifolia (Wats.) ; M. albicaulis var. integrifolia S.Wats.
King's Exp. 5 : 114. 1871;^ dispersa S. Wats. Proc. Am.
Acad. II : 115. 1876.

7. A. compacta (A. Nelson) ; M. compacta A.Nels. Bull. Torrey
Club, 25 : 275. 1898.

8. A. gracilenta (Torr. & Gray) ; M. gracilenta Torr. & Gray,
Fl. N. Am. I : 534. 1840.

9. A. affinis (Greene) ; M. affinis Greene, Pittonia, 2 : 103. 1890.

10. A. aurea (Lindl.) ; Bartonia aiirea Lindl. Bot. Reg. 22 : //.
1831. 1836. Not Mentzelia mirea Nutt. 1818 ; M, Lindleyi
Torr. & Gray, Fl. N. Am. i : 533. 1840.

11. A. nitens (Greene); M. nitens Greene, Fl. Francisc. 234.
1891.

12. A. Veatchiana;(Kellogg) ; M. f^'rt/r/!/^;/^ Kellogg, Proc. Cal.
Acad. 2 : 99. 1863.

13. A. pectinata (Kellogg); M. pectinata Kellogg, Proc. Cal.
Acad. 3 : 40. 1868.

1 4. A. micrantha (Torr. & Gray) ; M. viicrantha Torr. & Gray,
Fl. N. Am. I : 535. 1840.

IV. A fourth segregation is proposed in Syntliyris. Profes-
sor Greene has included this genus in Wnlfenia and he has been
followed by many American botanists. Wtilfenia, represented by
European and Asiatic plants, has a 4-valved capsule, which is
neither flattened nor emarginate at the apex, and a corolla with
a long tube. All the American species have flattened capsules,
which are 2-valved and obcordate at the apex. In this respect
they agree perfectly with Veronica. The original Synthyris (S.
reniformis) and a few related species have a corolla that is built
exactly on the same plan as that of Veronica, and it is only the
general habit that keeps them out of that genus. 6". alpina and
its relatives, on the contrary, have an altogether different corolla,
cleft to near the base into two distinct lips. 5. rubra and its allies
have no corollas but agree otherwise both in general appearance
and structure of the flower and fruit with S, alpina. It would
however hardly be advisable to place them in different genera
merely on account of the presence or absence of the corolla. The
related genera are distinguished as follows :

RvDBERG : Some Generic Segregations 279

Corolla scarcely a-lipped ; almost e<iually deeply 4-lol)t(l.

Corolla tubular-fuiinelform ; capsule neither llattened nor obcordate at the
apex, 4-valved.
Leaves alternate but mostly basal. [Vulfenia.

Leaves opposite ; stem leafy. Lept,imira.

Corolla rotate or shortcampanulate ; capside flattened, obcordate, 2-valvcd.

Leaves opposite or whorled ; stem leafy. Veronica.

Leaves alternate, but mostly basal ; plant with a mostly naked scape.

Synthyris.
Corolla, if present, deeply 2-lipped ; upper lip entire, broad, arcuate; lower lip
straight and deeply laciniate ; leaves alternate but mostly basal ; scape bracted.

Besseya.

Synthyris Benth. ; DC. Prodr. lo : 454. 1846

This genus was based on .S". rcniformis. S. rubra was also
included in the genus by the author, but he admits that it is a
rather anomalous member, lacking the corolla. It is, however,
more nearly related to iT. alpina, which has a corolla of a different
structure. The species belonging here are the following: ^. rcni-
formis (Dougl.) Benth.. .V. major (Hook.) Heller, 5. pinnatifida
S. Wats, and S. laciniata (Gray) Rydb. Whether .S. rotiindifolia
A. Gray should also be included or should be made the type of
a separate genus is a little doubtful. At present I am inclined to
regard it as a Syntliyris. So also 5". schisantlia Piper, with lacerate
corolla-lobes, which species is still more aberrant.

Besseya gen. nov.

Low perennials with a thick rootstock. Basal leaves petioled ;
blades ovate or oblong, crenate; stem-leaves bract-like, alternate:
inflorescence a bracted spike : calyx almost regularly 4-cleft to
near the base or cleft to the base on the upper side and then 2-3-
lobed : corolla 2-lipped, cleft to near the base, or lacking ; upper
lip, if present, entire, obovate or cuneate-obov^ate, concave ; lower
lip much shorter, often irregularly 2-3-cleft or laciniate : stamens
2, attached at the base of the corolla, or if this is absent on a
small disk : capsule flattened, obcordate, 2-valved.

It is with pleasure that I dedicate this genus of western moun-
tain plants to Professor Charles E. Bessey, of the University of
Nebraska, my former teacher and the most prominent botanist of
the region where these small plants grow.

The first species cited below is regarded as the type.

280 Rydberg : Some Generic Segregations

a. With corolla :

1. B. alpina (A. Gray); Synthyris alpina A. Gray, Am. Jour,

Sci. II. 34 : 251. 1862.

2. B. BuUii (Eaton) ; Gymnandra Biillii Eat.; Eat. & Wr. N. Am.

Bot. 259. 1840; Synthyris Hoiightoniana ^Qnih..; DC. Prodr.
10 : 454. 1846.

3. B. plantaginea (Benth.) ; 5. plantaginea Benth.; DC. Prodr.
10 : 455. 1846.

4. B. reflexa (Eastw.) ; 5. reflexa Eastw. Proc. Cal. Acad. III.

I : 124. 1898.

5. B. Ritteriana (Eastw.) ; 5. Ritteriana Eastw. 1. c. 123.

b. Without corolla :

6. B. rubra (Dough); Gyjunandra rubra Dough; Hook, Fh Bor,

Am. 2 : 103. 1838; Synthyris rjibra Benth.; DC. Prodr. 10 :
455, 1846.

7. B. gymnocarpa (A. Nels.); Widfenia gymnocarpa K. Nelson,

Bull. Torrey Club, 25: 282. 1898.

8. B. Wyomingensis (A, Nels.) ; W. Wyomingensis A, Nelson,

h c, 281.
New York Botanical Garden,

Explanation of Plates

Plate 13

Fig. I. Zygadentis glaberrimus Islxohx. a, section of flower ; b, sepal ; c, petal ;
d, stamens ; e, pistils.

Fig. 2. Toxicoscordion intermedium Rydb. a, b, etc., as above.

Fig, 3. Antidea elegans (Pursh) Rydb. a, b, etc., as above.

Fig, 4. Antidea Sibirica (L. ) Kunth. a, b, etc., as above.

Fig. 5. Rubtis nigrobaccus Bailey, a, section of receptacle ; b, a young drupelet ;
C, stone.

Fig. 6. Rubacer odoratum (L. ) Rydb. a, b, c, as in the preceding.

Fig. 7. Oreobatm delidosus (James) Rydb. a, b, C, as in the preceding.

Plate 14

Fig. I, Touterea decapetala (Pursh) Rydb. a, section of the flower showing two
of the petals and two of the similar staminodia ; b, one of the petals ; c, stamens ; d,
style ; e, partial section of the capsule showing one placenta in front view and half of
another in lateral view ; f , winged seed.

Fig. 2. Bicuspidaria tricuspis (A. Gray) Rydb. a, section of flower ; b, fruit ; c,
pistil ; d, stamen with the two cusps ; e, section of a part of the capsule, showing in

RvDBERG : Some Generic Segregations 281

lateral view one half of two placentae with seeds, between them the horizontal lamellae
and behind the vertical wings ; between the latter are seen seeds on the third placenta ;
f, a seed.

Fig. 3. Acrolasia alhiiaulis (Dougl.) Rydb. a, section of flower; b, petal ; c,
Stamens ; d, pistil ; e, capsule ; f , longitudinal section of the same showing the narrow
filiform placenta with seeds removed ; g, a muricate prismatic seed.

Fig. 4. Mentzelia oligosperma Nutt. a, section of flower ; b, stamen ; c, pistil ; d,
section of a part of the capsule, showing the broad placenta with the seeds removed,
except one ; e, a striate seed.

Fig. 5. Synthyris reniformis (Dougl.) Henth. a, a flower ; b, sepal ; c, corolla
laid open ; u, upper lip ; 1, lower lip ; d, pistil.

Fig. 6. Besseya alpina (A. Gray) Rydb. a, b, etc., as in the preceding.

Bull. Torrey Club, 30.

Pi.. 13

6

AMm^^a dUl.

SEGREGATES OF ZYGADENUS AND RUBUS.

Bull. Torrey Club, 30.

Pi.. I;

t.

A.Mar'-oi' J'l

SEGREGATES OF MENTZELIA AND SYNTHYRIS.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 38.

THE POLYPORACEAE OF NORTH AMERICA-IV.
THE GENUS ELFVINGIA

By WILLIAM ALFONSO MURRILL

NEW YORK
1903

[Reprinted from the Bulletin of the Torrey Botanical Club, 30: a«-fl8t. My 1903]

The Polyporaceae of North America.— IV. The Genus Elfvingia

Hy Wii.i.iAM Ai.iiioNso Mlrrii.l

Tlic mycological field of Europe is small as compared with
that of America. The European mycologist has in general a
small country with limited variation in altitude, geological forma-
tion and season, as well as a limited number of host plants, few of
which are readily attacked by wood-loving fungi ; hence it is pos-
sible to know more minutely the forms that do occur and to seg-
regate varieties according to host, position and other details of
which little note has yet been taken in America. The confusion
of forms in Europe is largely due to former difficulties in travel,
limited means, lack of interest in the work of others and lack of
generalization, a condition of affairs accounted for to some extent by
difference in language and the antipathy common among neighbors.
Upon receiving an exotic plant, be it from Guinea or Greenland,
the mycologist of limited outlook makes a heroic effort to corre-
late it with some species growing in his immediate vicinity and,
failing to do this, either loses interest altogether in the matter or
half-heartedly looks up a name for it in some foreign flora. To
send American fungi abroad for determination is usually to be
either disappointed or deceived. To many European botanists,
Kew is a very foreign country and America is off the map.

One should not be surprised to find, therefore, that many
American plants are parading under false names and that many
are entirely new to the botanical public ; and what is true of species
is likewise true of genera ; the old molds must be recast and
several new ones added to accommodate these large additions.
For our northern forms, the work of recasting has been largely
done by Dr. P. A. Karsten, of ^'inland, who has to deal with a
local flora very similar to our own in its lines of cleavage, although
the abundance of species present in our latitude is unknown in
Finland. The basis of this similarity between North American
and North European and Russian genera is found in the fact that
these countries have been practically continuous so far as distri-
bution is concerned, whereas the continents of the southern hemi-

296

297 Murrill: Polyporaceae of North America

sphere have very probably been for ages widely separated by im-
passable barriers of open sea. The principles above enunciated
are well illustrated in the genus Elfvingia of Karsten, a discussion
of which follows.

Elfvingia Karsten, Findlands Basidsv. 333. 1889
This genus was described by Karsten as follows :

" Hatten bekladd med en sprod, skorpartad hud, glatt. Cystiderna foga anmark-
ningsvarda. Basiderna n. klotrunda, 4-sporiga. Sporerna aggrunda, vartFulla, gul-
brunaktiga."

The only representative of the genus in Finland, according to
Karsten, is Elfvingia applanata (Pers.) Karst., a species first de-
scribed as Boletus Lipsiensis by Batsch (Elench. Fung. 2: 183-
185.//. 2§. Fig. I JO. a, b. 1786), then as Boletus applanatus by
Persoon (Obs. Myc. 2: 2. 1799), and later as Poly poms ineris-
moidcs by Corda (Sturm, Deutschl. Flora, 3 : 139-142./)/. 6j.
1837). Its close relationship to P. fomentarius L. may be judged
from the fact that Persoon in his Synopsis treats it as a variety of
the latter plant, although he had described it as a distinct spe-
cies only two years before. In the present treatment of the genus
Elfvingia, E. Lipsiensis (Batsch) will not appear, since it does not
occur in America, its place being taken by E. uiegalouia (Lev.).

Synopsis of the Biortti American species

1. Context ferruginous, spores hyaline, pileus usually ungulate. 2.
Context darker, fulvous to chocolate-brown, spores yellowish-brown, pileus usually

applanate. 3-

2. Pileus exactly ungulate, pores 3 to a mm., growing in temperate regions south to

Carolina. I. E. fometitaria.

Pileus compressed ungulate, pores 5 to a mm., growing in tropical America and the
gulf states. 2. E. fasciata.

3. Hymenophore annual, persisting above later growths ; pileus reniform, margin thin,

spores dark brown, roughly echinulate, 8-9 X 7,w- 3- ^- reniformis.

Hymenophore truly perennial, tubes stratified, spores smooth or nearly so. 4.

4. Crust white, becoming brown, not separating, spores smooth, 8-9 X 5 ,"' growing

in temperate regions. 4- ^- megaloma.

Cru.st brown, context floccose, softer, spores smooth or slightly echinulate, growing
in tropical America. 5-

5. Pileus thick, crust smooth and very hard, hymenium plane, margin truncate, very

smooth, often laccate, context very thin, spores smooth or slightly verrucose,
7-9X5-6". 5- E. turnata.

Pileus thin, crust radially wrinkled, thinner, hymenium concave, margin undulate,
never laccate, spores smooth, 8X5/"- (y. E. Lioneth.

Murrill: Polyporaceae of North America 298

1. Elfvingia fomentaria (L.)

Boletus fomentarius \.. S^. PI. 1 176. 1753.
Boletus unguhxtus Bull. Herb. France,/'/. 4c) i. fig. 2, C. D. E.
Polyporus fouicutarius Yx. Syst. i : 374. 1821.
Fo)ucs fouioitarins Gill. Champ. Vx. i : 686. 1878 ; Icon.
467.

See also Sowerby, Engl. Fungi, //. /??, and Schaeffer, //.

137-

Thi.s fungus is widely distributed on birch and beech in the

temperate regions of Furope and North America. It varies little
in shape and appearance in the north, though very small forms
occur in high latitudes. One of these brought from the Isle a
la Crosse by Dr. Richardson was described by Berkeley as var.
excavatus (Ann. Nat. Hist. 3: 387. 1839). The type specimens
at Kew agree in all respects with P. fouicutarius as it grows in the
Adirondack Mountains and in Sweden. A more southern form,
larger, thicker, more uniform and ashy in color, is known in
Europe as Polyporus luacugae De Not., but as this does not occur
in North America it need not be discussed at present. When
fruit-bodies of P. fouicutarius become old and indurated and lose
their outer pellicle the horny crust takes a high ivory-black polish
and is very different from the usual covering of the plant. It is
this form that has frequently been mistaken for P. nigricaus Fr.
At Kew, P. igniarius and P. applanatus have been variously con-
fused with P. fomcutarius.

In addition to the usual European and American exsiccati, all
of which contain this species, material has been examined from
Germany (Magnus), Sweden (Murrill), France (Patouillard), Fin-
land (Karsten), Italy (Bresadola), Scotland (Klotsch), Canada
(Macoun), Maine (Ricker), New Hampshire (Miss Minns), Massa-
chusetts (Mrs. French), Wisconsin (Baker), New York (Under-
wood), Connecticut (Miss White), and most of the other states
south to North Carolina and west to California.

2. Elfvingia fasciata (Sw.)

Boletus fasciatus Sw. Prod. 149. 1788; Y\. Ind. Occ. 3:
1 92 1. 1806.

Polyporus fasciatus Yx. Syst. I : 373. 1821.

299 MuRRiLL : Polyporaceae of North America

Polyporus sclerodermciis Lev. Ann. Sc. Nat. Bot. III. 5 : 129.

1846.

Polyporus marmoratus Berk. & Curt. Proc. Am. Acad. 4 : 122.

1858.

Fomes fasciatus Codk&, Grevillea, 14: 21. 1885.

Specimens examined : Ellis & Everhart, N. A. Fungi, no.
1702; Florida (Martin, Calkins, Rolfs, Lloyd, Rau), Louisiana
(Langlois), Alabama (Underwood, Earle, Baker), Texas (Egel-
ing), Jamaica (Earle).

This plant is fairly abundant in tropical America, being known
from nearly all localities that have been explored. Earle found it
at three different points in Jamaica growing on Cedrela and cocoa-
nut logs. It replaces E. fomentaria in the south, but cannot be
confused with it as some have supposed.

3. Elfvingia reniformis (Morg.)

Polyporus reniformis Morg. Jour. Cincinnati Soc. Nat. Hist. 8 :
103. 1885.

Fonies reniformis Sacc. Syll. 6: 187. 1888.

Collections examined : Ohio (Morgan), Kentucky (Miss
Price), Indiana (Underwood), Virginia (Murrill, Mrs. Murrill),
North Carolina (Memminger), Alabama (Underwood, Earle),
Louisiana (Langlois).

When first sent to Kew this plant was determined as P. appla-
natus Pers., a closely allied European species. Later it was placed
by Cooke under P. incrassatus Berk., a plant indigenous to Aus-
tralia ! It may be said, however, that the specimens now at Kew
are not fairly representative even of the form and much less of the
habit of Morgan's species. The following extract from one of D.
L. James' letters written in 1882 refers to the peculiarities of P.
reniformis : " I am rather surprised at my fungus turning out to
be P. applanatus. It grows at the foot of old oak stumps and trees
and seems to be an annual, the previous year's growth being found
distinct and above the new growth in a decayed condition. The
pileus does not harden as in P. applanatus:' By P. applanatus is,
of course, meant our ordinary species, P. niegaloma Lev. " Again,
the hymenial surface instead of being flat, is curved upwards in
every case, this feature being well shown in the specimen sent.

MURRILL : POLYPORACEAE OF NORTM AMERICA 300

However, your experience I shall certainly hold above my ignor-
ance."

During a recent trip through the southwestern part of Virginia,
I found this fungus very common on stumps and decaying trunks
of oak. In every case the fruit-bodies of previous years were dead,
with the later growths below. I have specimens taken from a
white oak trunk on which I saw the first fruit-bodies two years
ago and some of these have three layers, some two and some one,
only the last year's growth in each case being alive. This corrob-
orates James' testimony to his belief that the fruit-bodies are
annual. During the same trip, I collected a few specimens of E.
mcgalonia on its usual hosts and noted with interest that the two
species are as distinct in Virginia as they are farther north. Facts
thus far collected indicate that E. roiiforinis is a southern form
and confined to a much more restricted area than E. mcgalomn,
while within this area it is often abundant and quite destructive to
the oak, its favorite host.

4. Elfvingia megaloma (Lev.)

Polyporus megaloma Lev. Ann. Sc. Nat. Bot. III. 5: 128.
1846.

Polyporus leiicophaciis Mont. Syll. Crypt. 157. 1856.

Fomes megaloma CooVe, Grevillea, 14: 18. 1885.

Gajiodervia leucophaenm Pat. Bull. Soc. Myc. 5: 73. 1889.

Specimens have been examined from New York (Shear, Mur-
rill), Canada (Anderson), Pennsylvania (Ellis), Virginia (Miss V.
W. Murrill, Murrill), and nearly all the other states west to Cali-
fornia and south to Georgia. The plant occurs abundantly on
dead or dying wood of oak, sweet gum, poplar and other decid-
uous trees. In the popular mind it is associated with etching
because its hymenium changes to a darker color when bruised.
Owing to the close resemblance to P. apphmatus Pers. in habit
and appearance, it has been generally known under that name
among American mycologists. The correct name for it, however,
appears to be that assigned to it by Leveille in 1846 on receiving
specimens collected by Menand in New York. In connection
with his description, he notes that it is near P. applaiiattis in form,
consistency and volume, but differs essentially in the color of the

301 Murrill: Polyporaceae of North America

cap and the pores and especially in the presence of a wide sterile
border circumscribing the tubes below.

5. Elfvingia tornata (Pers.)

Poly poms tornatns Pers.; Gaud. Voy. Freyc. Bot. 173. 1826,

Polyporus aiistralis Fr, Elench. 108. 1828.

/^3W« «?/5/r«/w Cooke, Grevillea, 14 : 18. 1885.

Ganoderma austmle Pat. Bull. Soc. Myc. 5: 71. 1889.

Scindalma tornatinn Kuntze, Rev. Gen. 3: 517. 1898.

Described from islands in the Pacific ocean and apparently ot
general occurrence throughout tropical America, since most of the
collections from that region contain specimens of it. A large plant
was collected last year by Percy Wilson in Porto Rico, and Earle
on his recent visit to Jamaica found it at three different stations ;
at Port Maria on a dead limb of a leguminous tree, at Hope
Gardens on a dead deciduous trunk and at Port Antonio on the
stump of a hog plum and the fallen trunk of a cocoanut palm.

6. Elfvingia Lionetii (Rolland)
Gmioderma Lionetii Rolland, Bull. Soc. Myc. 17: 180.

//. 8. 1 901.

This plant was collected by M. Lionnet on trunks in the isth-
mus of Panama. It is closely allied to E. tornata, but is thinner
with a thinner crust, which is usually profoundly wrinkled from
the center outward. The context is floccose, elastic, brownish-
rufous, and the spores ovoid, smooth, fulvous, 8 x 5 /i. Several
specimens are in the New York Botanical Garden collected by C.
L. Smith in Nicaragua.

New York City.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 39

A PRELIMINARY ENUMERATION OF THE
GRASSES OF PORTO RICO

By GEORGE V. NASH

NEW YORK
1903

[Reprinted from the Bulletin of the Torrey Botanical Club, 30 : 369-389. Jl, 1903]

A Preliminary Enumeration of the Grasses of Porto Rico

I5y Gkorgk V. Nash

This enumeration is based mainly upon the material in the
herbarium of the New York Botanical Garden. The collections of
Heller, Underwood and Griggs, Wilson and Goll are there fully
represented, and those of Sintenis partially so. A few species, of
which we have not seen specimens, have been admitted upon the
authority of monographers. Further explorations of the island
must add more species, and it is hoped that this enumeration may
serve as a help in such further exploration. That this work may
be facilitated, it has been deemed advisable to incorporate analyt-
ical keys to the tribes, genera and species. This enumeration
credits the island with lO of the 13 grass tribes, 37 genera, and
75 species and varieties, among them 9 hitherto unknown.

Key to tlie Tribes

A. Spikelets falling from the pedicel entire, naked or enclosed in bristles or a hur-like
involucre, or immersed in the internodes of a readily disarticulating racliis, i-
flowered, or if a-flowered the lower flower staminate (perfect in Isachne) : in-
ternodes of the rachilla of the spikelet very short, not measurable.
Spikelets round or somewhat dorsally compressed : hilum punctiform.

Flowering scale and palet hyaline, thin, much more delicate in structure than
the thick-membranous to coriaceous empty scales.
Spikelets unisexual. Tribe I. Maydeae.

Spikelets in pairs, one sessile, the other pedicellate, the former perfect,
the latter sometimes perfect, often with a staminate flower, or fre-
quently empty, abortive or wanting. Tribe II. Andkoi-oooneae.
Flowering scale, at least that of tlie perfect flower, similar in te.xture to the
empty scales, or frecpiently thicker and tirmer, never liyaline and thiji.
[The preceding number of the Bulleti.n, Vol. 30, Nq. 6, for June, 1903 (30:
319-368), was issued II Je 1903.]

3GiJ

370 Nash : Enumeration of the

Flowering scale and palet membranous.

Inflorescence spicate. Tribe III. Zoysieae.

Inflorescence paniculate. Tribe IV. Tristegineae.

Flowering scale and palet chartaceous, cartilaginous or coriaceous, very
different in appearance from the remaining scales.

Tribe V. Paniceae.
Spikelets much compressed laterally : hiluni linear. Tribe VI. Oryzeae.
B. Spikelets with the empty scales persistent, the rachilla hence articulated above them,
l-many-flowered : internodes of the rachilla of spikelets of 2 or more flowers long
and readily measurable.
Stems herbaceous (except in one species of /'a«^Vw;«) : leaf-blade sessile, not articu-
lated with the sheath.
Spikelets borne in open or spike-like panicles or racemes, usually upon distinct
and often long pedicels.
Spikelets i-flowered. Tribe VII, Agrostideae.

Spikelets 2-many -flowered. Tribe IX. Festuceae.

Spikelets borne in two rows, on short equal pedicels or sessile.

Tribe VIII. Chlokideae.
Stems woody : leaf-blade with a petiole-like base which is articulated with the
sheath. Tribe X. Bamisuseae.

Tribe I. Maydeae

A tall stout grass with the pibtillnte spikelets enclosed in a globose white bead-like
organ. I- Coix.

Tribe II. Andropogoneae

First scale of the spikelet globose, rugose: racemes not hairy. 2. Hackelochloa.

Scales of the spikelet long, flat or convex : racemes hairy.
Racemes singly disposed.

Second scale of the sessile spikelet awnless : pedicellate spikelet smaller than

the sessile one. 3- Schiziuhyiium.

Second scale of the sessile spikelet awned : pedicellate spikelet larger than the
sessile one. 4- Diedoniis.

Racemes in pairs, or sometimes 3's-5's. 5. Andropogon.

Tribe III. Zoysieae
A low grass with a narrow spike-like inflorescence. 6. Anthephora,

Tribe IV. Tristegineae
A tall grass with a large panicle of awned spikelets. 7. Arundinclla.

Tribe V. Paniceae

I. Spikelets all perfect : leaf- blades sessile.
a. Spikelets not sunken in the rachis.
* Spikelets naked, not involucrate.
Scales 3.

Lowest scale with a thickened ring-like callus. S. Monachne.
' Lowest scale unappendaged.

Flowering scale with its opening turned from the rachis.

9. Paspalum.

Grasses of Porto Rico t\7\

FIoweriiiR scale witli its openinc; turned toward the rachis.

lO. Atiastroplius.
Scales 4.

Perfect flowers 2. il. Isachnc.

Perfect flower i.

Second empty scale not saccate at the base.
Empty scales not awned.

Spikelets in very slender I -sided racemes which are
usually whorled or approximate.

12. Synt/ietisnin.
Spikelets in panicles or panicled racemes, the latter

never whorled, commonly distant.
Spikelets long-hairy, acuminate.

13. Trichachnc.
Spikelets glabrous or with only short hairs

usually orbicular to ovate, if lanceolate
glabrous.
Spikelets hemispheric, orbicular or nearly so
in outline, the rachisof the l-sided racemes
winged. 14. Dimorp/iostac/iys.

Spikelets not hemispheric, varying in outline,
usually in panicles, rarely in racemes
and then the rachis not winged.
Flowenng scale without lateral pits or ap-
pendages at the base.

15. Panicuvi.
Flowering scale with pits or membra

nous appendages at the base.

16. Ichnanthus.
Empty scales awned or awn -pointed.

Stems creeping : leaf-blades short and broad.

17. Oplismcnus.
Stems erect : leaf-blades long and narrow.

18. Ecltiiiochloa.
Second empty scale saccate at the base. 19. Sacciohpis.

** Spikelets involucrate.

Involucre of bristles. 20. Chactochloa.

Involucre of 2 spine-bearing valves. 21. Cenchrus.

b. Spikelets .sunken in one side of the broad flattened rachis. 22. Steitotaphriim.

2. Spikelets monoecious : leaf-blades with a petiole-like base. 23. Olyra.

Tribe VI. Oryzkae
Grasses with broad flat leaf-blades and long narrow spikelets. 24. Pharus.

Tribe VII. Ar.ROSTinEAE

Flowering scale with three long awns. 25. Aristida.

Flowering scale awnless. 26. Sporobolus.

Tribe VIII. Chlorideae

I. Spikelets with i perfect flower.
Spikes digitate or in whorls.

374 Nash : Enumeration of the

8. MONACHNE Beauv. Agrost. 49,/^- lo. 1S12
Eriochloa H.B.K. Nov. Gen. & Sp. i : 94, pi. 30,31. 18 15.

Rachis of the inflorescence softly pubescent with short spreading hairs : outer scales of
the spikelet pubescent with many long stiff hairs. I. J/, punctata.

Rachis merely scabrous on the angles : outer scales less pubescent and with shorter
finer hairs. 2. J/, subglabra.

I. MONACHx\E PUNCTATA (L.).

Milmm ptinctatiini L. Syst. Ed. lo, 872. 1759.

Eriochloa punctata Hamilt. Prod. 5. 1825.

Humacao, Blauner ; Adjuntas road, near Ponce, Heller 6298.

2. Monachne subglabra sp. nov.

A tall somewhat branching grass with rather broad, flat leaf-
blades. Stems 8-12 dm. long, glabrous, the nodes densely pu-
bescent with short, spreading hairs : leaves numerous ; sheaths
usually shorter than the internodes and much shorter than the
blades, ciliate on the margins, the lower ones papillose-hirsute
between the nerves ; ligule a ring of short hairs ; blades linear to
lanceolate, flat, ascending, or the longer ones drooping, glabrous,
very rough on the margins, long-acuminate at the apex, generally
narrowed toward the base, 1-3 dm. long and 7-17 mm. wide:
inflorescence finally somewhat exserted, the rachis hispidulous on
the ridges: racemes ascending, commonly 6- 10, the larger 4-6
cm. long, the rachis hispidulous on the margins and more or less
setiferous : spikelets on shorter pedicels which usually bear a few
long hairs, 3.5 mm. long and about 1.3 mm. wide, elliptic to
ovate-lanceolate, acute, a short semicircular first scale usually pres-
ent, the second and third scales about equal in length, acute, rather
sparingly pubescent with appressed hairs, 5 -nerved, the third scale
bearing in its axil a palet of equal length and a staminate flower,
the fourth scale about three fourths as long as the spikelet, elliptic,
mucronate, finely transverse-rugose.

Collected at Martin Pena January 31, 1899, by Heller, no. 380.
9. PASPALUM L. Syst. Ed. lo, 855. 1759

I. Racemes scattered along the axis of the inflorescence.

a. Spikelets broadly elliptic to orbicular, one half as wide as long ; rachis of
racemes manifestly wing-margined.
* Spikelets not margined.

Stems erect or nearly so, stout : leaf-blades long : outer scales of the
spikelet with a midnerve.
Spikelets glabrous.

Third scale of the spikelet white or yellowish at maturity : basal

Grasses of Porto Rico .''75

leaf-sheaths very broad and reticulated, at least when dry-
Spikelets about 1,5 mm. long : racemes numerous.

1. P. dcfisum.
Spikelets about 2.5 mm. long : racemes l2-l8.

2. P. UnJerwoodii.
Third scale dark brown at maturity : basal sheaths not broad

nor reticulated. 3. P. Boscianuin.

Spikelets pubescent.

Spikelets elliptic, about one half as wide as long, the pubes-
cence appressed. 4. P. Helleri.
Spikelets broadly oval or obovate, the pubescence spreading.
Spikelets less than 1.5 mm. long. 5. P. heviisphaerictim.
Spikelets 2 5-3 mm. long. 6. /'. virgatum.
Stems slender, creeping at the base : leaf-blades short, rarely more than
5 cm. long : outer scales of the spikelet without a midnerve.

7. P. orbiciihltuin.
** Spikelets surrounded by a broad toothed margin. 8. P. fiDibriatuni.
b. Spikelets narrowly elliptic, less than one half as broad as long : rachis of racemes
barely, if at all, margined. 9. P. Porloricense.

2. Racemes long and slender, arising in pairs at the apex of the stem, rarely with another
one a short distance below. 10. /-*. conjugatum.

1. P.\SR\LL'.M DENSU.M Poir. ill Lam. Encycl. 5: 32. 1804.

A single specimen of this species from Porto Rico has been
seen by the writer, and is the type in the herbarium of Lamarck
at Paris ; it was collected by Ledru.

2. Paspalum Underwoodii sp. nov.

A tall perennial with reticulated basal leaf-sheaths, elongated
leaf-blades and rather dense inflorescence. Stems 8-12 dm. tall,
erect, rather stout at the base: leaves numerous at the base;
sheaths glabrous, the basal ones strongly reticulated, at least when
dry, rather abruptly narrowed into the blade at or below the mid-
dle, equitant ; ligule a scarious ring ; blades 4-7 dm. long, about

1 cm. wide, long-acuminate at the summit, exceedingly rough on
the lower surface, densely hirsute on the upper surface, serrulate
on the margins: inflorescence 1.5—2.5 dm. long: racemes 12—18,
sessile or nearly so, ascending or nearly erect, the lower ones 7-10
cm. long, the rachis including its serrulate wings about i mm.
wide and more or less setiferous, usually purplish : spikelets in
pairs in two rows, on short hispid pedicels, yellowish green, usu-
ally more or less purplish tinged, glabrous, broadly obovate, 2.5
mm. long, 2 mm. broad, and about 0.8 mm. thick, apiculate, the

2 outer scales 3-nerved, apiculate, the third scale yellowish-white,
oval, 2.2 mm. long and about r.7 mm. wide, minutely striate.

Type collected by Underwood and Griggs along roadsides,

376 Nash : Enumeration of the

Mayaguez to Joyua, June 14 to July 22, 1901, no. 149. The fol-
lowing specimens also belong here : Plaza de Humacao, Eggers
6^6, 1881 ; between Mayaguez and Afiasco, Sintenis 1223, 1885,
and near Guanica, 3412/7, 1885, of the same collector; Santurce,
Heller 594 ; near Mayaguez, Heller 4368.

Related to P. denswn Poir., but that has more numerous ra-
cemes and much smaller spikelets.

3. Paspalum Boscianum Fluegge, Gram. Mon. 171. 18 10.

Paspalum piirpiirascens Ell. Bot. S. C. & Ga. i : 108. 1817.
El Yunque, Luquillo Mts., Wilson 43.

4. Paspalum Helleri sp. nov.

A tufted nearly glabrous perennial. Stems 4-8 dm. tall,
finally branching above, rather slender, smooth and glabrous, the
nodes glabrous : leaves usually 3 ; sheaths smooth and glabrous,
shorter than the internodes ; ligule a scarious ring about i mm.
long ; blades erect, or drooping at the end, smooth and glabrous
on the surfaces with the exception of some long hairs on the
upper surface just back of the ligule, the margins rough, linear,
acuminate at the apex, the larger ones 1.5-2.5 dm. long and 5-7
mm. wide : racemes 2-4, sometimes single on the branches, slen-
der, 6-9 cm. long, the rachis more or less flexuous, winged and
about I mm. wide : spikelets in pairs, 2 mm. long and about i
mm. wide, elliptic, obtuse, the two outer scales 3-nerved, the first
scale pubescent with long weak hairs, the second one glabrous, the
third scale and palet greenish.

In fields and waste places. Type collected at Santurce, Jan.
9, 1899, by Heller, no. 10. No. 164, secured by this collector at
the same place, also belongs here.

5. Paspalum hemisphaericum Poir. in Lam. Encycl. 5:31. 1804.
Paspalum paniculaUnn Gaertn. Fruct. Sem. 2 : 2, pi. So.

1 79 1. In part. Not Linnaeus.

Near Mayaguez, Heller 4399; Adjuntas road, near Ponce,
Heller 6227.

6. Paspalum virgatum L. Syst. Ed. 10, 855. 1759.

In sandy soil, common. Luquillo Mts., Wilson 227 ; Rio
Piedras, Heller 130 and 625 ; Cataiio, Heller 1373 ; Underwood
and Griggs 794.

Grasses of Pokto Rico ;}77

7. Paspalum orbiculatum Poir. in Lam. Encycl. 5: 32. 1804.
Paspaluin pusillnin Vent. ; Flueggc, Gram. Mon. 100. 18 10.
Wet banks, south of San Juan, Heller 664.

8. Paspalum fimbriatu-m H.B.K. Nov. Gen. & Sp. i : 93, //.

2S. 181 5.

In ditch, Coamo to Ponce, Underwood and Griggs 576 ; Ponce,
Heller 6219.

9. Paspalum Portoricense sp. nov.

A low tufted nearly glabrous slender perennial. Stems about
I dm. tall, glabrous, with the exception of a few hairs at the nodes,
simple : leaves 2 or 3 ; sheaths glabrous, usually shorter than the
internodes and shorter than the blades ; ligule .scarious, acute, 1-2
mm. long ; blades ascending, flat, rather thin, the upper surface
with some long hairs at the very base and a few scattered ones
elsewhere, sparingly ciliate on the margins, the lower surface glab-
rous, linear, acuminate at the apex, usually more or less narrowed
toward the base : inflorescence very slender, about 2 cm. long, of
I or 2 erect branches : spikelets in pairs, one on a long and the
other on a short pedicel, elliptic, 1.8 mm. long and 0.7 mm. wide,
usually purplish tinged, glabrous, the first and second scales
3-nerved, the third scale greenish-white, a little shorter than the
outer scales, obtuse.

Type collected between Aibonito and Cayey, February 8,
1899, by Heller, no. 524.

10. Paspalum conjugatum Berg. Act. Helv. 7 : i 29. //. S. 1772'
Near Mayaguez, Heller 4397 ; Mayaguez to Joyua, Under-
wood and Griggs 175 ; Luquillo Mts., Wilson 159.

10. ANASTROPHUS Schlecht. Hot. Zeit. 8: 681. 1850.

I. Anastrophus compressus (Sw.) Schlecht. ; Doell. in Mart. Fl.
Bras. 2- : 102. 1877.
Miliimi coinpi'cssum Sw. Prod. 24. 1788.
Rio Blanco, Blauner.

II. ISACHNE R. Br. Prod. 196. 18 10

I. Isachne angustifolium sp. nov.

A prostrate leafy branching perennial, forming large mats,
with narrow leaf-blades and rather small panicles. Stems i m.
long or less, glabrous or very sparingly pubescent below, the

378 Nash : Enumeration of the

uppermost internode minutely appressed-pubescent : leaves nu-
merous ; sheaths glabrous or sparingly hispid, papillose-ciliate on
the margins, those on the main stem shorter than the internodes,
those on the branches overlapping ; ligule a ring of short hairs ;
blades erect or ascending, linear to linear-lanceolate, acuminate,
rough and green on the upper surface, smooth and glaucous on
the lower surface, the lower the shorter, only 2-3 cm. long, in-
creasing in length toward the summit where they are 5-10 cm.
long and 4-6 mm. wide : panicles long-exserted, 5-8 cm. long,
their main axis and widely spreading primary branches minutely
pubescent, the secondary branches similarly pubescent and at an
angle of about 45° with the primary branches: spikelets about
1.5 mm. long, on shorter pedicels, the two outer scales about
equal in length, equalling or a little shorter than the spikelet and
sparingh' pubescent at the apex, obtuse, the first broadly oval, the
second almost orbicular, the two flowering scales glabrous, yel-
lowish, about 1.3 mm. long.

On the summit of El Yunque, Luquillo Mountains, Porto
Rico, Wilson, no. 160.

12. SYNTHERISMA Wait. Fl. Car. 76. 1788

Spikelets one quarter as wide as long or mrre, acute, the fourth scale yellowish when

mature. I. S. fijiibriatuni.

Spikelets one fifth as wide as long, acuminate, the fourth scale greenish when mature.

2. S. setosuin.

1. Syntherisma FiMBRiATUM (Link) Nash, Bull. Torrey Club, 25 :

302. 1898.
Digitaria fiinbriata Link, Hort. Berol. i: 226. 1827.
Waste places, common. Cataiio, Heller 107, and Bayamon,
Heller 422.

2. Svntherisma setosum (Desv.) Nash, Bull. Torrey Club, 25 :

300. 1898.
Digitaria setosa Desv.; Hamilt. Prod. 6. 1825.
Near Mayaguez, Heller 4398.

13. TRICHACHNE Nees, Agrost. Bras. 85. 1829

I. Trichachne insulare (L.) Nees, Agrost. Bras. 86. 1829.

Andropogon insiilaris L. Syst. Ed. 10, 1304. 1759-
Paniciim lauophaeinn H.B.K. Nov. Gen. & Sp. i : 97. 18 15.
In fields and along streams, frequent. Heller 136 and 650;

Underwood and Griggs 131 and 564; Goll 479; Adjuntas road,

near Ponce, Heller 6158.

Grasses of Porto Rico 379

14. DIMORPHOSTACHVS Fourn. Comptes-Rendus Acad.

Par. 80 : 441. 1.S75

I. DlMORPHOSTACHVS PEDUNCULATA (Poir.) Fouill. McX. PI. 2:

15. 1886.

Paspahun pcdnnculatuin Poir. in Lam. Encycl. Suppl. 4: 315.
1816.

AdJLintas road, near Ponce, Heller 6354.

15. PANICUM L. Sp. PI. 55. 1753

I. Apex of the panicle- branches not extending beyond the insertion of the spikelets.
a. Stems herbaceous.

Spikelets scattered, on long pedicels of varying length.
I Fourth scale of the spikelet smooth : low grasses, less than 5 dm. tall.

Panicle large and eftuse, over i dm. long : spikelets numerous, less
than 1.5 mm. long.
Leaf-blades narrowly ovate, rarely exceeding 7 cm. long :

spikelets I mm. long. I. P. hrevifolium.

Leaf-blades lanceolate, rarely less than I dm. long : spikelets
1.3 mm. long. 2. P. trichanthttm.

Panicle small, less than I dm. long : spikelets few, 1.5 mm. long or
more.
Stems erect, if prostrate in age never rooting at the lower nodes.
Spikelets less than 2 mm. long.

Leaf-blades with the upper surface hirsute, the larger

over I cm. wide. 3. P. comophylhim.

Leaf-b'ades with the upper surface glabrous or merely
puberulent, less than 5 mm. wide.

4. P. A'ashianutn.
Spikelets more than 2 mm. long. 5. P. consangiiineum.
Stems erect only toward the end, the base prostrate and rooting
at the lower nodes.
Leaf-blades lanceolate, pubescent, less than 5 mm. wide.

6. P. pai-vi/o/im/i.
Leaf-blades ovate-lanceolate, more than 5 mm. wide.

7. P. oplismenoides.
Fourth scale of the spikelet transversely rugose : tall stout grasses.

8. P. tnaxi»iuni.
Spikelets crowded on short nearly equal pedicels, in short clusters on one side

of the panicle-branches.
Outer scales of the spikelet with cross-veinlets connecting the longitu-
dinal nerves. 9. P. paniculattim.
Outer scales without cross-veinlets.

Fourth scale of the spikelet transversely rugose.

Spikelets about 2 mm. long : stems slender : leaf-blades ovate-
lanceolate, usually less than 5 cm. long.

10. P. prostratttm.

380 Nash : Enumeration of the

Spikelets about 3 mm. long : stems stout : leaf-blades long and
linear. 11. P. molle.

Fourth scale of the spikelet smooth. 12. P. laxitm.

b. Stems woody, bamboo-like. 13. P. latifoliiini.

2. Apex of the panicle-branches extending beyond the insertion of the spikelets in an
awn-like appendage.
Second scale of the spikelet 3-nerved, the fourth scale smooth.

14. P. Sintenisii.
Second scale of the spikelet 5-7-nerved, the fourth scale transversely rugose.

15- P. JJavescens.

1. Panicum brevifolium L. Sp. PI. 59. 1753.

In fields and waste places, frequent. Santurce, Heller 157;
Martin Pena, Heller 387; Luquillo Mts., Wilson 283; Adjuntas
road, near Ponce, Heller 6094.

2. Panicum trich antrum Nees, Agrost. Bras. 210. 1829.

In moist ground, between Aibonito and Cayey, Heller 531.

3. Panicum comophyllum sp. nov.

A tufted densely pubescent perennial, finally much-branched,
with broad flat leaf-blades and rather large panicles. Stems 3-4
dm. tall, hirsute with long ascending hairs : leaves 5—7 ; sheaths
densely hirsute with spreading hairs, the primary ones shorter
than the internodes and the blades ; ligule a ring of stiff hairs
about 2 mm. long ; blades spreading or ascending, flat, softly
hirsute on both surfaces with spreading hairs, those on the
upper surface the longer, ciliate on the margin with stiffer hairs ;
those on the main stem oblong-linear to ovate-lanceolate, acute,
barely cordate at the somewhat clasping base, 5—7 cm, long, 10-
13 mm. wide: panicle somewhat exserted, 5—7 cm. long and
about as broad, broadly ovate, the main axis pubescent with short
spreading hairs, its branches somewhat ascending : spikelets broadly
obovate, 1.5 mm. long and about i mm. wide, pubescent with
spreading hairs.

Type collected in rich soil at Santurce, January 9, 1899, by
Heller, no. 12.

Related to P. ciliosinn Nash, of the southern United States,
but readily distinguished by its softly pubescent leaf-blades.

4. Panicum Nashianum Scribn. Bull. U. S. Dept. Agric. Div.
Agrost. 7 : 79. 1897.

In sand, quite common. Vega Baja, Heller, 639 ; Manati to

Grasses of Porto Rico 381

Vega Baja, Underwood and Griggs 955 ; Santurcc, Heller 982^^
and 6442.

5. Paxicum CONSANGUINEUM Kunth, Rev. Gram, i : 36. 1S29.
Sandy soil, Santurce, Heller 982.

6. Panicum PARViFOLiUM Lam. 111. i: 173. 1791.
Bayamon, Sintenis 12 16.

7. Panicum oplismenoides sp. nov.

A smooth and glabrous prostrate leafy perennial with much
of the habit of Oplismciuis sctariits, with short broadly lanceolate
leaf-blades and spreading panicles. Stems slender, rooting at the
lower nodes, branching : leaves numerous ; sheaths much shorter
than the internodes and about one half as long as their blades ;
ligule a very narrow scarious ring ; blades spreading, shorter than
the internodes, broadly lanceolate to ovate-lanceolate, often some-
what undulate on the margins, 2-3 cm. long, 6-10 mm. wide :
panicles but little exserted, broadly ovate, 3-4 cm. long, the
branches spreading: spikelets elliptic, 1.8-2 mm. long and a little
less than i mm. broad, the first scale a little more than one half
as long as the spikelet, 5 -nerved, broadly ovate, obtuse, the second
and third about equal in length, 5 -nerved, a little exceeding the
white flowering scale, the third scale with a manifest palet nearly
as long as itself

Collected on the edge of a ditch at Vega Baja, May 9, 1899,
by Heller, no. 13 16.

Related in habit and general appearance to P. polygonoides
Lam., but in that the leaf-blades are smaller and with hispid
sheaths and the spikelets globular.

8. Panicum maximum Jacq. Ic. PI. Rar. i : 13. 178 1-6.
Martin Pena, Heller 377 ; Martin Pena and Lecheria, Goll 31.

9. Panicum paniculatum (L.).

Paspalum paiiicnlaUivi L. Syst. Ed. 10, 855. 1759-

Panicum fasciciilatiim Sw. Prod. 22. 1788.

In sandy and rocky soil, frequent. Rio Piedras, Heller 135 ;
Aguadilla, Heller 4528 ; Sprengel ; H. Wydler 3 i i ; Adjuntas
road, near Ponce, Heller 6226 and 6302.

It may be of interest to note here that the specimen from which
Sloane's figure (Hist. Jam. //. 72, /. 2) was drawn, and on which
Linnaeus based his Paspalum paniculatum, has been examined by

382 Nash : Enumeration of the

the writer, and that it is what is commonly known as Panicinn
fasciculatuiii Sw.

10. Panicum prostratum Lam. III. i: 171. 1791.
Cane field between Ponce and Coamo, Heller 497.

11. Panicum MOLLE Sw. Prod. 22. 1788.

Along railroad, Bayamon, Heller 100; Yauco, Heller 6293.

12. Panicum laxum Sw. Prod. 23. 1788.

In fields, common. Between Aibonito and Cayey, Heller
522; Santurce, Heller 983 ; Cataiio, Heller 1378.

13. Panicum latifolium L. Sp. PI. 58. 1753.

In woods and thickets, common. Maricao, Sintenis 215;
Mayaguez, Heller 4375 ; Arecibo, Heller 343 ; Luquillo Mts.,
Wilson 350; Coamo, Goll 699; Bayamon, Goll 227.

14. Panicum Sintenisii sp. nov.

A tufted glabrous perennial with long narrow flat leaf-blades
and a slender panicle. Stem 4-6 dm. tall, slender, producing
panicle-bearing branches from the upper sheaths : leaves about 4 ;
sheaths several times shorter than the internodes and the blades ;
ligule a ring of short hairs ; blades narrow and elongated, flat toward
the apex, but folded below for a considerable distance, at least when
dry, giving the blade a long-stalked appearance, 1-2 dm. long
and about 2 mm. wide at the flat portion, the lower folded portion
about one quarter this width, long-acuminate : panicle very slender,
4-7 cm. long, its branches appressed, 1-2 cm. long, each running out
into a bristle-like termination as do its divisions : spikelets about
2 mm. long, acute, as are also all the scales, the first scale i-ner\'ed,
about one half as long as the spikelet, the second one 3-nerved,
about two thirds as long as the spikelet, the third scale 5-nerved
and equalling the fourth scale or a little shorter than it, the fourth
scale very acute.

In woods, Guanica, Sintenis 3463.

Related to P. Cliapmani Vasey, but that has stouter spikelets
and all the scales with more nerves.

15. Panicum flavescens Sw. Prod. 23. 1788.
Near Mayaguez, Heller 4373 and 4488.

Grasses of Porto Rico 3S3

i6. ICHNANTHUS Heauv. Aprost. 56,/)/. 12, f. 1. 1S12
I. IcHNANTHUS FALLENS (S\v.) Muiiro ; Heiitli. Fl. Hongk. 414-
1861.
Paiiiiimi pa/lcNS ^w. Prod. 23. 178S.

In wet shaded places, common. Bayamon, Heller 92 ; San-
tiirce, Heller 826 ; Mayaguez, Heller 4374 and 4479 ; Puente
Muco. GoU 884; Adjuntas road, near Ponce, Heller 6093.

17. OPLISMENUS Beauv. Fl. Owar. 2: 14. /-/. ^cV. 1807
I. Oplismenl'S SETAKiis (Lam.) R. & S. Syst. 2: 481. 1817.

Panicum sctariuni Lam. 111. i: 170. 179^-

In moist soil, frequent. Pinon, Goll 623 ; Adjuntas road, near
Ponce, Heller 6303.

18. PXHIXOCHLOA Beauv. Agrost. 53,//. //, / //•

1812.

I. EcHiNOCiiLOA COLONA (L.) Link, Hort. Berol. 2: 209. 1833.
Paiiicuin coloHiiJii 'L. Syst. Ed. 10, 870. 1759-
In fields and waste places, common. Catano, Heller 108 ;

Mayaguez, Heller 4409 ; Coamo Springs, Goll 660.

19. SACCIOLEPIS Nash, in Britt. Man. 89. 1901.
I. Sacciolepis striata (L.).

Holcus striatus L. Sp. PI. 1048. 1753-
Panicum striatum Lam. 111. i: 172. 1791.
Panicum gibbinn Ell. Bot. S. C. & Ga. i : i 16. , 18 17.
Sacciolepis gibba (Ell.) Nash, in Britt. Man. 89. 1901.
Wet ground, near Santurce, Heller 1363.

20. CHAETOCHLOA Scribn. Bull. U. S. Dept. Agric.
Div. Agrost. 4: i"^. 1897

Setae at the base of the spikelet 5 or more, involucrate, aiitrorsely scabrous.

Setae .short, not more than twice as long as the spikelet : inflorescence usually
slender. ^- C. purpiirasceus.

Setae longer; inflorescence stout. 2. C. iinb^rbis.

Setae 1-3 at the base of each spikelet, not involucrate, retrorsely scabrous above,
antrorsely below. 3- C- scntidens.

384 Nash : Enumeration of the

1. Chaetochloa purpurascens (H.B.K.) S. & M. Bull. U. S.

Dept. Agric. Div. Agrost. 21 : 13. 1900.

Setaria piirpiirasccns H.B.K. Nov. Gen. & Sp. I : 1 10. 181 5.
In fields and waste places, frequent. Santurce, Heller 14 and

329-

2. Chaetochloa imberbis (Poir.) Scribn. Bull. U. S. Dept. Agric.

Div. Agrost. 4: 39. 1897.

Panicimi iuibcrbc Poir. in Lam. Encycl. Suppl. 4 : 272. 18 16.
Sintenis 208 {fide Scribn. & Merr. 1. c.) ; Adjuntas road, near
Ponce, Heller 6240.

Chaetochloa imberbis geniculata (Lam.) S. & M. Bull. U. S.
Dept. Agric. Div. Agrost. 21 : 12. 1900.

Paniciiin gcnicidahini Lam. Encycl. 4: 727. 1797.

Differs from the species in having stouter geniculate stems,
broader leaf-blades, and a longer inflorescence.

Humacao, Blauner 233.

3. Chaetochloa scandens (Schrad.) S. & M. Bull. U. S. Dept.

Agric. Div. Agrost. 21 : 17. 1900.

Setaria scandens Schrad., Schult. Mant. 2 : 279. 1824.
Sintenis 6498 {fide Scribn. & Merr. 1. c).

21. CENCHRUS L. Sp. PI. 1049. i753i

Involucres 4-6-flowered, armed at the base with erect barbed bristles.

1. C. echinatits.
Involucres 2-flowered, armed at the base with generally shorter reflexed spines.

2. C. tribidoides.

1. Cenchrus echinatus L. Sp. PI. 1050. 1753.

In sandy waste places, common. Santurce, Heller 3 and
1346; Cayey, Underwood and Griggs 406.

2. Cenchrus tribuloides L. Sp. PI. 1050. 1753.
Fajardo, Blauner 242.

22. STENOTAPHRUM Trin. Fund. Agrost. 175. 1820

I. Stenotaphrum secundatum (Walt.) Kuntze, Rev. Gen. 794.
1891.

Ischaeimun secimdatiun Walt. Fl. Car. 249. 1788.

Roadside near Bayamon, Heller i 241 ; in sand, Manati to Vega
Baja, Underwood and Griggs 959.

Grasses of Pokto Rico 385

23. OLYRA L. Syst. Kd. 10. 126 1. 1759
I. Olvra latifolia L. Syst. Ed. 10, 1261. 1759.
Near Mayagucz, Heller 4443 and 45^3-

24. PII.XKUS V. Hr. Hist. Jam. 344. 1756
I. Pharus Brasiliensis Raddi, Ai,^rost. Bras. 21. 1823.

Rio Blanco, Blauner ; ravine, Coamo Springs, Underwood and
Griggs 505 and 845.

25. ARISTIDA L. Sp. PI. 82. 1753
1. Aristida dispersa Trin. & Rupr. Mem. Acad. Sci. St. Petersb.
VI. 7 : 129. 1829.
Guanica, Sintenis 3766.

26. SPOROBOLUS R. Br. Prod. 169. 18 10

Stems tufted, not arising from rootstocks.

Spikelets about 2 mm. long, the second scale half as long or less.

1. S. Indicus.
Spikelets about 3 mm. long, the second scale fully as long. 2. S. purpurascens.

Stems arising from long branching rootstocks. 3. S. Virginicus.

1. Sporobolus Indicus (L.) R. Br. Prod. 170. 18 10.
Agrostis Indie a L. Sp. Pi. 63. 1753-

In sandy soil, common. Santurce, Heller 985 and 1275;
Lecheria, Goll 22.

2. Sporouolus purpurascens (S\v.) Hamilt. Prod. 5. 1825.

Agrostis purpurascens ^w. Prod. 25. 1788.
Near Mayaguez, Heller 4590.

3. Sporobolus Virginicus (L.) Kunth, Rev. Gram, i : 67.

1829.
Agrostis Virgifiica L. Sp. PI. 63. 1753.

On sand dunes and in sandy soil, frequent. Santurce, Heller
141 2 ; jNIaunabo, Sintenis 5084.

27. CHLORIS Svv. Prod. 25. 1788.

Spikes slender : flowering scale lanceolate, not long-ciliate. i. C. radiatit.

Spikes stout : flowering scale oval, long-ciliate on the margins above.

2. C. Paraguiensis.

I. Chloris radiata S\v. Prod. 26. 1788.

In moist soil and waste places, common. Catafio, Heller 109 ;

386 Nash : Enumeration of the

Rio Piedras, Heller 195 and 196; Bayamon, Underwood and
Griggs 883 ; Santurce, Heller 6380.

2. Chloris Paraguiensis Steud. Syn. PI. Gram. 204. 1855.

Cliloris barbatuni Sw. Fl. Ind. Occ. i : 200. 1797. Not
Atidropogon barbaUnii L. Amoen. Acad. 5 : 412. 1759.

Near Mayaguez, Heller 4417.

28. P:USTACHYS Desv. Nouv. Bull. Soc. Philom. 2 :

188. 1810

I. Eustachys petraea (Sw.) Desv. Nouv. Bull. Soc. Philom. 2 :
189. 1810.

In sand, Vega Baja, Heller 131 ; Santurce, Heller 6422.

29. HETEROSTECA Desv. Nouv. Bull. Soc. Philom.

2: 188. 1810

I. Heterosteca rhadina sp. nov.

Stems tufted, 2.5-4 dm. long, slender, finally prostrate at the
base and branching above : leaves several ; sheaths strongly nerved,
glabrous, or the lower ones sometimes more or less pubescent ;
ligule a narrow scarious ring ; blades flat, narrowly linear, acumi-
nate, more or less hirsute on both surfaces, especially on the upper,
less than i dm. long, 1-2 mm. wide : spikes 3-7, short, spreading
or reflexed, of 1-3 spikelets, the 3-angled rachis hispidulous on
the angles, its internodes long, usually exceeding one half the
length of the empty scales : spikelets distant, appressed to the
rachis, the purple empty scales very acute, i -nerved, hispidulous
on the stout midrib, the first shorter than the second which is 5-6
mm. long, the first flowering scale 3-nerved, the nerves extending
into short awns, the lateral ones barely reaching the apex of the
body of the scale, the central one extending about i mm. beyond
it, enclosing a perfect flower and a palet longer than its body, the
second flowering scale likewise 3-nerved, the nerves extending
into hispidulous awns longer than the body, enclosing a staminate
flower and a longer palet.

Collected on the Adjuntas road, near Ponce, by Heller, no.

6057, and distributed as Bouteloiia Americana (Sw.) Scribn. The

Aristida Americana cited in the synonymy of Bouteloua litigiosa

Lag. by Grisebach (Fl. Brit. W. I. 537)'is accredited to Swartz

(Obs. Bot. t. 2, f. 2). Swartz, however, distinctly indicates there

the real authorship of the species by quoting the original Linnaean

Grasses of Porto Rico -iH?

description in the Systema Naturae. Both authors refer the plant
to Jamaica, so that the plate and description of Swartz may in all
probability be safely taken as characterizing the Linnaean plant
also. The genus Hctcrostcca was. founded by Desvau.x on one
species, //. juncifolia. At a later date he probably became ac-
quainted with the Aiistida Aiiuricana Sw., for in a subsequent
publication (Jour, de Bot. 3: 68. 181 3) he changes the name to
HetcrosUca Americana, citing H. juncifolia as a synonym.

The plant of Swartz is quite different from the one we have
under consideration, differing in the more numerous spikes with
more spikelets, and in the second flowering scale being reduced to
a 3-awned empty rudiment.

30. ELF.USINE Gaertn. Fruct. i: 7. 1788
I. Eleusine Indic.v (L.) Gaertn. Fruct. i : 8. 1788.

Cy no sums Indie us L. Sp. PI. 72. 1753.

In cultivated and waste ground, common. Catano, Heller
no; between Ponce and Coamo, Heller 499 ; Mayaguez, Heller
4410; Lecheria, Goll 20.

31. DACTYLOCTENIUM Willd. Enum. 1029. 1809
I. D.\CTVLOCTENiUM Aegvptium (L.) Willd. Elnum. 1029. 1809.

Cynosurus Aegyptius L. Sp. PI. 72. 1753.

In cultivated and waste grounds, common. Santurce, Heller
155 ; Mayaguez, Heller 441 1 ; Guanica, Sintenis 3469 ; Guayama,
Goll, 513 ; Coamo Springs, Goll 661 ; Bayamon, Underwood and
Griggs 900.

12. LEPTOCHLOA Beauv. Agrost. 71,//. 15 J. i.

1812
I. Leptochloa viRGATA Beauv. Agrost. 166. 181 2.
Moist ground near Aiiasco, Heller 4535.

33. GYNERIUM H. c^- B. PI. Aequin. 2: 112,//. 113.

1809
I. Gynerium saccharoides H. & B. PI. Aequin. 2 : \\2, pi. 113.
1809.
Along streams, Yauco, Underwood and Griggs 718.

3^88 Nash : Enumeration of the

34. PHRAGMITES Trin. Fund. Agrost. 134. 1820

I. Phragmites occidentalis (Sieb.) Trin.; Steud. Nomen. Ed.
2, 2 : 324. 1841.

Arundo occidentalis Sieb.; Schult. Mant. 2 : 289. 1824.

Near Guanica, Heller 6289. This specimen is doubtfully re-
ferred here, its immaturity preventing a more definite determination.
It was distributed as Ariindo Donax ; it hardly appears to be that
plant, however. It may only be a form of the world-wide P.
Phragmites. Mature flowering specimens of this are greatly desired.

35. ERAGROSTIS Beauv. Agrost. 70,//. 14., f. 11. 1812

Flowers perfect : stems not rooting at the lower nodes.

Palet long-ciliate on the margins : panicle narrow and spike-like.

1. E. ciliaris.
Palet not long ciliate : panicle loose and open.

Annual : leaf-blades less than I dm. long. 2. E. Purshii.

Perennial : leaf-blades more than i dm. long. 3. E. ElHottii.

Flowers dioecious : stems creeping. 4. E. hypnoides.

1. Eragrostis CILIARIS (L.) Link, Hort. Berol. i: 192. 1827.
Poa ciliaris L. Syst. Ed. 10, 875. 1759.

In waste and cultivated soil, common. Bayamon, Heller loi ;
Santurce, Heller 156 and 6404 ; Cabo Rojo, Sintenis 838 ; Baya-
mon, Underwood and Griggs 895 ; Ponce, Heller 6218.

2. Eragrostis Purshii Schrad; Shlecht. Linnaea, 12 : 45 I. 1838.

Rio Piedras, Heller 197 ; between Aibonito and Cayey, Hel-
ler 551 ; Montellano, Goll 473; Adjuntas road, near Ponce,
Heller 6198. These specimens are doubtfully referred here.

3. Eragrostis Elliottii S. Wats. Proc. Am. Acad. 25: 140.

1890.
Poa nitida Ell. Bot. S. C. & Ga. i : 162. 18 17. Not Lam.
1791.

Eragrostis nitida Chapm. Fl. S. U. S. 564. 1 860.
Moist sandy soil, Santurce, Heller 584.

4. Eragrostis hypnoides (Lam.) B.S.P. Prel. Cat. N. Y. 69.

1888.
Poa Jiypnoides Lam. 111. 185. 1791-
Roadsides, Rio Piedras, Heller, 649 ; Bayamon, Sintenis 1225.

Grasses of Pokto Rico 3H9

36. UNIOLA L. Sp. PI. 71. 1753
I. Uniola racemifloka Trill. Hull. Sc. Acad. St. Petcr.sb. i : 68.
1836.
Near Salinas do Cabo Rojo, Sinteni.s 839/

37. ARTIIROSTYLIDIUM Rupr. Mem. Acad. St. Petcrsb.

VI. 5: 117. 1S39

Leaf- blades very narrowly linear or hair-like. I. A. capillifolium.

Lcaf-hlades lanceolate. 2. A. iniilthpualum.

1. Artiirostvlidium capillifolium Griseb. Mem. Am. Acad. II.

8: 531- ''"i^'S-
Near Sabanagrandc, climbing on tall trees, Sinteni.s 3891.

2. Arthrostylidium multispicatum Pilger in Urban, Symb. Ant.

2: 341. T901.

In woods near Maricao and Adjuntas, Sintenis 209 and 4016
(according to Pilger).

A number of other specimens of the Rambuseae are in collec-
tions, but they are without inflorescence and so undeterminable

New York Botanical Garden.

JPU13L.I0^4lTIO1VS

OF

The New York Botanical Garden

Journal of the New York Botanical Garden, monthly, illustrated, con-
laining notes, news and non-technical articles of general interest. Free to all mem-
bers of the Garden. To others, lo cents a copy ; $l.oo a year. [Not offered in
exchaYige.] Vol. I, 1900, viii -f- 213 pp. Vol. II, 1901, viii -|- 204 pp. Vol. Ill,
1902, viii -f- 244 pp.

Bulletin of the New York Botanical Garden, containing the annual reports
of the Director-in-Chief and other official documents, and technical articles embodying
the results of investigations carried out in the Garden. Free to all members of the
Garden; toothers, $3.00 per volume. Vol. I, No.s. 1-5, 449 pp., 3 maps, and 12
plates, 1896-1900. Vol. II, Nos. 6-8, 518 pp., 30 plates, 1901-1903.

Memoirs of the New York Botanical Garden. Price to members of the
Garden, ^I. 00 per volume. To others, $2.00. [Not offered in exchange.]

Vol. I. An Annotated Catalogue of the Flora of Montana and the Yellowstone
Park, by Dr. Per Axel Rydberg, assistant curator of the museums. An arrangement
and critical discussion of the Pteridophytes and Phanerogams of the region with
notes from the author's field book and including descriptions of 163 new species.
ix -f- 492 pp. Roy. 8vo, with detailed map.

Vol. II. The Influence of Light and Darkness upon Growth and Development,
by Dr. D. T. MacDougal, first assistant and director of the laboratories. An ac-
count of the author's extensive researches together with a general consideration of
the relation of light to plants. The principal morphological features are illustrated.
xvi -|- 320 pp. Roy. 8vo, with 1 76 figures.

Contributions from the New York Botanical Garden. A series of tech-
nical papers written by students or members of the staff, and reprinted from journals
other than above. Price, 25 cents each.

Vol. I. Inclusive of Nos. 1-25, vi -f- 400 pp. 35 figures in the text and 34
plates. ^5.00.

CURRENT NUMBERS 25 CENTS EACH.

No, 33. Studies in the Asclepiadacese, by Miss A. M. Vail.

No. 34. The genus Riella with descriptions of new species from North America
and the Canary Islands, by Dr. M. A. Howe and Prof. L. M. Underwood.

No. 35. The Pol)rporace3e of North America. — III. The genus Fames, by Dr.
W. A. Murrill.

No. 36. Studies on the Rocky Mountain Flora. — X, by Dr. P. A. Rydberg.

No. 37. Some Generic Segregations, by Dr. Per Axel Rydberg.

No. 38. The Polyporaceae of North America — IV. The genus Elfvingia, by
Dr. W. A. Murrill.

All subscriptions and remittances should be sent to

New York Botanical Garden

Bronx Park, New York City

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 40

THE PHYLLODES OF OXYPOLIS FILIFORMIS,
A SWAMP XEROPHYTE

By ROSINA J. RENNERT

NEW YORK
1903

[Reprinted from the Bulletin of the Torrey Botanical Club, 30 : 403-411. Jl, 1903]

The Phyllodes of Oxypolis filiformis, a Swamp Xerophyte*

P>v KosiNA J. Rennert

A large number of species of higher plants which live in ponds,
marshes and other extremely moist habitats present structural
characters of a xerophytic nature which have been most difficult
of interpretation. A special study of the phyllodes of Oxypolis
filiforviis t was undertaken by the author in the hope that some
additional light might be thrown on the subject. This work was
carried on in the New York Botanical Garden under the direction
of Dr. D. T. MacDougal, to whom I am indebted for valuable
advice and fruitful suggestions.

Oxypolis filiforviis is an umbcllifer which grows in swamps and
along the borders of ponds in the southeastern United States, as
far north as Delaware. The leaves exhibit curious modifications
by which they depart widely from the uniformly broad or widely
dissected umbelliferous type, being reduced to stiff grayish-green
awl-shaped phyllodes, which have a structure as if made up of a
number of sections of varying lengths, the joints being distinctly
noticeable and marking the position of peculiar septa in the interior
tissues. This habit of the leaves gives the plant a rush-like ap-
pearance, an effect which is heightened by the small size of the
cauline, and the strong development of the basal phyllodes. The
entire phyllode is flattened along one side. A minute pit is to be
observed externally to every septum, occurring in the middle oi
the slightly flattened surface.

Briquet % has given a detailed account of the results of his ex-
amination of this plant with respect to the anatomy of the phyl-
lodes. He describes the epidermis as a single layer of cells with
curved outer walls protected by a thickened outer membrane and
a folded cuticle. The stomata were noted as either flush with the

* Read before the Botanical Society of America, by invitation, at Pittsburg, Pa. ,
July I, 1902.

f Oxypolis filiformis (Walt.) Britton. Tiedeniannia teretifolia (Muhl. ) DC.
' \ Briquet, J. Etude de la feuille du Tiedmannia teretifolia J)C. Bull. Herb
Boiss. 5 : 461-465. 1897.

403

404 Rennert : The Phyllodes of Oxypolis filiformis

surface, or sunken slightly below, being about 48 to 54 fx in length.
Directly underneath the epidermis is a hypoderm composed of
small sclerotic cells with a small lumen. The hypoderm is one to
three layers in thickness, and is interrupted only underneath the
stomata. The chlorophyl-bearing tissue consists of three to six
layers of cells, the outer ones of which are palisaded. The inner-
most layers are more loosely arranged. Internal to the chloro-
phyl-bearing cells is a cylinder of four or five layers of large thin-
walled cells, designated as endoderm by Briquet. The fibro-vascular
bundles are situated in this tissue. The pericycle comes into con-
tact with the chlorophyl-bearing tissue in places, however. The
axial portions of the phyllodes were found to consist of thin-walled
stellate elements which are in contact at the tips of the rays. The
septa were found to be composed of sclerotic cells with numerous
branching canals extending radially through the walls. The fibro-
vascular bundles are unbranched in the intervals but break up at
the septa and form a mesh of anastomosing branches, which how-
ever do not traverse the septa radially to any great distance.

Briquet explains the seemingly anomalous possession of trans-
piration-hindering hypoderm by this marsh plant by the fact that
it is subject to two extremes of conditions. The mechanical ad-
vantages from the cylindrical form of the phyllodes, and the aerat-
ing capacity of the stellate tissue would be of advantage during
the early spring season, when the plant is more or less submerged.
The shape of the leaf would present a reduced surface and the
hypoderm would check transpiration during the mid-summer
season after the waters have subsided.

My own examination of this plant and the results of the
experimental tests made with it have revealed some structural fea-
tures not mentioned by Briquet, and lead to a somewhat different
interpretation of its adaptative features.

Some plants brought from Georgia and Florida in 1901 and
1902 were grown in the greenhouse under various conditions as
described below. The septa in the phyllodes of these specimens
were found to consist chiefly of small thin -walled cells containing
chlorophyl, and resembling the elements of the other chlorophyl-
bearing tissue, with only an occasional thick-walled sclerotic cell.
The endoderm was found to be continued on the surfaces of the

Rennekt : The Phvllodes of Owpolis filifurmjs 405

septa. I am not able to account for the difference in these septa
as observed by Briquet and myself, except to suggest that his
plants were grown under conditions which induced the sclerosis of
the elements of the septa and which he has not described.

The most important additional fact concerning the structure of
these organs, however, is that all of the specimens in the herbarium
of the New York Botanical Garden, and of those cultivated by my-
self in the open, exhibit the pits or depressions mentioned above,
and this feature is to be seen even in the first leaf of the seedlings.
These pits seem to be the external openings of oil or resin glands
which are situated immediately below each septum. The pit is
partly closed by a chlorophylless outgrowth of the lower margin,
which can be distinguished by the unaided eye as a minute scale.
Goebel * describes similar outgrowths on the phyllodes of Crantzia
linearis, an umbellifer of the same habit as Oxypolis, and despite the
fact that they occur along one side of the leaf only, he holds the
opinion that they are vestigial leaf-divisions. This conclusion can
hardly be valid with respect to the pits and scales of Oxypolis.
These formations are absent from the basal septa as they would be
if vestigial leaf-divisions, but as a matter of fact the scales and pits
originate on a line ninety degrees from the plane of the stipules.
Furthermore the specialized structute of the formations in question
militates against their acceptance as vestiges.

It was found that the floor of the pit is on a level with the
plane of the diaphragm. It is this floor and the flap which closes
the orifice only, whose structure indicates a specialization of func-
tion. With these exceptions the invagination is lined by a con-
tinuation of the normal epidermal cells of the phyllode, covered by
a cuticle continuous with that on the outside of the leaf. The floor
and flap have an epidermis of smaller cells the outer walls of which
are slightly convex, forming small papillae. No cuticle is present.
The small openings which occur in this epidermis measure about
19/y. in diameter and are formed by two cells broadly elliptical in
shape. Directly below these epidermal cells and making up the
greater part of the flap is a mass of irregularly shaped cells dove-
tailed compactly into each other. The epidermal cells as well as
the cells of this tissue contain a great amount of resin or oil. En-

* Goebel, K. Organographie der Pflanzen, 494. 1901.

406 Rennert : The Phyllodes of Oxypolis filiformis

larged tracheae are to be found in great numbers leading into this
compact tissue and forming a solid mass below it. The anasto-
mosing bundles of the septa end here. In the light of this peculiar

nrn

Fig. I. Oil-gland of phyllode, X I40-

structure, it seems reasonable to regard these pits as glands for the
secretion of oil or resin.

In opposition to Briquet, I am led to the conclusion that Oxy-

Rennert : The PiivLLonES of Oxvpoi.is filifokmis 407

polls does not owe its peculiar structure to an adaptation to wet
and dry periods, but that it is a true swamp or bog xerophyte in
accordance with the generahzation of Schimper, who supposes that
the water of bogs and swamps contains such a large proportion of
humic acid as to render absorption difficult. Plants growing in
such places therefore have difficulty in securing a requisite supply,
as if growing under arid conditions. Not all plants are capable of
absorbing water from a mixture exhibiting a higher pressure than
that of the cell-sap of the absorbing organs.

Dr. Shaw * has pointed out that the vegetation of a pond in
its final stage is always xerophytic and has drawn attention to the
fact that the continual deposit of organic material around the mar-
gin of a pond will produce a zone so much filled with humic acid
as to be entirely devoid of vegetation. Dr. Cowles f has upheld
this view of the xerophytic character of an undrained swamp and
has frequently noted the remarkable assemblage of xerophytic
adaptations such as leathery or hairy leaves and special structures
for water absorption exhibited by the flora of peat bogs. In con-
firmation of this standpoint Dr. Cowles has pointed out that if
plants growing in swamps of this type are xerophytic we ought
to find them thrving in dry regions. And this he reports to be
true ; Clcthra alnlfolla, Spli'aca touicntosa and Myrlca ccrlfera, gen-
erally conceded swamp plants, have been found thriving and in
all respects normal on dry hillsides.^

As Briquet has indicated, the most superficial examination of
the plant brings to light the unmistakably xerophytic character of
the reduced leaf-surface and the strong development of hypoderm.
On the other hand many of the characters which Briquet regards
as demonstrating the hydrophytic nature of the plant, seem when
viewed in the light of recent ecological work, to be adaptations
either for water storage or the prevention of transpiration. The
large thin-walled empty celled endoderm certainly exhibits capac-
ity for water storage. The thick cuticle and the waxy incrustation
of the epidermis, the accessory cells and the sunken position of
the stomata, and the presence of resin glands all point to the con-

*Shaw, C. H. The Development of Vegetation in the Morainal Depressions of
the Vicinity of Wood's Hole. Bot. Gaz. 33 : 437. 1902.
f Cowles, H. C. Bot. Gaz. 27 : 293. 1899.

408 Rennert: Th^ Phyllodes of Oxypolis filiformis

elusion that the conservation of water is the chief purpose met by
the striking structural adaptations.

This view is strengthened by the experimental results. Two
series of cultural tests were made. In one series a number of
plants in pots of soil were enclosed in bell-jars and were thus grown
in a saturated atmosphere. Others were entirely submerged in

Fig. 2. Partial cross-section of phyllode, xerophytic form, X I40-

tanks of water. The phyllodes produced under these conditions
were similar. In both cases these organs were delicate, hollow,
terete, translucent, light green in color, and exhibited no trace of
glands or pits. The stipules were membranous and reduced in
width. The epidermal cells were without wax or cuticle and were
greatly elongated longitudinally. The stomata were partly ex-
serted and were of a reduced number, but retained their character-
istic structure. In consequence of the regularity of the epidermal

Rennert : The Phyllodes of Oxvpous fii.iformis 409

rows of cells, the stomata occurred in definite rows separated from
each other by five sterile epidermal rows. Hypoderm was absent.
The chlorophyl-cells followed directly upon the epidermis. This
tissue was reduced to about three rows, and the palisade character
of the first row of cells was not very marked. The endoderm was
also less developed. It consisted of only one to three rows of
cells. The stellate tissue of the other form was entirely absent
here. As a consequence of the meager development of endoderm
the fibro-vascular bundles jutted out into the central cavity. The
xylem in the bundles was reduced and the position of the resin
duct changed. In the normal type it was always to be found out-
side of the phloem ; in these phyllodes it was on the other side of
the bundle near the xylem. This position of the duct has been

Fig. 3. Partial cross-section of phyllode, hydrophytic form, X I40-

pointed out by Van Tieghem as more primitive or juvenile. The
diaphragms were composed entirely of thin -walled cells and the
anastomosing bundles contained in them were much weaker and,
contrary to the habit in the solid phyllode, anastomosed as freely
at the center of the septum as about the edge. No trace of glands
in the phyllode have been found and there is consequently no
general convergence of bundles toward any single point in the
margin.

The phyllodes of the submerged plants possessed in all cases
the same structure as those grown under the bell-jar and exhibited
an additional modification in their method of propagation. New
plants were developed from the nodes of the xerophytic stems
which had been developed before the plants were submerged and
also in turn from the nodes of plants produced under water.

The water in which the immersed plants were grown presented

410 Rennert : The Phyllodes of Oxypolis filiformis

none of the difficulties of absorption of the liquid in bogs. These
submerged individuals as well as those in the bell-jars were seen
to undergo true aquatic adaptations consisting in the loss of hypo-
derm of the central stellate tissue and of the glands ; the reduction
of the chlorophyl-bearing elements, of the endoderm, and of the
number of stomata as well as of the xylem ; the change in the
position of the resin duct ; the modification of the epidermal cells^
stomata, cells of the septa and the course of the bundles through
the septa ; and the adoption of a new method of propagation.

These changes may be divided naturally into two classes, those
which involve a suppression of xerophytic characters and those
which are a means of adaptation to moist conditions and therefore
hydrophytic in their nature. In the first class we must consider
the loss of hypoderm and glands, the reduction of palisade-cells
and water-storage tissue of the endoderm and the loss of cuticle ;
while we find in the modification of the epidermal cells, the re-
duction and modification of the stomata and the adoption of a new
method of propagation types of structure peculiarly distinctive of
hydrophytes. If the action of water free from high concentration
of osmotic substances will induce such sweeping departures from
the normal, it seems but reasonable to conclude that the plant is
naturally a swamp xerophyte.

The adaptations described seem to have been most thoroughly
adopted by the species since the characteristic structure of the
phyllodes is to be found in the seedlings. These seedlings were
raised from seed developed by the plants with which the ex-
periments were made. The cotyledons were flat and linear in
shape and 5 to 6 cm. in length. When the first leaf developed it
exhibited the typical awl-shaped, septate character and bore glands
and the accompanying scales along one side throughout its length.
In internal structure it also resembled the later leaves in all re-
spects, except that the hypoderm was absent and that the cells of
the endoderm bore chlorophyl.

The chief results of importance brought out in the foregoing
paper may be briefly summarized as follows :

I. The leaf-structures of Oxypolis filifonnis are terete, awl-
pointed phyllodes with numerous septa or diaphragms, and this-
form is exhibited even by the first leaf of the seedling. The

Rennekt: The Phvllodes of Oxvpolis filiformis 411

structure of the septa as examined by myself differs notabl)' from
that described by Bricjuet.

2. The phvllodes normally bear numerous glands sunk in pits
which are edged with a minute scale. These glands are on a line
at 90*^ from the plane of the stipules and the attendant scales may
not be considered as vestigial leaflet structures as suggested by
Goebel as to the scales of Craiitzia.

3. Plants of 0 xy polls filif or uiis when grown in a saturated at-
mosphere or under the influence of normal soil-water (submerged)
exhibits marked hydrophytic adaptations of the phyllode. The
changes ensuing under such circumstances emphasize stili further
the xerophytic character of the phyllodes as indicated by their
anatomical character, and reenforce the conclusion that this species
should be regarded as a swamp xerophyte.

JPTJBLIC^VTIOIVS

OF

The New York Botanical Garden

Journal of the New York Botanical Garden, monthly, illustrated, con-
taining notes, news and non-technical articles of general interest. Free to all mem-
bers of the Garden. Toothers, lo cents a copy; ^l.oo a year. [Not offered in
exchange.] Vol. I, 1900, viii -f- 213 pp. Vol. II, 1901, viii -)- 204 pp. Vol. Ill,
1902, viii -f 244 pp.

Bulletin of the New York Botanical Garden, containing the annual reports
of the Director-in-Chief and other official documents, and technical articles embodying
the results of investigations carried out in the Garden. Free to all members of the
Garden; toothers, ;^3.00 per volume. Vol. I, Nos. 1-5, 449 pp., 3 maps, and 12
plates, 1896-1900. Vol. II, Nos. 6-8, 518 pp., 30 plates, 1901-1903.

Memoirs of the Nevsr York Botanical Garden. Price to members of the
Garden, ^ 1. 00 per volume. To others, ;^2. 00. [Not offered in exchange.]

Vol. I. An Annotated Catalogue of the Plora of Montana and the Yellowstone
Park, by Dr. Per Axel Rydberg, assistant curator of the museums. An arrangement
and critical discussion of the Pteridophytes and Phanerogams of the region with
notes from the author's field book and including descriptions of 163 new species.
ix -|- 492 pp. Roy. 8vo, with detailed map.

Vol. II. The Influence of Light and Darkness upon Growth and Development,
by Dr. D. T. MacDougal, first assistant and director of the laboratories. An ac-
count of the author's extensive researches together with a general consideration of
the relation of light to plants. The principal morphological features are illustrated.
xvi -|- 320 pp. Roy. Svo, with 176 figures.

Contributions from the New York Botanical Garden. A series of tech-
nical papers written by students or members of the staff, and reprinted from journals
other than above. Price, 25 cents each.

Vol. I. Inclusive of Nos. 1-25, vi -{- 400 pp. 35 figures in the text and 34
plates. $5.00.

CURRENT NUMBERS 25 CENTS EACH.
No. 33. Studies in the Asclepiadaceae, by Miss A. M. Vail.

No. 34. The genus /?iel/a with descriptions of new species from North America
and the Canary Islands, by Dr. M. A. Howe and Prof. L. M. Underwood.

No. 35. The Polyporacese of North America. — III. The genus Fames, by Dr.
W. A. Murrill.

No. 36. Studies on the Rocky Mountain Flora. — X, by Dr. P. A. Rydberg.

No. 37. Some Generic Segregations, by Dr. Per Axel Rydberg.

No. 38. The Polyporacese of North America — IV. The genus Elfvingia, by
Dr. W. A. Murrill.

All subscriptions and remittances should be sent to

NEW YORK BOTANICAL GARDEN

Bronx Park, New York City

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 41

THE POLYPORACEAE OF NORTH AMERICA-V.

THE GENERA CRYPTOPORUS, PIPTOPORUS,

SCUTIGER AND PORODISCUS

By WILLIAM ALPHONSO MURRILL

NEW YORK
1903

LReprinted from the Bulletin or thb Tokkky Botanical Club, 30 : 42a-434. August, 1903J

The Polyporaceae of North America.— V, The Genera Cryptoporus,
Piptoporus, Scutiger and Porodiscus

By William Ali'honso Murrill

The most important groups of the genus Fomes, as this term
is generally used, have been treated in the preceding papers of this
series. The plants heretofore considered are, for the most part,
perennial, with large stratified sporophores which grow conspicu-
ously upon the dead or dying wood infested by their very active
mycelium. The present paper deals chiefly with species found
under the genus Polyponis in Saccardo. These plants are annual
and their fruit-bodies are less conspicuous than those of the former
group, while their mycelium is usually comparatively limited in
extent. Most of the species here discussed are terrestrial and
somewhat fleshy, being allied to the Boletaceae. These terres-
trial species are exceedingly rare and beautiful in North America,
and present an interesting problem in distribution. The genus
Porodisais introduces the wood-loving forms of the genus Poly-
porus, which will form the subject of another article.

Cryptoporus Shear, Bull. Torrey Club, 29: 450. 1902
The name Cryptoporus was first used by Peck (Bull. Torrey
Club, 7 : 104. 1880) to designate a new section of the Placodermei
made necessary by the discovery of a species of Polyporus having a
volva. His description of the section is as follows :

" Pileus at first subcarnose, becoming harder or corky, sessile or
stipitate, the margins greatly prolonged beyond and beneath the

[The preceding number of the Bulletin, Vol. 30, No. 7, for July, 1903 (30 : 369-
422), was issued lO Jl 1903.]

423

424 MuRRiLL : Polyporaceae of North America

mouths of the pores and wholly or partly concealing them from
view, generally opening beneath by a small aperture ; pores not
stratose, the dissepiments thickened at the mouths and there dif-
ferently colored."

In raising this section to generic rank, Shear cites Peck's de-
scription and the description of the single species contained in the
section following, with a brief discussion of varieties and several
additional collections from new localities.

Cryptoporus volvatus (Peck) Shear

Polyporusvolvahis Peck, Rep. N. Y. State Mus. 27 : 98. 1877.

Polyporiis obvolut2is Berk. & Cooke, Grevillea, 7: i. 1878.

Fomes volvatus Cooke, Grevillea, 13: 119. 1884.

Cryptoporus volvatus Shear, Bull. Torrey Club, 29 : 450. 1902.

This peculiar plant was described from specimens collected on
Abies nigra. It is also found rarely on Abies halsaniea, but seems
to prefer species of pine, upon which it is quite common. I have
seen dead trunks of Pinus rigida filled with the mycelium, the
fruit bodies emerging through holes made by beetles. It is pos-
sible that the specimens found on charred wood in the West are
larger than our Eastern plants because no effort is required to
reach the surface through a heavy layer of cortex. Torrey's stip-
itate Western form figured by Gerard is duplicated at times in the
East when the plant grows on the top of a log. Henning's variety
Helix from California is fully described by Peck and differs in no
way from the type, which had not been seen when the variety was
described. Peck's varieties obvoliitus and Torreyi seem hardly
worthy of separation. P. evolvens Berk. & Cooke, Herb. Berk.
1879 from California, fastened to a sheet of P. evolutus B. & C,
is only P. volvatus and is very distinct from P. evolutus.

Exsicc. : California, Harkness, Torrey, McClatchie ; Colorado,
Crandall, Shear & Bessey ; Oregon, Shear, Carpenter; West Vir-
ginia, Nuttall ; New York, Murrill ; Michigan, Robbins ; British
Columbia, Macoun ; Delaware, Commons ; District of Columbia,
Shear, Hicks.

PiPTOPORUS Karst. Rev. Myc. 3: 17. 1881
This genus was originally established upon a single species,
Polyporus betulinus (Bull.) Fr., and has since remained monotypic.

Mukrill: Polyporaceae of North America 425

It is chiefly distinguished from nearly related genera by its separ-
able hymenium, which, upon investigation, is found to be cut off
from the context proper by a thin flesh-colored layer of mycelium
connected very loosely with the adjacent layers and allowing them
to separate easily. In age this layer disintegrates and the tubes
fall away in irregular masses, leaving the smooth white surface of
the context conspicuously exposed.

Piptoporus SU15ER0SUS (L.) Murrill

Boletus siibcrosus L. Sp. PI. 1176. 1753.

Boletus bctulinus Bull. Herb. France, //. ^12. 1786.

Polyporus betidinus Fr. Syst. Myc. i : 358. 1821.

Piptoporus betnlimis Karst. Rev. Myc. 3: 17. 1881.

Piptoponis siiberosus Murrill, Jour. Myc. 9: 94. 1 903.

Pileus fleshy to corky, compressed ungulate, convex above,
plane below, attached by a short umbo behind, varying to bell-
shaped when hanging from horizontal trunks, 5-30 X 5-20 x 2-5
cm. ; surface smoky, covered with a thin separating pellicle, gla-
brous, devoid of markings, cracking with age ; margin velvety,
concolorous, obtuse, projecting beyond the hymenium nearly a
centimeter : context fleshy-tough, elastic, homogeneous, 3 cm.
thick, milk white ; tubes separated from context by a thin pink
layer, 0.5 cm. long, 2-3 to a mm., sodden white; mouths very
irregular, dissepiments thicker than pores, obtuse, entire, crumb-
ling away in age, leaving the smooth white context : spores white,
cylindrical, curved, 4-5 n in length.

This plant is confined to species of birch, upon the dead or decay-
ing trunks of which it is quite abundant in the northern United
States and Canada. Its common occurence in Sweden explains why
it was noticed and described by Linnaeus in Flora Suecica as early
as 1745. From the following selected exsiccati it will be seen
that this species is as cosmopolitan as its hosts. France, Roiune-
gtiere 2102; Germany, Sydow 62 j ; Saxony, Krieger po/;
Sweden, Roniell 10, Murrill ; Russia, Jaczewski j6 ; Wisconsin,
Baker ; New Jersey, Murrill ; New York, Underwood, Murrill;
New Hampshire, Miss Minns, Grout.

Scutiger Paul. Icon. Champ, pi. ji.f. i-j. 1793
Seven species were included in this genus by Paulet, one of
them an agaric, four hydnums and two polypores. The first species

426 MuRRiLL : Polyporaceae of North America

listed and described is Scutigcr tiiberosns Paul, later changed to
Polyporiis asperdliis by Leveille, a well-known European species
belonging to the section Ovini of Fries and to the genus Polyporiis
of Karsten and Caloporus of Quelet. Albatrelhis of S. F. Gray,
another synonym of Scntiger, antedates the two last mentioned by
more than half a century. Plants belonging to this genus are
central-stemmed, fleshy-tough and terrestrial with white flesh,
rounded pores and smooth, hyaline spores. Some of the members
of this group approach the closely related genus Boletus, but are
firmer, tougher and more lasting.

As to distribution, the species of Scntiger are, in general, limited
to rather small areas ; few of them are common, some are rare and
local, and, so far as known, European and American species are
entirely distinct. The indications are that the plants lack vigor
and are easily affected by soil, climate and other environmental
conditions. Polyportis poripes Fr. was placed in this group before
the plant was fully known. Mature specimens show that it
belongs rather with multiplex forms such as Polyporiis frondosus
under the genus Grifola. A somewhat older stage of P. poripes
was described by Berkeley and Ravenel in 1872 as P. flavovirens.

Synopsis of tbe North American Species

1. Surface of pileus uneven, squamose or rugose. 2.
Surface of pileus smooth, tomentose or glabrous. 4-

2. Pileus sulfur-yellow, pleuropus, surface ornamented with imbricated floccose wart-

like scales, context white or yellowish, tubes small, angular, decurrent, white,
becoming greenish when wounded, yellowish when dry, spores 9X6//.

I. S. Ellisii.

Pileus brown. 3-

3. Tubes large, 1.5 mm. or more in diameter, hexagonal, surface of pileus smoky-

brown ornamented with darker imbricated tufts of appressed hairs, context
white, stipe excentric, its entire surface reticulate. 2. S. retipes.

Tubes small, 0.5 mm. in diameter, polygonal, decurrent, white, pileus reddish-
brown, rugose, stipe central, not reticulate. 3. S. dectirrens.

4. Pileus light-colored, white, red or blue. 5-
Pileus dark-colored, gray or brown. 7-

5. Pileus white, context white, tubes irregular, dissepiments thin, white, plants small,

growing upon grass roots, stipe short, dark-brown. 4. S. cryptoptis.

Pileus red, glabrous, stipe short, concolorous, context rose-colored, tubes short,
small, 1-2X0.2 mm., decurrent, orange-colored, spores ovoid, hyaline,
4X5-6/i. 5- S. laeticolor.

Pileus blue when fresh, changing to brown on drying. 6.

Murrill: Polvporaceae of North America 427

6. Tubes entire, becoming reddislibrown on drying, context ochraceous and pileus

and stipe reddislibrown in herbarium sp<-'cimens. 6. .V. taeruUoporus.

Tubes lacerate, fading to grayish-brown or dirty white, context nearly white, pileus

and stipe dull smoky-brown when dry. 7. S. holocyancus.

7. Stipe black and rooting. 8.
Stipe neither black nor rooting. 9.

8. Pileus smoky-brown, subtomentose, margin ihiri, intlexed, context white, lubes

regular, polygonal, entire, 2 mm. long, 0.5 mm. in diameter, stipe cylindrical,
light-brown above, black and rooting below, spores white, elliptical, 7X5/'-

8. S. radicatus.

Pileus drab-colored, nearly glabrous, margin thin, indexed when young, context

milk-white even wlien dry, tubes white, irregular, toothed, I mm. long, 0.25

mm. in diameter, stipe short, sooty-black as far as the decurrent tubes, attached

to buried wood, spores white, 3-4 X 5-7^- 9- S. sitbradicalus.

9. Pileus gray, glabrous or nearly so, margin very thin, context rosy-gray, soft, fleshy,

thin when drj', tubes small, 0.25-0.5 mm., unequal, decurrent, stipe short,

concolorous. 10. S. grisens.

Pileus brown. lo.

10. Stipe dark purple, very thick, pileus fulvous-brown, purplish at times, clothed with
short tomentum, margin very obtuse, context reddish beneath the cuticle, marked
when dry with a black concentric line limiting growth, tubes white, 2 to a mm.

II. S. petsicivus.

Stipe yellowish-brown, usually excentric, plants caespitose, pileus yellowish-brown,

pruinose, margin thin, context rose-tinted when dry, dark red next to the tubes,

which are small, 1-3 X 0.3 mm., decurrent, rose-colored when dry, the edges

fimbriate. 12. S. Whiteae.

I. Scutiger Ellisii (Berk.)

/'<:^/>y><?rwi- £7/w// Berk. Grevillea, 7 : 5. 1878.

Poly poms flavosquaniosiis Underw. Bull. Torrey Club, 24 : 84.
1897.

This plant was first collected by Ellis in low woods near New-
field, New Jersey. In 1886 only two specimens of it were known
to Ellis, the one just mentioned and one collected in South Caro-
lina by Ravenel. In 1897 it was found in Alabama by Mrs.
Earle and redescribed by Underwood as P. flavosquavtosus. The
specimen thus described was growing on clayey soil in pine woods
and w-as large and of advanced age, for it was in February that it
was found. Very young plants of this species were collected in
the same locality by Bradford in December, 1900, which measured
over three inches across, emitting when fresh a "strong unpleasant
odor." The indications are, judging from the meager knowledge
at our command, that this attractive species is at home in the gulf
states and grows but sparingly above the Carolinas.

428 MuRRiLL : Polyporaceae of North America

2. Scutiger retipes (Underw.)

Polyponis retipes Underw. Bull. Torrey Club, 24 : 85. 1897.

The type collection of this species was made in Alabama in
1897 by Mrs. F. S. Earle. The plants were found in autumn
growing on the ground in pine woods. Collections were made
later in Alabama by Earle and Baker. In color and stature this
species resembles specimens of Polyponis ovinus sent from Upsala
to Kew, but the pores are very much larger and the species are
very distinct.

3. Scutiger decurrens (Underw.)

Polyponis decurrens Underw. Bull. Torrey Club, 24: 83.

1897.

This is a very distinct species, about which little is known.

The type plants were collected by Dr. Hasse near Pasadena, Cali-
fornia, growing in the soil on the side of a caiion. The collection
was made in February, 1896, and no new collections of the spe-
cies have been reported since.

4. Scutiger cryptopus (Ell. & Barth.)

Polyponis cryptopus Ell. & Barth. Erythea, 4: 79. 1896.

A very distinct and interesting species found by Bartholomew
in Kansas attached to dead grass roots in sandy pastures. The
name refers to its habit of lying flat on the ground and thus con-
cealing the stipe, in which position the pilei look like small whitish

discs.

5. Scutiger laeticolor sp. nov.

A thick central-stemmed plant of rather large size with a
smooth red pileus and orange-colored tubes. Pileus circular in
outline, depressed at the center, 10 x 10 X i cm. ; surface gla-
brous, brick-colored to purplish-red, margin acute, at first inflexed,
irregularly undulate : context fleshy-tough, pale rose-colored,
homogeneous, 0.75 cm. thick ; tubes 1-2 mm. long, 5 to a mm.,
decurrent, rather dark orange throughout, mouth subcircular or
polygonal, dissepiments thin, toothed at times, especially when
decurrent : spores ovoid, smooth, hyaline, copious, 4 x 5-6 p. :
stipe short, thick, increasing upward, 2 x 1-2.5 cm., resembling
the pileus in color, but solid, firm and tough, with darker flesh.

The above description is made from a rather old and some-
what faded dried specimen collected by Underwood in Alabama in

Murrii-l: Polyporaceae of North America 429

November, 1895, and now in the Underwood herbarium. Al-
though the description is on this account unsatisfactory, the plant
is nevertheless very distinct and must be highly attractive when
seen at its best. The only other plant of this species that I have
seen is one at Kew sent by Ravenel under the name P. confluens.
This specimen however, does not figure in my description, since I
have made no special study of it. Like most of the other mem-
bers of the genus, this species is doubtless rare and probably con-
fined to the states south of Virginia. The appearance of the stipe
indicates that it is terrestrial, growing unattached in soil rich in
humus. The name I have chosen for it refers to its brilliant color-
ing throughout.

6. Scutiger caeruleoporus (Peck)
Poly poms caeruleoporus Peck, Rep. N. Y. State Mus. 26 : 68.

1874.

Exsicc. : New NoxV, Peck, Ellis; Vermont, Morgan; North
Carolina, Atkinson.

This rare and beautiful species is found on the ground in woods
during midsummer. In sheltered spots the whole plant is exter-
nally grayish-blue in color, the blue fading gradually to gray upon
exposure. Specimens have been found with one half of the pileus
exposed and faded, while the other sheltered half was normal. In
drying, the blue color disappears, being retained longer in the
hymenium than in the surface of the pileus. It was from par-
tially faded specimens that the plant was first described. Herba-
rium specimens in the Ellis collection arc fulvous with darker hy-
menium and ochraceous flesh. How long the grayish-blue color
remains after drying I cannot tell.

7. Scutiger holocyaneus (Atk.)
Polyponis holocyaneus Atk. Jour. Myc. 8: 117. 1902.
This species was collected near Blowing Rock, North Caro-
lina, by Atkinson and Troyer in September, 1901. Several plants
were found growing gregariously on the ground beneath coniferous
trees. When fresh they were entirely blue, but in drying the color
changed to dull brown. In this and several other respects the
species resembles Scutiger caeruleoporus, from which, however, it
is quite distinct.

430 MURRILL : POLYPORACEAE OF NoRTH AMERICA

8. Scutiger radicatus (Schw.)

Poly poms radicatus Schw. Proc. Acad. Sci. Phila. 4: 155.

1834.

Polyporus Morgani Peck. Rep. N. Y. State Mus. 32 : 34.

1879-

Polyporus Kansensis Ell. & Barth. Erythea, 4:1. 1896.

Exsicc. : Ohio, Lhyd, Morgan ; Indiana, Underwood ; New
York, Ovcracker ; Canada, Deariicss, Macon n ; Kansas, Barthol-

oineiv.

This species is found from September to November in woody
earth or on much-decayed wood about stumps or dead trunks. It
is large and conspicuous and is quite easily recognized by its
brown cap, large pores and long stipe, which is black and radicate
at the base. The species ranges from Pennsylvania north to
Canada and west to Ohio and Kansas, where it has been twice
renamed. A specimen from Washington agrees with this plant
in many respects, but has a darker, rougher pileus and firmer sub-
stance, with some differences in pore-structure. I hesitate, how-
ever, to describe it as new without more material, since 5. radicatus
is somewhat variable even in its type locality.

9. Scutiger subradicatus sp. nov.

A rather large thin plant with light brown almost glabrous
surface, small white serrated tubes and short black stipe. Pileus
irregular in outline, convex to plane, 12 x 9 x 0.5 cm. ; surface
fibrillose, drab-colored to isabelline, margin very thin, inflexed
when young, irregularly undulate at maturity : context fleshy-
tough, 1-7 mm. thick, pure milk-white even when dry; tubes
mere areoles at first, short and small at maturity, scarcely i mm.
in length, 3-4 to a mm., decurrent to the blackened part of the
stipe, white, yellowish when dry, mouths polygonal, regular, at
length much elongated by confluence or otherwise irregular, edges
thin, toothed or fimbriate when mature : spores ovate to ellipsoidal,
smooth, hyaline, not abundant, 3-4 x 5-7 H : stipe short, thick,
central, tapering and attached at the base, sooty-black up to the
pores, 4 X 2.5 cm. ; context milk-white, firm, fleshy-tough, sur-
face minutely tomentose, rugose-reticulate when dry.

Two young plants of this species were collected by Dearness at
London, Canada, in September, 1896, and a finely developed plant
was recently found in New York by Mrs. M. L. Overacker. The

Murrill: Polyporaceae of North America 431

three plants form a perfect series in development. All grew at-
tached to decaying wood beneath the soil. The pores are not
rounded and punctiform when very young, but large and areolate
like those of Menilius, the short walls being built up from these
ridges like the cells of a honey-comb without increasing in diam-
eter. At first glance, one would say that this species is a peculiar
form of .v. radicatiis, but a little investigation shows that it is quite
distinct. Because of this superficial resemblance, however, I have
assigned to it the name given above,

lo. Scutiger griseus (Peck)
Polyponis griseus Peck, Rep. N. Y. State Mus. 26 : 68.

1874.

Polyponis Ear/iiV nderw. Bull. Torrey Club, 24: 84. 1897.

Exsicc: New York, Per /c; l<iev/ Jersey, P//is, StcvL/isou ; Ala-
bama, Earlc.

This species has been found on the ground in open deciduous
woods in New York and New Jersey and in pine woods in Ala-
bama. It is the only member of the genus that has been collected
in any quantity. Its place in the American flora is very much the
same as that occupied by P. Iciicomclas in Europe, though the two
species are certainly distinct.

1 1. Scutiger persicinus (Berk. & Curt.)
Poly poms pcrsicinus Berk. & Curt. Grevillea, I : IJ. 1872.
Little is known of this plant beyond the original description in
Grevillea and a longer one by the same authors in the Centuries of
North American Fungi. It was first found by Ravenel in South
Carolina growing at the base of trunks in pine woods. A study
of the type at Kew shows it to be a very distinct species, while
its large size and dark brown or purple color must make it con-
spicuous among related species when seen in the fresh state.
While bearing an outward resemblance to P. Sclnocinitzii, its con-
text and pores are white when fresh, changing to gray when dried,
and this color is uniform, except for a black line showing in cross-
section as a limit of growth for a certain period. The margin of
the pileus is very obtuse and lobcd or quite irregular from con-
fluence.

432 MuRRiLL : Polyporaceae of North America

12. Scutiger Whiteae sp. nov.

A plant of medium size, growing in clusters, with yellowish-
brown pileus and tubes that are white when fresh and rose-colored
when dry. Pileus subcircular in outline, convex, depressed at the
center, 8-12 x 0.5-1 cm.; surface pruinose, velvety to the touch,
isabelline to fulvous, margin acute, at first inflexed, irregularly
undulate at maturity: context fleshy-tough, 0.2-0.5 cm. thick, of
nutty flavor, rose-tinted when dry, dark red next to the tubes ;
tubes 0.1-0.3 cm. long, 3 to a mm., very decurrent, white when
young and fresh, rose-colored when bruised or dried, mouths cir-
cular or subcircular, edges thin, fimbriate : spores ovoid, hyaline,
smooth, copious, 3.5 x 5/^: stipe short, usually excentric, en-
larged at the base, 3 x 2-zi. cm., concolorous, tough : plants caes-
pitose, often confluent at the base, all stages of development being
found in one cluster.

A fine collection of this species was made by Miss V. White
at Bar Harbor, Maine, in August, 1901. The plant was found
growing among moss on a damp river bank and seemed to her
to resemble Hydmnn rcpandnm in general habit. Accompanying
the specimens are excellent field notes and drawings in water
colors.

Plants sent to Kew as P. ovinus by F"arlow from New Hamp-
shire seemed to me to agree with Miss White's specimens, but no
special study of them was made. A few other stray American
plants in foreign herbaria lead me to think that this species may
possibly range as far south as Ohio, but that its proper home is
in New England.

Porodiscus gen. nov.

Hymenophore small, annual, tough, epixylous, erumpent from
the lenticels of dead branches ; stipe attached to the vertex of the
pileus, usually curved at maturity ; context white, fibrous, tubes
cylindrical, short, one-layered, mouths constricted ; spores globose,
smooth, hyaline.

The type of this genus is Polyporiis pocula (Schw.) Berk. &
Curt., first described by Schweinitz as Pcziza pendida and later as
Sphaeria pocula. By Fries it was first assigned to the genus
CypheUa and later formed the basis of the new Friesian genus
Enslinia, which name, however, was preoccupied by Enslinia
Rchb. The fruit body matures slowly and resembles a discomy-
cete in its early stages, hence the confusion in regard to its sys-

MURRILL : POLYPORACEAE OF NoRTH AMERICA 43.'>

tematic position. The name I have chosen for the genus refers to
this resemblance. There is only one species known.

Porodiscus pendulus (Schw.)

Peziza poidtila Schw. Syn. Funj:^. Car. 92. no. 1183. 1818.

Cvphclia piudii/a Fr. Syst. 2: 203. 1822.

Sp/iacria pocitla Schw. Proc. Acad. Sci. Phila. 4 : 1 89. pi. 2. f.
6. 1834.

Etisliiiia pocnla Yx. Summ. Veg. Scand. 2: 399. 1849.

Polyp07'us pocula Berk. & Curt. Proc. Am. Acad. Arts & Sci. 4:
122. 1859.

Polyporiis cjipidacforniis Berk. & Curt. Grevillea, l: 38. 1872.

This species has been much confused because of its similarity
to members of other groups. Schweinitz at fir.st called it Pedza
digitalis Alb. & Schw. (Consp. Fung. 315, no. 943. //. 5. f. i.
1805), a European plant which it outwardly resembles, but in pub-
lishing it he adds "nisi nova, minor, durior, pcitdnla a me nomi-
nata." That this name pcndula was unpublished previous to this
time is indicated by the fact that he cites this as the place of its
publication under Cyphella pcndula in Proc. Acad. Sci. Phila. 4 : 184.
1834. That it was in common use may be judged from the fact
that Fries got it in manuscript from Schwaegrichen, a friend of
Schweinitz and the editor of his work on the fungi of Carolina.

When Schweinitz received more mature specimens of the same
species from Dr. Torrey collected on branches of ash, he named it
Sphacria pocula, at the same time stating that the plant was entirely
similar to his Cyphella pcndula, except for the presence of peri-
thecia. A southern form on sumac was later named Polyporus
cupulaefonnis by Berkeley and Curtis. The confusion that has
arisen is partly due to the small size of the plant and the changes
that take place in it as it develops. On emerging from the bark
of its host it is erect, sessile and depressed at the center without
the appearance of a Polyporus. Later, the central depression is
occupied by a porous hymenium and a stipe develops, which varies
in length and is more or less curved according to its position on
the branch. The surface could hardly be called tomentose, but is
uniformly covered with a brown powder, which bleaches and
partly disappears in age.

434 MuRRiLL : Polyporaceae of North America

As to locality and host, this species shows little choice, being
found on dead branches of chestnut, oak, hickory, sumac, red
cedar and other trees from New England to Florida. I have
found it very common on dead branches pruned from chestnut
trees in the woods and parks about New York City. Some of
these branches are entirely infested with the fungus, even to the
smallest twigs. According to Schweinitz, it was also common on
chestnut around Bethlehem, Pennsylvania.

Specimens examined: Rav. Fung. Car. lo, 210, Ell. & Ev.
N. A. Fung. 308, 2728, Rabenhorst- Winter Fung. Eur. 3328 (from
America) ; Georgia, Ellis ; South Carolina, Rav end ; Florida, Cal-
kins ; West Virginia, Niittall ; Missouri, Demetrio ; Pennsylvania,
Ellis ; New Jersey, Ellis, Mjirrill ; Delaware, Commons ; Connect-
icut, Thaxtcr.

New York City.

JPXJBLIC^^TIOIVS

OF

The New York Botanical Garden

Journal of the Nevr York Botanical Garden, monthly, illustrated, con-
taining notes, news and non-technical articles of general interest. Free to all mem-
bers of the Garden. Toothers, ID cents a copy; ^l.oo a year. [Not offered in
exchange.] Vol. I, 1900, viii -f- 213 pp. Vol. II, 1901, viii + 204 pp. Vol. Ill,
1902, viii -f 244 pp.

Bulletin of the New York Botanical Garden, containing the annual reports
of the Director-in-Chief and other official documents, and technical articles embodying
the results of investigations carried out in the Garden. Free to all members of the
Garden; toothers, ^3.00 per volume. Vol. I, Nos. 1-5, 449 pp., 3 maps, and 12
plates, 1896-1900. Vol. II, Nos. 6-8, 518 pp., 30 plates, 1901-1903.

Memoirs of the New York Botanical Garden. Price to members of the
Garden, ^i.oo per volume. To others, ;?2.oo. [Not offered in exchange.]

Vol. I. An Annotated Catalogue of the Flora of Montana and the Yellowstone
Park, by Dr. Per Axel Rydberg, assistant curator of the museums. An arrangement
and critical discussion of the Pteridophytes and Phanerogams of the region with
notes from the author's field book and including descriptions of 163 new species,
ix + 492 pp. Roy. 8vo, with detailed map.

Vol. II. The Influence of Light and Darkness upon Growth and Development,
by Dr. D. T. MacDougal, first assistant and director of the laboratories. An ac-
count of the author's extensive researches together with a general consideration of
the relation of light to plants. The principal morphological features are illustrated,
xvi -f- 320 pp. Roy. 8vo, with 176 figures.

Contributions from the New York Botanical Garden. A series of tech-
nical papers written by students or members of the staff, and reprinted from journals
other than above. Price, 25 cents each.

Vol. I. Inclusive of Nos. 1-25, vi -|- 400 pp. 35 figures in the text and 34
plates. $5.00.

CURRENT NUMBERS 25 CENTS EACH.
No. 33. Studies in the Asclepiadaceae, by Miss A. M. Vail.

No. 34. The genus Riella with descriptions of new species from North America
and the Canary Islands, by Dr. M. A. Howe and Prof. L. M. Underwood.

No. 35. The Polyporacese of North America. — III. The genus Fonies, by Dr.
W. A. Murrill.

No. 36. Studies on the Rocky Mountain Flora. — X, by Dr. P. A. Rydberg.

No. 37. Some Generic Segregations, by Dr. Per Axel Rydberg.

No. 38. The Polyporacese of North America — IV. The genus Elfvingia, by
Dr. W. A. Murrill.

All subscriptions and remittances should be sent to

New York Botanical Garden

Bronx Park, New York City

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 42

A REVISION OF THE FAMILY
FOUQUIERIACEAE

By GEORGE V. NASH

NEW YORK
1903

[Reprinted from the Bulletin of the Torrky Botanical Club, 30 : 449-459. Auguet, 1903]

A Revision of the Family Fouquieriaceae

By George V. Nash

During an attempt to identify one of the species of Fougiticiia
which flowered in the conservatories of the New York Botanical
Garden during the past June, much confusion was found to exist,
both in the identification of herbarium material and in the litera-
ture bearing upon this family. So great was this confusion that
the writer was eventually compelled to extend his examinations
beyond the point he had anticipated, and finally realized that a re-
vision of the family was necessary before any definite results could
be obtained. No recent treatment, involving a consideration of
the species, could be found, and the apparent need of such work
encouraged the author to enter upon the following revision, which,
it is hoped, will throw some light upon an interesting family and
one but little understood. As here regarded, it embraces two
genera and seven species, three of which are here described for the
first time.

The relationship of the family is rather puzzling, and a satis-
factory solution of the difificulty has not as yet been proposed.
By Bentham and Hooker (Gen. PI. I : i6i) it was made a tribe
of Tamaricaceae, and the same treatment was accorded to it by
Engler and Prantl (Nat. Pfl. 3'' : 298). Subsequently PIngler
(Nat. Pfl. Nachtr. 251) maintained that the family was better kept
separate from Tamaricaceae on account of its oily endosperm and
gamopetalous corolla. Its distribution would also tend to confirm
this distribution of the group, for Tamaricaceae, with Fouquieria-
ceae removed, is strictly Old World, while Fouquieriaceae itself is
confined to North America, and primarily to its arid regions.

While Engler changed the rank of this family, he indicated no
change in its relationship. It is hardly within the scope of a re-
vision of this kind to discuss at length a matter of this nature, but
the strong resemblance in many ways to certain forms of the Pole-
moniaceae cannot be passed by without some comment. In this
connection it is well to remember that it was in this family, as a
Cantiia, that the original species was published. The 3-celled

449

450 Nash : Revision of the Family Fouquieriaceae

ovary, the more or less united styles and the gamopetalous tubu-
lar corolla, to the base of which the filaments are slightly adnate,
all markedly point in this direction. The spongy central column
found in the dehiscing capsules would indicate that the inner edges
of the septa unite. A transverse section of the ovary made at this
time, and before the septa break away from the walls, would much
resemble the condition of things existing in the genus Gilia, also
found in the same region. The general resemblance of the flow-
ers to those of some of the large-flowered red Gilias is perhaps the
most striking feature of the plants. It is true that the sepals in
this are distinct, while in Gilia they are more or less united, but
this is but one character weighing against many others. It would
seem to me that the ovarian, placental, style and corolla charac-
ters indicate a strong affinity with that family, much stronger than
with the Tamaricaceae, to which family, polypetalous in all its
other forms as are its immediate relatives, the gamopetalous corolla
of Fouquieriaceae does violence.

FOUQUIERIACEAE DC. Prod. 3 : 349. 1828.

Shrubs or trees, the trunks simple, columnar and stout, or
much branched : branches spine-branching. Spines developed
within the petioles of the leaves on the new growth and becoming
apparent when these fall. Leaves with the blades flat, entire, or
rarely obcordate or emarginate : those on the young growth
petioled, the petioles from one half as long as to equalling the
blade. Secondary leaves borne in fascicles in the axils of the
spines and sessile or nearly so. Inflorescence spicate to paniculate,
terminal. Flowers sessile or pedicelled. Calyx of five free im-
bricated sepals. Corolla yellow or red, hypogynous, the segments
united for one half their length or more : tube cylindric, some-
times broadened toward the apex : lobes flat or concave, imbri-
cated in the bud, incurved, erect, spreading, reflexed or enrolled.
Stamens ten to fifteen, of unequal length, exserted from the corolla-
tube, and adnate to it at the very base : filaments subulate, usually
broadened and compressed dorsally at the more or less pubescent
base where they are sometimes coherent, sometimes provided
with a scale-like appendage near the base : anthers acute at the
apex, cordate at the base, elliptic to nearly ovate, introrse, versa-
tile, attached below the middle, 2-celled, the cells opening longi-
tudinally. Ovary 3-celled, the inner edges of the septa united at
the base and at the very apex, free in middle, the septa finally

Nash : Revision of thk Family Fouquiekiaceae 451

uniting by the inner edges and breaking away from the walls,
forming a central column attached at the apex and at the base.
Ovules 4-6, in two rows, in each cell, borne on the edge of the
free portion of the septa. Styles 3, slender or stout, included in
or exserted from the corolla, united only at the base or for their
entire lengtli. Capsule dehiscing loculicidally, 3-valved, the
valves thick and coriaceous. Seeds oblong, comjircssed, at first
broadly winged, the wing finally breaking up into long filaments
similar to those on the body of the seed. Albumen thin, oily.
Cotyledons flat, oblong to ovate, cordate at the base. Radicle
short.

Genera two, both natives of Mexico, one, Fouqitkria, also found
in the adjacent parts of the United States. The Idria of Kellogg
is often united with Fouqiiicria, but it seems more natural to keep
it separate, both on account of habital characters and differences
in the flowers. The short stout included styles united for their
entire length, yellow flowers and columnar, normally unbranched
trunk, justify this treatment.

Styles more or less united, but free at the apex, exserted, the column and branches
slender: shrubs or trees with branching trunk : corolla red. I. Fouqtiieria.

Styles wholly united, stout, short, included, forming a 3-angled stout body: tree with
a stout columnar undivided trunk : corolla yellow. 2. IJria.

I. FOUQUIERIA H.B.K. Nov. Gen. & Sp. 3 : 452. 1820

Bromiia H.B.K. Nov. Gen. & Sp. 6 : 83. 1823.

PJiUctaena'U€<ym. Vidensk. Selsk. Skr. V. 2 : 283. pi. 185 i.

Shrubs or trees with spicate or paniculate inflorescence and red
sessile or pedicelled flowers. Sepals less than one half as long as
the corolla-tube. Corolla red, cylindric to campanulate, the seg-
ments united into a tube for one half their length or more, the tube
sometimes gradually enlarged toward the apex : lobes from erect
to recurved or enrolled. Stamens ten to fifteen, more or less
broadened and compressed dorsally at the pubescent base. Styles
slender, exserted, united only at the base, or nearly to the apex.

Species six ; Mexico and the adjacent parts of the United
States.

Inflorescence corymbiform-paniculate. I. F. fasciculata.

Inflorescence conic to elongated paniculate, or spicate.

Corolla-tube cylindric, several times longer than the lobes.

Inflorescence spicate. 2. F. formosa.

Inflorescence paniculate.

Filaments unappendaged : panicle conic to ovate : bushy shrubs or trees.

452 Nash : Revision of the Family Fouquieriaceae

Corolla-tube 3-4 mm. in diameter : sepals oval : panicle-branches
slender. 3. F. Macdoiigalii.

Corolla-tube 5 mm. in diameter : sepals orbicular : panicle-branches
stout. 4. F. peninsularis.

Filaments with a scale-like appendage near the base : panicle long and
narrow : shrubs with long slender simple branches.

5. F. splendens.
Corolla campanulate, the tube about equalling the lobes. 6. F. campanulata.

I. Fouquieria fasciculata (R. & S.)

Canhia fasciciilata R. & S. Sy.st. 4: 369. 18 19.

Fouquieria spinosa H.B.K. Nov. Gen. & Sp. 3: 452. 1820.

Bronnia spinosa H.B.K. Nov. Gen. & Sp. 6 : 84. //. ^28.

1823.
Caiitua spinosa Willd.; H.B.K. Nov. Gen. & Sp. 6 : 84. 1823.
Echcveria paniciilata Mocino & Sesse ; DC. Prod. 3: 350.

1828.

A tree 4 m. tall or more, with white fragile wood and round
glabrous spiny branches. Leaves fascicled in the axils of the
spines, obovate-oblong, rounded at the apex or sometimes emar-
ginate, cuneate at the base, glabrous, shining, about 2.5 cm. long
and 9-10 mm. wide : panicle terminal, corymbose, much-branched,
its branches glabrous : capsule about i cm. long : sepals nearly
orbicular, about one fourth as long as the capsule : seeds oblong,
compressed, the margin membranous-winged, the inner surface
concave, the outer convex, brown, glabrous, the wings nearly en-
tire, emarginate at the apex and at the base ; episperm thinly
membranous, adhering to the endosperm which is thin and fleshy;
embryo included, straight, almost as long as the endosperm, com-
pressed ; cotyledons leafy, ovate, obtuse, cordate at the base,
fleshy ; radicle cylindric, somewhat acute, one third as long as the
cotyledons.

The above description is drawn from the original in the work
of Humboldt, Bonpland and Kunth, cited above. I have seen no
specimens of this species, but its corymbiform panicle is unlike
that of any of the others. The type material was in fruit only,
and was secured at Puente de la Madre de Dios, at an altitude of
about 5.280 feet. This place is somewhere in the neighborhood
of Mexico City or Actopan, but it has not been possible for me,
up to the present time, to locate it more definitely.

It is interesting to note here that this is the type of Fouquieria,
a monotypic genus at that time, which was based on Cantua fascic-

Nash : Rf.vision of iiik Family Fouquif.riaceae 453

nlata R. & S. (although II.B.K., for no apparent reason, credit it
to W^illdenow). Subsequently lumqiiicria was again published,
this time being based on an entirely different plant, /'". forniosa,
and at the same time F. spinosa, alluded to above, was made the
type of the new genus Brouiia. The genera are, therefore, syn-
onymous, Fouquicria taking presence on account of the priority of
publication.

2. Fouquieria FORMOSA H.B.K. Nov. Gen. & Sp. 6 : 83, //.
527. 1823.
Echci'eria spicata Mocino & Sesse ; DC. Prod. 3 : 349,

1828.
PJiiletaena horrida Liebm. Vidensk. Selsk. Skr. V. 2 : 283. pi.

1851.

A branching shrub 2-3 m. tall, with a racemose inflorescence
and large red flowers. Leaves on the new growth 3-4 cm. long,
petioled ; petiole about one half as long as the blade ; blade 2—2.5
cm. long, 10-13 mm. wide, elliptic, apiculate, cuneate at the base :
fascicled leaves in the axils of the spines smaller, sessile or nearly
so, elliptic, 1.5-3 cm. long, usually less than i cm. wide, rounded
at the apex, cuneate at the base : spike 1.5 dm. long or less, the
flowers ascending: sepals red, 8— ii mm. long, broadly oval to
orbicular : corolla red, the tube a little curved, cylindric, about 2
cm. long and about 7 mm. in diameter, the lobes spreading or
reflexed, orbicular, abruptly acuminate, 6—8 mm. long : stamens
exserted, unequal in length, sometimes twice as long as the cor-
olla, the filaments a little broadened and compressed below, gla-
brous at the base, then pubescent for a short distance with long
ascending hairs, the remainder of the filament glabrous, the
anthers oblong-ovate, cordate at the base, acute at the apex, 5—6
mm. long : styles united except at the apex, shorter than the
longest stamens, the divisions 3-5 mm. long.

Southern Mexico.

Specimens examined. — Jalisco : Guadalajara, Pringle 2^20,
1889. Puebla : Tehuacan, /*/7;/^A' 62(^6, 1895. Mexico: Chi-
quihuite, Boiirgeaii 1120, 1 865-6.

This is quite distinct from any of the other species in its spicate
inflorescence. The exact locality from which it was originally
secured is not indicated. The specimens cited above would point
to the southern part of Mexico as its home. Philetaeria horrida
Liebm. was obtained in valleys at an altitude of i 500-1800 meters,

454 Nash : Revision of the Family Fouquieriaceae

in the district of Tehuacan, State of Pueblo, from which place
Pringle also secured the same plant many years later. Lieb-
mann's excellent plate and description leave no doubt as to the
identity of his plant with the one of H.B.K.

3. Fouquieria Macdougalii sp. nov.

A much-branched tree, reaching a height of 7 m. and a trunk
diameter of 1-2 dm., with bark yellowish green on the trunk and
brown on the spiny branches, the spines 1-2 cm. long, and termi-
nal slender panicles of few bright red flowers. Leaves on the new
growth 6-8 cm. long, petioled ; blade 3.5-4 cm. long and about i
cm. wide, acute at the apex, and rather abruptly narrowed into a
petiole of the same length : fascicled leaves in the axils of the
spines gradually narrowed into a sessile base, 3.5-4 cm. long and
about I cm. wide : panicle slender, 7-10 cm. long, its deHcate
simple branches widely spreading, the lower ones 3-4 cm. long
and bearing 2-4 flowers on slender pedicels which are 1-2 cm. in
length and abruptly thickened at the apex : flowers few : sepals
broadly oval, about 6 mm. long, the outer two acute, the inner
three rounded and apiculate at the apex : coroUa about 2.5 cm.
long, the tube cylindric, 3-4 mm. in diameter, the lobes broadly
ovate, erect, acute, about 5 mm. long : stamens 10, exserted, the
filaments red above, white below, broader and dorsally compressed
near the base, the interior surface of this compressed portion gla-
brous, the exterior surface pubescent with long stout hairs which
gradually increase in length upward and extend but a short dis-
tance on the rounded part of the filament which is glabrous to the
summit, the anthers 2-3 mm. long : styles exceeding the stamens,
united almost to apex : capsule about 2 cm. long.

Type specimen from living plants, collected at Torres, Mexico,
in 1902, by MacDougal, no. 28, which flowered in the conserva-
tories of the New York Botanical Garden in June, 1903.

Sonora and Sinaloa.

Specimens examined. — Sonora: Rayon, Thurber gj2, 1851;
Torres, MacDougal 28, 1902; Alamos, Palmer j 06, 1890; Gra-
nados, Hartman 226, 1 890. Sinaloa : Culiacan, Palmej- 1804,
1891.

This plant was first secured by Thurber at Rayon, about
eighty miles north of the place where it was recently obtained by
Dr. MacDougal. Thurber's plant was distributed as F. spinosa,
from which it differs materially. It is evidently, however, the

Nash : Revision of tiik Fa.mii.v Fouquiekiaceai-: 455

plant referred to under that name in the preface to Gray's Plantae
Novae Thurberianae (Mem. Am. Acad. Arts & Sci. II. 5: 303).
His description of this tree agrees with a photo<^raph of one made
by Dr. MacDougal. The trunk arises from the ground for two
or three feet, and then divides into crooked branches, the ultiniate
divisions of which are pendulous.

4. Fouquieria peninsularis sp. nov.

Bronnia spviosa Benth. Voy. Sulph. 16. 1844. Not II.H.K.

1823.

A shrub 2-3 m. tall, with a conic panicle and red flowers.
Leaves on the new growth 5-6 cm. long, petioled : petiole about
3 cm. long, about as long as the blade: panicle conic, 1.5 dm.
long or less, its branches ascending, stout, the lower ones some-
times 4-6 cm. long and usually divided, bearing 2-4 flowers on
the ultimate divisions in a rather crowded manner, on short stout
pedicels usually less than 5 mm. long : sepals orbicular or nearly
so, 5-6 mm. long, apiculate, reddish : corolla red, the tube slightly
if at all curved, about 1.5 cm. long and 5 mm. in diameter, the
lobes erect or nearly so, orbicular, acute, 5—6 mm. long : stamens
e.xserted, unequal in length, the filaments broadened and com-
pressed at the base, the inner surface of the compressed portion
glabrous, the outer surface pubescent with long ascending hairs,
the remainder of the filament glabrous, the anthers 3-4 mm. long :
styles united except at the apex : capsule fully 2 cm. long.

Lower California and western Sonora.

Specimens examined. — Lower California : La Paz, Maj. W.

Rich, Dec. 11, 1847 (type); Turtle Bay, Anthony 7./^, July-Oct.,

1896; San Bartolome Bay, Chas. F. Pond, March, 1889; Cal-

malli, Purpus 141a, Jan. -March, 1898 ; Cape San Lucas, Xantus

j8. Sonora: Gxydiymdi?,, Palmer 266, 1890.

Related to F. splendens, but distinguished by the absence of the
appendage at the base of the filaments, the more open panicle and
the larger capsule.

I have ventured to identify the Bronnia spinosa of the Voyage
of the Sulphur with this plant, as I have seen 'material from Cape
San Lucas, the place from which that plant was secured. The
type of this species was secured at La Paz, only about ninety miles
north of Cape San Lucas. Bentham describes the filaments as
glabrous, a condition unknown in the genus so far as I have ex-

456 Nash : Revision of the Family Fouquieriaceae

amined it. Otherwise his description agrees well with this plant,
and as the filaments are pubescent only toward the base the pubes-
cence might readily be overlooked.

5. Fououieria splendens Engelm. in Wisl. Mem. Tour Mex. 98.

1848.

Fouqideria spinosa Torr. in Emory, Mil. Reconn. \4.y,pl.8.
1848. Not H.B.K. 1820.

A branching shrub, the long slender branches arising from near
the base, sometimes to a height of 4-6 m., with a narrow panicu-
late inflorescence and red flowers. Leaves of the new growth 3-5
cm. long, petioled ; petiole about one half as long as the blade ;
blade 2-3 cm. long, 5-7 mm. wide, narrowly elliptic to oblanceo-
late, acute at the apex, narrowed at the base : fascicled leaves
I- 1. 5 cm. long, 5-8 mm. wide, narrowly obovate to obcordate :
panicles narrow, solitary or several at the apex of the stem, 5-20
cm. long, the branches usually short and with the few flowers
crowded : sepals broadly oval to nearly orbicular, 5-8 mm. long,
obtuse or rounded at the apex : corolla red, the tube straight, i 5
18 mm. long, gradually a little enlarged toward the apex, 4-5
mm. in diameter at the middle, the lobes spreading and recurved
or enrolled, broadly oval to orbicular, obtuse at the apex or some-
times apiculate, 4-5 mm. long : stamens exserted, unequal in
length, the filaments broadened and dorsally compressed at the
base, the compressed portion running out laterally into a scale-like
appendage which is pubescent on the upper surface and 0.5-1 mm.
long, the inner surface of this broadened portion glabrous, the
outer surface pubescent with long stout hairs, the remainder of the
filament glabrous ; anthers about 4 mm. long, oblong, abruptly
acute, cordate at the base : styles more or less united, only toward
the base, or nearly to the apex : capsule 1-1.8 cm. long.

Western Texas and northern Mexico to southern California
and northern Lower California.

Specimens examined. — Texas : El Paso, G. R. Vasej, May,
1 88 1, and L. H. Deivey, June 15, 1891 ; Wright 228, October,
1849. Mexico : Chihuahua and Sonora, Tliurbcr 4.01, May, 185 i ;
Coahuila, Prt/7wr 80, 1880; San Pablo, Gregg, A^^rW 30, 1847.
New Mexico : Grant Co., Mcanis 46, 1892; Little Mt., near Las
Cruces, Wooton, May and July, 1893. Arizona : Tucson, Tourney,
April 20, 1894 ; Tucson Mts., Tourney 465, 1892 ; Willow Springs,
Jones, May 29, 1890; Gila Valley, Rothrockjig; foothills, Prin-

Nash : Revision ok the Family FouguiEKiACEAE 457

^/r, April 6, 1891 ; San Francisco Mts., W/uclir, 1872; Squaw
Peak, Mcartis ijj, May 6, 1887; Ft. Huachuca, Wilcox, June i,
1 892, and 106, 1 894. California : liniory, November 29, i 846 ; Fre-
mont, 1849 ; The Needles, /6';/r.y?c'?j/, May 6, 1884 ; southern Cali-
fornia, Parish. Lower California : Rosario,(9r6V//'/ 1354, May i , 1 886.
This species has a much more extended range than any of the
others and is the only one found within the confines of the United
States. It may be distinguished at once from narrow-panicled
forms of F. pciiiiisularis by the prominent appendage near the
base of the filaments and by the larger capsule. It has a num-
ber of common names, among them being Ocotillo, Coach-whip,
Vine-cactus, and Jacob's Staff.

6. Fouquieria campanulata sp. nov.

A woody plant with narrow panicle and red campanulatc flow-
ers. Leaves on the new growth 3-4 cm. long, petioled ; petiole
less than one half as long as the blade, which is 2-3 cm. long and
4—5 mm. wide, narrowly oblong or oblanceolate : fascicled leaves
2-3 cm. long, 3-6 mm. wide, oblanceolate, acutish, narrowed to
the sessile base: inflorescence a narrow panicle 1. 5 dm. long or
less, the branches short and spreading and the flowers on them
crowded : sepals broadly oval to orbicular, 5-6 cm. long, pale :
corolla, from the tip of the recurved spreading lobes to the base
of the tube, 12-14 """^i- ^ong, campanulate, the tube, which is
about as long as the lobes, 3 mm. broad at the base, enlarging to
the summit where it is 5-6 mm. in diameter : stamens exserted,
unequal in length, the filaments broadened and compressed at the
base, the broadened portion running out into a spreading or re-
flexed scale-like appendage which is pubescent on the upper sur-
face and about 0.75 mm. long, the inner surface of the compressed
portion glabrous, the outer surface pubescent with long hairs
toward the summit, the remainder of the filament glabrous, anthers
ovate-elliptic, cordate at the base, acute at the apex.

Durango.

Spixiviois examined. — Santiago Papasquiaro, Palmer 8 j , 1896.

An exception in the genus in having campanulate flowers.
This and F. splendcns are the only species which have well-de-
veloped appendages toward the base of the filaments.

2. IDRIA Kellogg, Hesperian, 4: loi,//. i860

Trees with a stout columnar trunk from which arise short
spreading spiny branches. Inflorescence paniculate, arising from

458 Nash : Revision of the Family Fouquieriaceae

the apex of the trunk, the flowers rather crowded and almost ses-
sile upon the ultimate divisions of the panicle. Corolla yellow,
campanulate, the lobes orbicular, concave, incurved, shorter than
the tube. Stamens lo, adnate at the very base to the corolla-
tube : filaments pubescent below. Styles short, thick, united
their entire length, but little longer than the ovary, forming a stout
3 -angled body, the angles rounded.

Species one ; Lower California.

I. Idria columnaris Kellogg, Hesperian, 4 : loi. />/. i860.

Fouqiiicra cohinmaris Kellogg ; Curran, Bull. Cal. Acad. Sci.

i: \ZZ.pl. 1885.
Fouqiiiera gigmitea Orcutt, West Am. Sci. 2 : 48. 1886.

A tall tree with a tapering trunk up to 3-4 m. in height or
even taller, and a diameter at the base of about 3 dm. or more,
from which arise the short spreading spiny and leafy branches.
Leavesof the new growth unknown : fascicled leaves 1.5— 2 cm. long
and 5—8 mm. wide, oblanceolate to narrowly obovate : panicles
3-4 dm. long, flowers rather crowded and nearly sessile upon the
ultimate divisions : flowers, including the exserted stamens, 12-14
mm. long : sepals orbicular, about 4 mm. long : corolla yellow,
6-7 mm. long : stamens exserted, the filaments pubescent below,
the anthers about 3 mm. long : styles about 2 mm. long, thick.

Lower California. ,

Specimens examined. — Rosalia Bay, Anthony 120, July to
October, 1896.

Originally collected by Dr. Veatch near the Bay of Sebastian
Viscaino, on the mainland east of Cedros Island. The tree was
described by Dr. Kellogg as spineless, but this must have been an
error, as others who have visited the same region remark upon the
long spines which are found on the short branches arising from
the trunk. Moreover, a specimen in the herbarium of the New
York Botanical Garden, collected by Anthony at Rosalia Bay,
but a few miles north of the original station, shows these spine-
bearing branches, the spines being similar to those occurring in
the other members of this family. Dr. Kellogg describes the
trunk as undivided, while Orcutt, in the description of his
Fonquicra gigantea, states that the trunk branches above the
middle, sending up a few simple branches to nearly the height of
the main stem. Brandegee, in his account of a collection of plants

Nash : Revision of thk Family Fuu(juieriaci-:ak 4r>{)

made in Lower California in 1889 (Proc. Cal. Acad. Sci. II.

2 : 132, 133), also refers to this branched condition, but considers

this state as exceptional and due to accident or injury. Both

Brandegee and Orcutt claim the trunk attains a height of fifty feet

or more. Brandegee remarks that the old capsules are 8-10 nun.

long and sessile in the panicle, and that the trunks are 6-9 dm. in

diameter at the base, gradually tapering upward into a pointed

apex, so that the general shape is much like that of an inverted

carrot.

New York Botanical Garden.

X*XJBLie\A.TIO]VS

OF

The New York Botanical Garden

Journal of the New York Botanical Garden, monthly, illustrated, con-
laining notes, news and non-technical articles of general interest. Free to all mem-
bers of the Garden. Toothers, lo cents a copy; $i.oo a year. [Not offered in
exchange.] Vol. I, 1900, viii + 213 pp. Vol. 11, 1901, viii + 204 pp. Vol. Ill,
1902, viii -(- 244 pp.

Bulletin of the New York Botanical Garden, containing the annual reports
of the Director-in-Chief and other official documents, and technical articles embodying
the results of investigations carried out in the Garden. Free to all members of the
Garden ; to others, ^3.00 per volume. Vol. I, Nos. I-5, 449 pp., 3 maps, and 12
plates, 1896-1900. Vol. II, Nos. 6-8, 518 pp., 30 plates, 1901-1903.

Memoirs of the New York Botanical Garden. Price to members of the
Garden, ;^i.oo per volume. To others, ^2.00. [Not offered in exchange.]

Vol. I. An Annotated Catalogue of the Flora of Montana and the Yellowstone
Park, by Dr. Per Axel Rydberg, assistant curator of the museums. An arrangement
and critical discussion of the Pteridophytes and Phanerogams of the region with
notes from the author's field book and including descriptions of 163 new species,
ix -|- 492 pp. Roy. 8vo, with detailed map.

Vol. II. The Influence of Light and Darkness upon Growth and Development,
by Dr. D. T. MacDougal, first assistant and director of the laboratories. An ac-
count of the author's extensive researches together with a general consideration of
the relation of light to plants. The principal morphological features are illustrated,
xvi -f 320 pp. Roy. 8vo, with 176 figures.

Contributions from the New York Botanical Garden. A series of tech-
nical papers written by students or members of the staff, and reprinted from journals
other than above. Price, 25 cents each.

Vol. I. Inclusive of Nos. 1-25, vi -\- 400 pp. 35 figures in the text and 34
plates. ^5.00.

CURRENT NUMBERS 25 CENTS EACH.

No, 33. Studies in the Asclepiadaceae, by Miss A, M. Vail.

No. 34. The genus /HeHa with descriptions of new species from North America
and the Canary Islands, by Dr. M. A. Howe and Prof. L. M. Underwood.

No. 35. The Polyporacese of North America. — III. The genus Fames, by Dr.
W. A. Murrill.

No. 36. Studies on the Rocky Mountain Flora. — X, by Dr. P. A. Rydberg.

No. 37. Some Generic Segregations, by Dr. Per Axel Rydberg.

No. 38. The Polyporacese of North America — IV. The genus Elfvingia, by
Dr. W. A. Murrill.

All subscriptions and remittances should be sent to

NEW York Botanical Garden

Bronx Park, New York City

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 43

SOME CORRELATIONS OF LEAVES

By DANIEL TREMBLY MACDOUGAL

NEW YORK
1903

[Reprinted from the Bulletin of the Torrey Botanical Club, 30 : 503-512,

September, 1903]

Some Correlations of Leaves

liv l)ANii:i. TKKMiii.v Mac1Jou(;al

As a result of a long series o( experimental tests made in the
New York Botanical Garden, Mr. Charles Zeleny has found that
the excision of one of the leaOets of such plants as Trifoliimi pra-
tcusc (clover), Parthcnocissus quinqucfolia (Virginia creeper) and
Lupimis albiis (white lupine) resulted in alterations in the positions
of the remaining leaflets, alterations in tlic intervals between the
remaining members and divergences from the normal size.

Our knowledge of correlations at the present time would lead
to the expectation that organs, the activity of which was comple-
mentary or dependent upon a removed organ would show a de-
creased or diminished development. On the other hand tlic small
amount of evidence available would have led to the generalization
that the loss of an organ or a part of an organ would stimulate the
development of the remainder of the organ, or of the tissues carry-
ing on the same function, in a supposed effort to bring the total
functional performance up to the normal average. Mr. Zeleny's
results, however, demonstrate unequivocally that the excision of a
leaflet in the above species is followed by a diminished development
on the part of the remainder, which results in the accomplish-
ment of a superficial extension, as represented b\' the length of the
members, about .seven per cent, less than the normal* Results
fairly in agreement with those of Mr. Zeleny have also been ob-
tained by Nemec by less exact methods, f

Goebel cites the fact that the stipels of the compound leaves
of Robuna Psaidacacia (locust) reach an abnormall}' large size
when the leaflets are removed, which is an example of the induced
enlargement of a structure in the effort to carry out the functions
of lost tissue. I He has also found that the destruction of
the vegetative points on such leaves as those of Bryop]iyllum is

* Read before the Botanical Society of America, Pittsburg, July l, 1902. Now in
press in Bull. N. Y. Bot. Garden, vol. 3, no. 9.

t Nemec, B. Ueber die Folgen einer Symmetriestorung bei zusammengesetzten
Blattern. Bull. Internal. Acad. Sci. Boheme. 1902.

j Goebel. Organography of Plants, i : 210. 1900.

503

504

MacDougal : Some Correlations of Leaves

followed by the formation of new ones at places nearer the base
of the leaf-blades. * It is evident therefore that the epipodium, or
laminar portion of the leaf, is capable of the most diverse correla-
tive reactions according to the structure, and adaptive, or secondary-
functions it has acquired.

The lack of uniformity in the facts at hand suggested that a
repetition of some of the experimental tests with additional species
might yield results of value. To this end the author made a
series of observations upon the effect of the removal of the epipo-
dium and mesopodium of the leaves of one species upon the stip-

A B C

Fig. I. Primus serotina. A, normal leaf, X >2. B, normal stipules, X 3- C,
leal-base with enlarged stipules; the stump of the excised petiole is shown, X 3-

ules of the hypopodium and upon the vegetative points of the
stems from which the leaves arose. Additional observations were
carried out on the influence of the removal of a leaf showing
only a distinct mesopodium (petiole) and epipodium (lamina) upon
the vegetative points of the stem and upon the development and
differentiation of the tissues of the stem.

A small specimen of Pnnuis serotina growing near the prop-
agating houses of the New York Botanical Garden was selected
for the first test. All of the young unfolding leaves, except those
of one basal branch, were dissected by the excision of the petiole
near the stipules on April 24, 1903. None of the leaves had un-
folded on the above date. The plant was visited daily and the
successive leaves removed in the same manner leaving the stipules
intact. A number of the other small trees of the same species

*Goebel. Regeneration in Plants. Bull. Torrey Club, 30 : 197-205. 1903.

MacDougal: Some Cokkklations of Leaves 505

growing near by were also available for comparison. The earlier
part of the season was one of extreme drought. During the
period of fifty-two days between April i6, a week before the be-
ginning of the test, a total precipitation of (mly i cm. was recorded.
It was noted that the scales and stipules were of longer duration
and presented the appearance of increased superficial extension
upon the delaminated branches within a fortnight after beginning
the tests. The development of leaves and buds appeared to have
almost ceased and an anatomical examination was made June 3-10
with the following results :

The greatest length made by branches which had been defoliated
from the beginning was 18 cm., while a length of 25 and 30 cm.
was found in normal branches; 18-20 leaves had been formed
on defoliated branches and but 14 or 15 on normal ones. The
basal internodes of the twigs showed but little difference in length
in the two instances, being as much as 2 cm., although less in some
cases. The internodes formed later in the season on the normal
twigs attained a length of as much as 3.5 cm. in some instances,
while those of the defoliated twigs varied from 1.5 cm. to one
third that length. Consequently the defoliated twigs were closely
crowded at the terminal part with scales and stipules. This
crowding effect was accentuated by the fact that one or two small
leaves were thrust out of the axillary buds, although no general
activity of the bud was exhibited. In contrast with the behavior
o{ Acer, it could not be made out that any extra number of lateral
buds on the elder parts of the branches below had been induced
to awaken.

The correlative alterations in the stipules were very marked
and consisted in changes in form, size, position, structure and
period of endurance. Normal stipules of the cherry are thin-
membranaceous, and generally of a reddish tinge containing almost
no chlorophyl. These stipules are usually short-lived and fall off
in such manner that on the above date they could be found onh-
on two or three leaves on the terminal portions of twigs. The
stipules on twigs on which the leaves had been dissected were
both broader and longer, offering a superficial expansion of about
five times the normal (see /. /). The maximum measurements
were but little greater than those of normal stipules formed earlier

506

MacDougal : Some Correlations of Leaves

on the first internode, but their length, width and thickness was
such as to give the exaggerated extension noted above when com-
pared to correspondent structures on normal branches. Further-
more, three pairs of stipules were present on normal twigs and ten

Fig. 2. A, Acer ; normal and defoliated branches. The dotted line shows the
bases of new growth. H, Frunus ; transverse section of normal stipule. (Z, Primus ;
transverse section of stipule of delaminated leaf. B and C are drawn to same scale.

pairs on those which had been defoliated. It is thus to be seen
that twigs which had been deprived of the laminar portions of the
leaves made such development and retention of the stipules that
they were able to expose about fifteen or sixteen times as much
stipular surface as normal correspondent branches.

The increase of the stipular surface consequent upon the re-
moval of the laminae is not unknown, and Goebel notes that simi-

MacDougal : Somk Cokkri.ations of Leaves 507

lar reactions are shown by the elder [Savibucits nigra)* He also
found that the stipules of Vicia Faba were increased in superficial
extension two to six times beyond the normal by the excision of
the laminae in an early stage of development, and a similar exag-
geration in the stipules of Lathynis Aphaca resulted from the ces-
sation of elongation or extension of the shoots. This correlation
is not present in all plants, however, and not even in all Papilion-
aceae, since it was seen that the excision of the laminae of Phascolus
multiJJorus had but little effect upon the size and structure of the

laminae.f

The excision of the leaves of Aristolochia, in the course of the
present experiments, did not cause any notable increase in size of
the stipules as seen by general inspection, although no examina-
tion of the internal structure was made.

Kronfeld made a series of experimental tests of this matter in
1886, and found that the stipules of Pyrctlirnui Iudic7tin, Rosa
sevipcrjionns, Rubus fruticosus, R. Idaais, Sida Napaca, Trifoliiim
filifovme and Urtica urcns were not sensibly increased in size by
the destruction of the laminae. The destruction of the laminae of
Pyrns Mains was followed by an increase of the superficial expan-
sion of the stipules amounting to a hundred per cent. The exci-
sion of the laminae of Pisnm sativum was followed by an increase
of the size of the stipules amounting to from fifty to one hundred
per cent., while the excision of the stipules alone seemed to have
the effect of causing earlier flower-formation. %

Stomata were found on both normal and abnormal stipules in
Pnmns in my own experiments. No examination of the number
of stomata on the adjoining surfaces of the stem were made, but
Braun has found that defoliation causes the formation of an in-
creased number of these organs. § The enlargement of the stimu-
lated stipules of Prnnns had apparently not been accompanied by
a multiplication of the epidermal cells, since these elements were
both larger and wider in surface view than the normal.

*Goebel. Organography of Plants, i : 210. 1900.

f Goebel. ( Beitrage zur Morphologic und Physiologic dcs Blattes. ) Bot. Zeitung,
38 : 836-837. 1880.

% Kronfeld, M. Ueber die '• Correlation dcs Wachstums." Bot. Zeitung, 44 :
846-849. 1886.

\ Braun, K. Ucber Verandcrungen ira Gewcbe entlaubter Stengel und Zweige.
(Inaug. Diss.) Erlangen. 1899.

508 MacDougal : Some Correlations of Leaves

The enlarged stipules exhibited a thickness four or five times
as great as the normal, chiefly due to increase in tissues and dif-
ferentiations toward the structure of a typical lamina. A median
layer of loosely arranged parenchymatous tissue contained much
chlorophyl which was almost wholly lacking from the normal
organ. This mesophyllary tissue also exhibited numerous inter-
cellular spaces, and was altogether well adapted to carrying on
photosynthetic and transpiratory functions.

The entire stipule, which is usually closely appressed to the
petiole, was held at a more widely divergent angle than the normal.
These adaptations have greater significance when it is understood
that the total amount of stipular surface presented by a delami-
nated branch would amount to about fifteen or sixteen times as
much as the normal. This comparatively normal increase, how-
ever, does not result in developing a foliar surface of more than
one or two per cent, of that of the normal leafy branch.

Sections of the sixth internode from the base of the twigs
formed in 1903 were examined in order to ascertain the effect of
deprivation of the laminar structures and their partial replacement
by the stipules. It has already been pointed out that the branches
which had been delaminated had developed more internodes, the
total lengths of which were less than of twigs normally grown, and
had the appearance of being slightly thicker, but no measurements
were made which might form the basis of an exact comparison.
The epidermal tissues of the treated branch had greater radial
and tangential diameters, and the underlying collenchymatous
layers were but slightly thickened. The medio-cortex contained
much more chlorophyl than the normal. The bast fibers were
not so heavily thickened as in the normal, and the walls of all the
internal tissues exhibited a slight yellowish tinge. The cambium
showed a greater number of layers than the normal. The xylem
was irregularly developed and seemed to have attained a less
•advanced stage of differentiation in its various elements than in
the normal. The tangential development was notably deficient
and the separate bundles were separated by wide rays of pith.

On June 1 5 the delaminated branches had assumed the appear-
ance of a resting stage. The formation of leaves had ceased, and
nearly all of the stipules had fallen off The buds in the axils of

MacDougal : Some Correlations of Leaves 509

the newly formed branches were enclosed and in the winter con-
dition. My absence from the Garden during July prevented obser-
vations being taken during a period of four weeks, but an exami-
nation of the plants on July 30 showed that all of the newly formed
delaminated branches of this season had perished and that buds on
the older branches had awakened and extended to a length some-
thing less than the normal growth of the earlier part of the season.

The increase in the amount of chlorophyl in the cortex may be
considered as a further effort to meet the losses ensuing from the
lack of the laminar portions of leaves. The non-development of
the phellogen and the imperfect construction of the bast fibers and
fibrovascular bundles may be ascribed in part to a lack of formative
material and in part to the absence of the stimulation ensuing from
the decreased transpiration stream and the material usually formed
in the laminae. In addition, the mechanical strains and stresses
e.xerted by the foliar organs undoubtedly exert some stimulative
effect on the rigidity and pliability of the stems by the induction of
the development of various mechanical elements. The absence of
stimuli of this character would account in part for the imperfect
development of the stems.

The tests with Ac\r were much more restricted. The petioles
were excised near the base and the effect upon the activity of the
vegetative points noted. Only a single basal branch of a small
tree was treated in this manner, and comparisons were made with
corresponding portions of similar branches arising from the same
part of the central stem.

The single normal branch taken for comparison developed
four internodes which had a diameter about thirty per cent, greater
than the internodes of the defoliated branch. The total length of
the four newly formed internodes amounted to 78 mm., the longest
being the third from the base of the yearly growth, measuring 25
mm. Eight active functional leaves were developed. (One has
been cutaway in the illustration,/. 2, A.) The newly-grown part
of the defoliated branch measured 48 mm. and comprised seven
internodes, the longest of which was the sixth from the base,
measuring 8 mm. The excision of the leaves had induced the de-
velopment of one bud in the axil of the last internode of the
previous season. The branchlet thus formed showed two long

510 MacDougal : Some Correlations of Leaves

internodes, the basal one of which measured 14 mm. Two
branchlets were produced at the next node below on the newly-
formed portion of the main branch, which showed two and three
internodes and were 10 and 15 mm. long respectively. The
third internode below produced two branchlets 6 and 12 mm.
long and composed of two internodes.

The lateral branchlets of the main branch showed less total
growth and also less marked development of the lateral bnds cor-
respondent to the general habit of the normal. Activity of the
vegetative points on the defoliated and normal branches had
practically ceased on June 15, the mutilated branch showing no
prolongation of the period. The departures from the normal
structure of the stem were fairly similar to those described in
Pnimis. Chlorophyl was quite as profusely developed in the
normal as in the treated branch, however, and was found in the
medullary rays in great abundance.

The above results are found to agree in the main with those
of Boirivant, who practiced defoliation and delamination and noted
the effect upon stems and upon petioles of Faba vulgaris, Sa-
rothainmis scoparins, Genista tiiictoria, Lathynis odoratus, CJicno-
podiinn album, Atriplex nitens, Linum usitatissimum, Galium
Cruciata, Asparagus officinalis, Robinia Pseudacacia, Araclds
hypogaca and Ailantus glandidosa^' It is to be noted, however,
that the epidermal tissues in my material showed enlargement in
the tangential axis, and one or two other minor differences appear.
Boirivant's studies included a consideration of the anatomy of the
petioles of leaves of which the laminae had been cut away, and in
these organs he found an exaggerated production of chlorophyl,
modifications of the cortical tissues consisting of added differentia-
tions of palisade tissue, and enlargement of the vessels and more
lignification of the tracheids and pith. I am not able to find that
he noted the effects of the treatment upon the stipels of Robinia,
or made observations upon the reaction of the hypopodium
especially.

The observations of Braun also extend over the matter of the
effect of delamination on stems. His observations included

* Boirivant, A. Recherches sur les organes de remplacement chez les plantes.
Ann. Sci. Nat. Bot. VIII. 6 : 309-400. 1897.

MacDougal : Some Corkklations of Leaves 511

Laniiuin Orcala, Aconitinn Stocrkianutn, Clematis I Italba, Syringa
I'lilgaris, Corylus Avcllaiia titropnrpiircn, Gillcnia trifoliata, Pninus
Mahalib, Acsculus Ili/f/>ocastanmn, Ampclopsis hcdiracea, Aristo-
locliia Sip/io, Anr plataiioidcs, Rosa centifolia and Phascolns uiulti-
Jloriis* In addition to results previously cited, Braun concludes
also that the effect of delamination is to increase the photosyn-
thetic capacity of stems, retard the development of internodes and
interrupt the formation of woody tissue, all of which is confirmed
by my own results. Braun also suggests that the condition of
delaminated stems may be regarded as pathological, and that it is
directly due to mechanical interferences with the functions
involved. The condition of delaminated branches is undoubtedly
one of hypoplasia, but a careful consideration of the various fea-
tures involved leaves but little doubt that the failure to develop
is largely due to the absence of the customary stimuli in response
to which the several tissues carry on growth and differentiation in
a normal manner. The results in question are quite as much a
matter of correlation dependent upon stimulative factors, as the
variations in position and growth in compound leaves as discovered
by Zeleny.

The more important facts observed as a result of cutting away
the laminae of the leaves of Acer and Pninits by severing the
petiole near the base may be briefly summarized as follows :

I. The duration of the stipules of Pruniis is increased, so that
ten pairs of these organs are to be found on branches at a time
when only two or three pairs are present on a normal branch.

II. The stipular surfaces exposed on delaminated branches of
Pntniis were sixteen times as great as on normal branches.

III. The individual stipules of Pnimis were of a maximum
measurement in length and width on the delaminated branches and
were much larger than the stipules present on the corresponding
part of normal branches.

IV. The position of the stipules of Prunus on delaminated
leaves was more divergent than the normal, and appro.ximated
that of a typical leaf.

V. The structure of the stipules of Pruiins was modified in

* Braun, .\. Ueber VerSnderungen im Gewebe entlaubter Stengel und Zweige.
(Inaug. Diss.) Erlangen. 1 899.

512 MacDougal : Some Correlations of Leaves

such manner as vastly to increase their capacity for photosynthe-
sis and transpiration.

VI. The loss of the lamina acts as a stimulus which induces a
development and differentiation of spongy parenchymatous and
other tissues in the stipule, and the formation of chlorophyl.

VII. The lateral buds of Acer on internodes formed during the
previous season were awakened by defoliation of the extending
apical portion.

VIII. The period of activity of the vegetative point of the
terminal portion of a woody branch is not lengthened as a result
of defoliation.

IX. The total length of the newly formed portion of the
branches defoliated during growth was thirty to sixty per cent,
less than that of normal branches occupying the same relative
positions on the shoot.

X. The number of internodes developed on defoliated branches
was greater than in the normal.

XI. The defoliated internodes were of a diameter not greater,
sometimes less than the normal, and had accomplished less perfect
differentiation.

XII. Branches compelled to accomplish growth in a defoliated
condition generally are found to be in a state of hypoplasia, the
undeveloped and imperfectly differentiated condition of the tissues
being due to the lack of correlative stimuli.

New York Botanical Garden.

r»UBLio^%^Tioi>rs

OF

The New York Botanical Garden

Journal of the New York Botanical Garden, monthly, illustrated, con-
taining notes, news and non-technical articles of general interest. Free to all mem-
bers of the Garden. To others, lo cents a copy ; ^i.oo a year. [Not offered in
exchange.] Vol. I, 1900, viii -f- 213 pp. Vol. II, 1901, viii + 204 pp. Vol. Ill,
1902, viii -|- 244 pp.

Bulletin of the New York Botanical Garden, containing the annual reports
of the Director-in-Chief and other official documents, and technical articles embodying
the results of investigations carried out in the Garden. Free to all members of the
Garden; toothers, ^3.00 per volume. Vol. I, Nos. 1-5, 449 pp., 3 maps, and 12
plates, 1896-1900. Vol. II, Nos. 6-8, 518 pp., 30 plates, 1901-1903.

Memoirs of the New York Botanical Garden. Price to members of the
Garden, ^l. 00 per volume. To others, $2.00. [Not offered in exch.-inge.]

Vol. I. An Annotated Catalogue of the Flora of Montana and the Yellowstone
Park, by Dr. Per Axel Rydberg, assistant curator of the museums. An arrangement
and critical discussion of the Pteridophytes and Phanerogams of the region with
notes from the author's field book and including descriptions of 163 new species.
ix + 492 pp. Roy. 8vo, with detailed map.

Vol. II. The Influence of Light and Darkness upon Growth and Development,
by Dr. D. T. MacDougal, first assistant and director of the laboratories. An ac-
count of the author's extensive researches together with a general consideration of
the relation of light to plants. The principal morphological features are illustrated,
xvi -f 320 pp. Roy. 8vo, with 176 figures.

Contributions from the New York Botanical Garden. A series of tech-
nical papers written by students or members of the staff, and reprinted from journals
other than above. Price, 25 cents each.

Vol. I. Inclusive of Nos. 1-25, vi -|- 400 pp. 35 figures in the text and 34
plates. Is. 00.

CURRENT NUMBERS 25 CENTS EACH.

No. 38. The Polyporaceae of North America — IV. The genus Elfuingia, by
Dr. W. A. Murrill.

No 39. A preliminary enumeration of the grasses of Porto Rico, by George V.
Nash.

No. 40. The phyllodes of Oxypolis filiformis, a swamp xeropbyte, by Miss Rosina
J. Rennert.

No. 41. The Polyporaceae of North America. — V. The genera Cryptoporus,
Piptoporiis, Scutiger and Porodiscus, by Dr. W. A. Murrill.

No. 42. A revision of the family Fouquieraceae, by George V. Nash.

No. 43. Some correlations of leaves, by Dr. Daniel Trembly MacDougal.

All subscriptions and remittances should be sent to

New York botanical Garden

Bronx Park, New York City

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 44.

SOIL TEMPERATURES AND VEGETATION

By DANIEL TREMBLY MACDOUGAL

1903

[Reprinted from the Monthly Weather Review for August, 1903.]

SOIL TEMPERATURES AND VEGETATION.

V\ I>ANii;i. Tkkmui.y MAcI)(if<iAi,.

A committee ou the relations of plants to climate was ajj-
pointed at the New York meeting of the American Association
for the Advancement of Science in June, I'JOl, to which was
delegated the task of carrying out some work uj^on the rela-
tions of plants to various climatic factors. The actual inves-
tigations planned by the committee wei'e entrusted to the
autlior for execution, and a set of thermographs -was jiut in
action in northern Idaho in the summer of the same year in
addition to the battery that was installed in the New York
Botanical Garden. During the following year some tliermo-
metric observations were made in the Mission Mountains and
Kootenai Mountains,, in northern Montana, and a j^aper was
presented at the meeting of the Association in Denver, August,
1901, describing a method of estimating the total temj^era-
ture exposure of a plant which Avould be specially ajJiilicable
and useful in measuring the iuliuence of temj^eratui-e uj^on the
shoots of plants.

The basal portions of a typical plant, often the larger part
of the body, are imbedded in the soil at various dejiths, and
no adequate study of the influence of temperature upon phys-
iological processes could be made until some accurate, grajjhic,
and convenient method was devised for taking continuous
records of the soil. The committee was given a second grant
by the Association, and additional funds were also voted at
the Pittsburg meeting in July, 1902. By the aid of additional
contributions from the New York Botanical Garden efforts
were made to devise an instrument that would meet the above
needs. The committee was so fortunate as to enlist the active
interest and practical cooperation of Prof. "William Hallock, of
Columbia University, who undertook to design a thermograph
that would make a continuous record of the soil at any desired
depth. A single working model was constructed in the autumn
of 1901 and was tested for several months in Professor Hal-
lock's laboratory before being installed in the Botanical Gar-
den, May 2, 1902. A description of this instrument, together
with the records obtained for May, 1902, have already been
published.^ It will be profitable to repeat this description
hei'e, together with the accompanying illustration (see fig. 1).

The thermal element of the instrument consists of a copper
bulb or globe 11 centimeters in diameter (fig. 1, A), -^dth a

^ MacDougal. The temperature of the soil. Journal of the New York
Botanical Garden. 3: 125-131. July, 1902.

streugtlieniug equatorial ridge of solid metal, tilled with com-
mercial kerosene. A short section of copjier tubing with the
walls flattened on two sides for convenience of manipulation
during construction is soldered to a suitable opening in one
pole of the coiJ2:)er globe, and the free end of the heavy tube is
likewise soldered to a small copper tube with an external diam-
eter of about 4 millimeters and an internal diameter of 1 milli-
meter. This tube is also tilled with j^etroleum and may be of
any reasonable length up to 10 meters or i:)erhaps more without
vitiating the accuracy of the instrument to any apjireciable
extent. Furthermore, in practise this tube may be variously
bent and curved in making adjustments without detriment to
the results obtained (fig. 1, C).

Fig. 1. — Hallock thei'mograph. A, bulb containing kerosene. B,
heavy tube, soldered to polar opening in bulb. C, capillary tube con-
necting bulb with expanding chambei-. D, double expansion chamber.
E, upright post with adjustment clamp at E, on expansion chamber
and fastened to short lever arm at F. G, lever extending from axis,
and bearing pen, which traces temperature record on the cylinder. H,
adjustment to raise or lower pen in making corrections and calibrations.
K, cylinder driven by internal clockwork which makes one revolution
per week. The cylinder carries paper ruled horizontally to degrees
centigrade. The rate of movement carries the paper past the pen at
such rate that two hours is taken to pass over the interval between two
curved lines. L, upright rod on hinged base to remove pen from con-
tact with paper. M, base of cast iron.

The free end of the capillary tube is connected through an
opening in the side wdth the chamber in the interior of a solid
brass block fastened to the base of the apparatus. Two pairs
of convex, corrugated brass disks 8 centimeters in diameter
are fastened together at the edges and the united pairs are then
soldered together forming an expansion chamber comj^osed of
two disk-shaj^ed cells, with an opening through the soldered
portion (tig. 1, D). Next an opening is made in the center of
one of the convex surfaces and this opening is joined by solder
to the chamber in the brass block. The double expansion
chamber is also tilled with petroleum and care is taken to ex-
clude all air from the tubes and chambers. The entire system
is now seen to consist of a I'igi^l bulb, tubes, and small base

cbamber tenuiiiiitin^ in an t-xpiinsioii structure whieli may
dilate or contract with the alterations in volume of the Hiiid.
It now remains to attach devices for 2)roi)erly demonstrating
the raovementH of the upjjer wall of the expansion chamber in
order to obtain the chanj^'es induced by variations in tempera-
ture. To do this an ii])ri^'lit ))ostis fastened to the up])er wall
of the exi)ansi(m chamber and this is provide<l with a sleeve
and set screw which allows adjustments of its length in con-
nection with a lever (tig. 1, E and F) 4 centimeters long run-
ning to an axis held in bearings between two upright stand-
ards. A second arm 17 centimeters in length, bearing a suitable
pen is attached to this axis and extends so as to come in contact
with a suitable uju'ight recording cylinder revolved by clock-
work. The long lever is built up of three parts and is provided
with a device for adjustment and calibration (tig. 1 H). The
cylinder carries double ruled paper and makes a single revo-
lution in a week. In the construction of the working model
described here, it was found convenient to use the levers and
recoi'ding mechanism of a thermograjihof German manufacture,
the slips of which were ruled for degrees centigrade. The
instrument was calibrated to meet these requirements, but new
models now under construction are designed on the Fahren-
heit scale. The improved model may also be constructed to
carry thermal apparatus for taking air temperatures on one
end of the base and the soil temperatures on the other end,
thus making synchronous records.

The peculiar value of the above apparatus consisted in the
fact that the recorder could be kept iu an instrument shelter
of any convenient pattern and the sensitive bulb carried out
and buried in the soil at any desired distance from the shelter,
or at any depth in the soil, by which the actual temperatures
encountered by the roots of jDlants might be obtained. In
practise it was found that the amount of liquid in the capillary
tube and ex2)ansion chamber was so small in proi^ortion to that
in the bulb that the temperatures encountered by the tube or
chamber made no observable error in the reading of the instru-
ment. This was true to such an extent that the Hame of an
alcohol lamp applied to the tubes did not produce any change
in the position of the pen upon the recording slip. The same
is true of the etTects of direct exposure to the sun's rays.

The location for the installation of the instrument for the
trial series of records for the first year was a small area of
clayey soil mixed with loam lying near the propagating houses
and nurseries in the New York Botariical Garden.- The drain-
age is free, as the surface of the ground is slightly sloi)ing
to the eastward. During the summer the place is covered by
the shade of some cherry trees after 4 p. m.

In the installation of the instrument the recorder, together
with an air thermograph, was placed in an instrument shelter
of the United States "Weather Bureau pattern. The capillary
tube passed through the floor of the shelter to the surface

of the ground, being protected from mechanical injury by
being placed in a grooved board. A narrow trench 2 meters
long and about 20 centimeters in depth was dug and was then
prolonged in a tunnel to a farther distance of 50 centimeters.
This tunnel was only large enough to receive the bulb, and
after that had been pushed to the farther end of the tunnel the
excavated soil Avas packed around it in a manner as nearly
natural as possible. By this arrangement the layer of soil
above the bulb was left undisturbed and artificial drainage
conditions were not introduced. This arrangement brought
the upper surface of the bulb in contact with the soil pene-
trated by the matted roots of grasses and other small herba-
ceous plants, which also extended much deeper around the
excavation.

The greater bulk or volume of the root system of the smaller
plants usually lies nearer the surface of the soil than the posi-
tion of the bulb of the thermograph, but the important func-
tion of absorption is carried on chiefly by the younger, most
recently formed terminal portions of the roots which lie deeper
in the soil. The rate at which mineral and other substances
in solution in the soil are taken up, the permeability of the
protoplasmic membranes, the turgidity and consequently the
rigidity of the roots are also greatly affected by temperatures,
while the growth of these organs, together with their endur-
ance and that of all underground organs, depends not only
upon the actual temperatures encountered, biit also upon the
rapidity with which changes ensue.

Before being placed in position the air thermograph and
the soil thermograph were placed in a closed room with a
standardized mercurial thermometer (No. 2792, G. S. Keichs-
Anstalt, Berlin), and were found to read alike to within a
small fraction of a degree, allowing for a slowness of response
by the thermographs.

The records on hand now embrace a period of fourteen
months, and the facts obtained seem to be of sufficient inter-
est to warrant their discussion at the present time, together
with some of the actual curves traced by the instruments.

It may be seen from these that the maximum daily tempera-
tures occurred between 8 and 11 p. m., and the minima twelve
hours later, or between 8 and 10 a. m. The optimum tem-
perature for absorption by roots lies well above that of the
soil at the depth at which the observations were made. It
follows, therefore, that the temperature of the soil approaches
this optimum most nearly, and offers most favorable condi-
tions for the taking up of watery solutions at a time of the
day when the amount of water thrown off by the shoot and of
mineral matter used in metabolism are nearing the minimum
by reason of the absence of light, lowered air temperature,
and consequent increased humidity of the air. These inhar-
monious conditions account almost wholly for the profusion
of guttation excretions or " dew drops " formed on the tips

5

iiutl luurj^ans of jj^niss bliules aiul loaves ot" low y^rowiiif^" plants
early iu the eveuiiijt?. Absorption by the root continues quite
vigorously after sunset by reason of the favorable tempera-
tures, and the au^^niented amount of fluid in the cortex of the
roots sets up a jjressuri" which ultimately f(jrces water into the
central cylinder and up throuj>;h the woody cells faster than it
may be used and transpired by the thin-walled cells of tin; leaves.
The vessels become filled with water which is forced out in
lic^uid form through the excretory openings iu the form of drops.
In some species the amount of water coming away from the
plant in this manner may reach quite an appreciable (piautity.

On the other hand the forenoon witnesses the rapid acceler-
ation of transjjiration by all parts of the shoot at a time when
the soil temperature is decreasing to a minimum. The in-
crease of transijiration continues until mid-afternoon, while
the temi)erature of the soil reaches a minimum two or three
hours earlier and then begins to rise, but does not do so suffi-
ciently to favor absorjition to any great extent. It is true of
course that the needs of the leaves may be j^artially met by the
activity of the rootlets w^liich lie nearer the surface.

The greatest amount of variation during twenty-four hours
that was recorded during thej'^ear amounted to 2° C. (3.6° F.).
This amplitude was shown on two occasions. June 28, 1902,
the temperature rose from 8° C. (46.4° F.) at 10 a. m. to 10° C.
(50° F.) at 10 p. m., in consequence of an increased tempera-
ture of the air, reaching a maximum of 27.8° C. (82° F.) at 1
p. m. The temperature of the soil fell from 12° C. (53.6° F.)
at 11 p. m., on July 16, 1902, to 10° C. (50° F. ) at noon on the
following day. Curiously enough this minimum coincided
exactly with that of the air a meter above the surface four
hours earlier. A reverse movement of any kind greater than
1° C. has not been recorded within the limits of twenty-four
hours. These slow changes in the temperature of the soil
would exert no stimulative iutiuence upon the action of the
roots. Thus, for instance, a sudden and great diminution of
the temperature of a root temporarily would increase the rate
of absorption which, however, soon begins to decrease again.
The least dailv variations occurred during the winter mouths,
when it did not exceed (1° C. (1.8° F.).

It is interesting to recall iu this connection some of my OAvn
observations on the temperatures of the roots of small herbs
and grasses in Arizona Avhich were growing in volcanic sand.
The observations were made by mercurial thermometers in
1898, and it was found that these organs Avere carrying on
their functions under temperatures as high as 44° C. (Ill -f
°F.) to 45.5° C. (114 °F.).

The general course of the temperature and the nature of
the variations is shown by the accompanying tracings (see
tigs. 2 to 7) and the following data:

6

Temperatures of soil at a depth of 1 foot (30 centimeters ) fro7n June, 1902, to
June, 1903 (degrees centigrade).

1902.

1903.

June. July.

i

Aug.

Sept.

Oct.

Nov.

Dec.

Jan.

Feb.

Mar.

Apr.

May.

Maximum. . .
Minimum . . .

Variation . . .

12.7
7.6

5.1

13.2
7.5

5.7

12.0
7.4

4.6

10.0
3.7

6.3

6. 0 0. 0
—1.5 —2.5

7.5 ! 2.5

3.5
-.3.0

4.0
3.0

1.0

7.0
5.0

2.0

9.5
7.0

2.5

8.0
4.0

4.0

9.0
6.0

3.0

The maximum temperature for the year, 13.2° C. (56° F.),
occurred in July, and the minimum, —3° C. (26.6° F.), in De-
cember, giving a total annual variation of 16.2° C. (29.4° F.).

When plants are cultivated in pots, in green houses, the
small volume of soil around the roots responds much more
rapily to changes in the temperature of the surrounding air
and to the influence of streams and sprays of water than does
the upper layer of soil in the open. In general the soil in the
green house will show a much higher average temperature,
which, with the other conditions mentioned, makes necessary
special treatment on the part of the gardener. If the natural
conditions of water supply by precipitation were complied with
the needs of the plant would by no means be met under the
altered conditions of temperature.

Fig. 2. — Tracings of thermographic curves of air, and of soil at a
depth of 30 centimeters (1 foot), June 9-30, 1902. The daily periodicity
and the direct effect upon the soil of changes in air temperatures are
illustrated. Centigrade scale. The temperature of the air is shown by
the uppermost of each pair of tracings.

Not the least interesting feature of the records are the data
concerning temperatures during the winter niontlis when vege-
tation is supposed!}' in a resting condition. The first genei-al
drop occurs earlv in Sei)tember when the temperature is h)w-
ered 5° to G° C. '(0° to 10° F. ). A month later a second gen-
eral downward movement was noticeable, in October, whicli
reached the freezing point on October 22. After a few days

^

a\rik^:\uunnnw

Fkj. 3. — Tnicings of tlicniio^'rapliic curves of tlie aii', ami of the soil
at a depth of 30 centiineters (1 foot), Juno 3()-August 4, li)02. The maxi-
mum temperatures of the year are iucluded. Centigrade scale. The
temperature of the air is sliown liy the uppermost of each pair of tracings.

fluctuation about this point a decrease began which reached
the annual minimum of — i^° 0. (2(5.4:° F.) on December 10.
A heavy snowfall was followed b}' a steady rise which carried
the temperature above the freezing point on December 15.
The total jDeriod during which temi^eratures below the freez-

8

iug point occuiTed included only fifty-four days, and the
thermograph recorded freezing temperatures on but fifty days
of this time. After the above date the rise was steady and
continuous with minor fluctuations.

I.V. Vr'V£^"g' ^^ ■ Sept- 1^ [gPJ-

Fig. 4. — Tracings of thermograpliic curves of the air, and of tlie soil
at a deptli of 30 centimeters (1 foot), August 4-September 1, 1902. A
period during which comparatively little fluctuation and variation occurs.
Centigrade scale. The temperature of the air is shown by the uppermost
of each pair of tracings.

Both the maxima and minima are higher in January than in
December and are still higher in February, at which time the
temperature of the air reaches its annual minimum. As a con-
sequence of the above conditions the roots of plants at a depth
of a foot (30 centimeters) in this locality find increasingly
more favorable conditions for activity after the latter part of
December. Growth and division of the cells, and the attend-
ant respiration would ensue in man}- species at an accelerated
rate over that prevalent in the earlier mouths of the winter
and the absorjDtion of mineral and other compounds could take
place to the utmost capacity of the plant. This will still be
more apj^arent when it is seen that the temperature of the soil
at a depth of a foot is but little lower in February than it is
in April or May at a time when the growth and formation of

9

new organs in the shoot is jit a iiiuxiniiiin. The jipproiich of
the sprinp: season in this locality therefore finds the root sys-
tems and absorbing organs of the vegetation which pene-
trates the soil to the given depth in a state of comi)aratively

Fig. 5.— Tracings of thermographic curves of the air, aud of tiie soil
at a depth of 30 centimeters (1 foot), September 1-29, 1902. A general
decrease in the temperature of the soil amounting to about 5° C. was
shown. Centigrade scale. The temperature of the air is shown by the
uppermost of each pair of tracings.

great activity, and it needs but the exposure of a few days or
even of a few hours in some plants to alloAv for very marked
development of the stems, leaves, and flowers.

Another i:)oint of interest in the present connection is the
fact that such notable differences are found between the temp-
eratures of the subterranean and aerial portions of the bodies
of plants at almost all seasons. During June, 11)02, the shoots
of herbaceous plants were in an atmosphere that varied be-
tween 8° C. (46.5° F.) and 34° C. (92.5° F.), while the roots
were between 8° C. (46.4° F.) and 18° C. (55.4° F.). As the
maxima and minima were not synchi'onous the actual differ-
ence between the temperature of twigs and leaves on the up-
per part of the plant, and roots on the lower amounted to as

10

much as 22° C. (nearlj^ 40° F. ) at certain times of the day.
Such conditions occur, though slightly less accentuated, during
the entire summer in this locality. It is evident without further
discussion that such differences in the temperature conditions
of the two poles of the j^lant must exert a more or less im-
portant influence on the transport of fluids and solutions from
one part of the j^lant to another. Eeferring to the in-evious
discussion concerning the comjiarative transpiration and ab-
sorj^tion during the day it is to be seen that the heightened
temjDerature of the shoot must operate in a simple jDhysical
way to greatly augment the amount of water thrown off while
the roots must take in water at the same time to meet the loss
at a temijerature as much as 40° F. lower.

Fio. G. — Tracings of thermographic curves of the air, and of the soil at
a depth of 30 centimeters (1 foot), September 29-November 3, 1902. A
further drop in temperature is shown which carries the soil below the
freezing point temporarily on the 22d, and finally for the season on the
29th. Centigrade scale. The temperature of tlie air is shown by the
uppermost of each pair of tracings.

During the movement of the water from the roots to the
leaves of grasses and other low growing plants, a total distance

11

of no more than 50 centimeters (20 inches) may be traversed,
occui)yin^- a matter of a few minutes, or an hour at most,
during which time the tem])erature is raised the al)Ove amount.
The warming of tlie licjuid as it passes upward tlirough tlie
living and nonliving cells is attended by alterations in its
solubility of mineral and organic substances and by a de-
creased capacity for holding gases in solution. The down-
Avard movement of solutions of sugars, acids, and nitrogenous

"X TmH^uVv^\huvU»^:icrDec. 's^gg/^goa.p

Fig. 7. — Tracings of thermographic curves of air, and of the soil at a
depth of 30 centimeters ( 1 foot), November 3-December 1.'), 1002. The
soil romaiued below the freezing point until December 8, when it began
to rise steadily, and did not fall below freezing point during the re-
mainder of the winter. Centigrade .scale. The temperature of the air
is shown by the uppermost of each pair of tracings.

12

substances from the leaves encounters the opposite set of
conditions. This movement takes place almost entirely by
osmose and diffusion and is a much more complicated process,
both chemically and physically, taking place in living cells
only. The cooling of the liquid would entail alterations in
its j)ower of carrying substances in solution and would also
alter its physical relations to the atmosjiheric gases present.

It may be said, in conclusion, that the facts disclosed as to
the actual temperatures in the soil, the diurnal and seasonal
changes therein, and as to the differences in temperature of
the aerial and underground portions of plants can not fail to
be of very great importance in the physical and chemical pro-
cesses, upon which growth, cell division, nutrition, and projDa-
gation dej)end. The determination of the effect of differences
in temperature between the roots and aerial shoots has received
but little consideration from the physiologist and the geogra-
pher. A careful analysis of the conditions and results of
experimental observations carried on with plants under arti-
ficial conditions, with the roots and shoots under abnormally
similar temperatures, would no doubt result in the detection
of many mistaken conclusions, especially in regard to absorp-
tion, trajislocation, and transpiration.

That soil temperatures and the relations of these tempera-
tures to those of the air must be of very great importance in
the cultivation of economic plants is self-evident, es^Decially
in species in which the desired useful portion is formed under-
ground and receives storage material formed by the activity
of the aerial organs. Thus, in the case of such j^lants as the
l^otato, certain mineral substances are absorbed from the soil at
a comparatively low temperature, carried aloft into the heated
leaves, where they participate in activities resulting in the
formation of sugars, starches, and other carbohydrates, per-
haps some nitrogenous substances as well, and then these
complex bodies are slowly diffused downward, with many ac-
companying chemical and physical modifications, to under-
. ground cool storage organs, where a condensation occurs and
the products are stored in insoluble form in the tuber.

contributions from the new york botanical

garden—No. 45

STUDIES IN PLANT HYBRIDS:
THE SPERMATOGENESIS OF HYBRID PEAS

By WILLIAM AUSTIN CANNON

NEW YORK
1903

[Reprinted from the Bulletin of the Torrey Botanical Club, 30 : 519-543, pi. 17-19.

October, 1903]

Studies in Plant Hybrids: The Spermatogenesis of Hybrid Peas*

By William Austin Cannon
(With Plati-.s 17-19)

Up to the present the studies on the origin and development
of the spores and sex-cells of hybrids have been made with no
definite knowledge of the manner of the variation of the hybrid
race and without regard to or accurate knowledge of the degree
of fertility or sterility of the forms, and it seems almost unneces-
sary to remark that no study can be at all complete, whatever its
results or conclusions may be, that does not take into considera-
tion each of these conditions.

Studies on the sporogenesis and spermatogenesis of hybrids in-
clude the researches on cannas by Guyer.f on cotton by Cannon,
on Drosera by Rosenberg, on Gladiolus by Metcalf, on pigeons by
Guyer and on Syringa by Juel. Nothing definite is known, so far
as I am aware, about the variation of any of these hybrids, and
as regards their fertility probably not one is entirely fertile, and
Syringa at least is thought to be quite sterile.

There are two or three conditions at least to which a hybrid
plant should conform in order to be well adapted for study from a
morphological standpoint. These are briefly : the hybrid must be
fertile ; it must vary in a definite and known manner ; and, finally,
it must be suitable in itself, as distinguished from all other require-
ments, for histological study.

One reason why the pea hybrids are so well adapted to experi-
mental researches is that they are fertile. This, indeed, was one
of the reasons that led Mendel to select them for his studies. I

*This study was made with the aid of Grant No. 27 of the Carnegie Institution,
t See bibliography at the end of this paper.

519

520 Cannon : Studies in Plant Hybrids :

have observed also that the peas used in the present study were
fertile ; they produced pods which were usually well filled with
seeds capable of germination. These observations are in harmony
with the manner of the formation of the spores ; no irregularities
whatever, which in any way were traceable to the plants as hybrid
organisms, were seen. If these observations are correct, the con-
elusion can hardly be avoided that the variations and reversions of
hybrid peas are associated with normal processes, such as are found
in the pure types, and hence that abnormalities are merely indica-
tions of the hybrid nature of the forms that bear them, although
by no means diagnostic characters. It is then a question fairly
open for the cytologist to determine, what the nature of those
processes are, and this has yet to be done.

The hybrids used in this study, * Fillbasket x Debarbieux and
Express x Serpette, showed variation after the law of Mendel,
although I should state that they were not examined more than was
necessary to surely establish this point. The seeds from which
the plants studied were derived were from those of the first gen-
eration. The Fillbasket x Debarbieux seeds were yellow and
green, nearly in the ratio 3:1. Of course most of the flower-buds
were removed for study, but in a few cases I left the flowers to
fertilize themselves. The plants which came from the green seeds
bore green seeds only, and those from the yellow seeds bore seeds
that were both yellow and green in the ratio 3:1. In the crosses
in which Serpette was used as one of the parents, as for example
Express x Serpette, the Serpette dwarf character was noticed in
the offspring as presumably the recessive one ; it appeared in at
least one third to one half the total number of third-generation
plants. In Express x Serpette there were seventy-two plants
which resembled the Express ancestor, and thirty which were
dwarfed and which were like the Serpette race. The illustrations
(/. 34-37) show fairly well the appearance of these reversionary
types and the pure forms also.

* I am indebted to Professor William Bateson for the seed from which the hybrids
were raised. Professor Bateson very generously sent me some first crosses, several of
the second generation and also the pure types.

I wish also to acknowledge the assistance of Mr. Homer D. House, of Colum-
bia University, for aid in preparing material for study, and that of Miss Palmyre de C.
Clarke, of the New York Botanical Garden, who prepared the bibliography.

The Spermatogenesis of Hybrid Peas 521

FlIXBASKET AND DeBARBIEUX, PuRE RaCES

No difference in the ori^^in and development of the microspore
was noted between the pure races Fillbasket and Debarbieux. In
both the reduced number of chromosomes was seen to be the
same, namely seven, and the chromosomes appeared to be quite
aHke in each ; to avoid repetition the general account of the matu-
ration processes will be deferred until the next topic. Favorable
material, however, gave opportunity to observe the sporogenous
divisions in Fillbasket and the results are here presented.

In the most of the sporogenous divisions the mitotic figures
are likely the same as in the somatic nuclei ; but in late anaphases
or early telophases certain characters were seen which were appar-
ently somewhat different. In early telophase of one of the later spor-
ogenous divisions the chromosomes are associated in pairs, touch-
ing by one, the polar, end. A nucleus with such chromosomes is
shown in /. 4. This division was evidently the last sporogenous
one, that is, the nucleus figured would become that of the spore
mother-cell without further division. Whether the chromosomes
are ever in touch throughout their whole length, as appears to oc-
bur in Fillbasket x Debarbieux (/. 6), was not determined. The
later history of the chromosomes is of great interest, since upon a
proper understanding of it may depend an appreciation of the sig-
nificance of the loops of chromatin which appear subsequently in
the nucleus as one of the prophases of the first maturation di-
vision. If, as supposed, the spireme of the nucleus of the spore
mother-cell is double, we might look for a previous doubling of
the chromosomes, although this might perhaps not follow. I have
seen no clear indication that the chromosomes form a double
spireme in the telophases of the last sporogenous division. The
behavior of the chromosomes at this time is as follows : As they
approach the poles of the spindle they are appreciably more uni-
formly arranged and more nearly parallel to the long diameter of
the spindle than in the other divisions in the sporogenous cells,
when at the poles the chromosomes unite or touch two and two
by the polar ends. In this condition they enter the resting stage ;
a continuous spireme was not seen. In some nuclei the adjoining
pairs of chromosomes were apparently united by their free ends,
but I did not surely determine this point. If the chromosome

522 Cannon : Studies in Plant Hybrids :

pairs are united in this manner a continuous spireme of single
chromosomes united end to end would be the result. The results
of my observations upon this point were somewhat contradictory^
and this may have been owing to the possibility that not all of the
nuclei studied were undergoing the last sporogenous division, and
that the pairing of the chromosomes and the form of telophase fol-
lowing are, in the pea, peculiar to the presynaptic nuclear division.
I hope at another time to give a more complete account of this
division.

When I observed the association of chromosomes as above de-
scribed I supposed that it was merely from chance, which indeed
may yet prove to be the case, and not the usual and normal occur-
rence in such cells, because it surely is not at all hkely that so no-
ticeable a thing could for any length of time escape the eyes of
trained cytologists. The better to reinforce the correctness of my
observation, or to prove it false as the case might be, I studied the
mitoses in the nuclei of somatic cells. As a rule, there is no indica-
tion whatever in the somatic nuclei of peas of the association of the
chromosomes in twos, but in one nucleus an appearance recalling
the pairing was observed. In this case a closer examination
showed that the association did not include all of the chromo-
somes and was very evidently one of chance merely ; so that it
seems to me that the association of the chromosomes in pairs in
the last sporogenous division in Fillbasket does not find a coun-
terpart in any somatic cells and is not the result of chance.

Fillbasket x Debarbieux

In studying the sporogeny of the hybrid Fillbasket x Debar-
bieux my aim has been rather to notice departures from the " nor-
mal," and by it to discover if possible the structural causes of the
remarkable " splitting " of the Mendelian hybrids, than to examine
critically and minutely the several nuclear divisions for themselves.
Accordingly I have observed the behavior of the chromosomes
more especially in the two maturation mitoses, in the first division
of the microspore and in the last division of the sporogenous cells.

The sporogenous cells of the hybrids were rather easily dis-
tinguished even in young anthers because of their position and by
the greater density of the cytoplasm, although it should be said

The Spermatogenesis of Hyijrid Peas 523

that it is not seldom difficult surely to determine the character of
a given cell. To study the divisions from which the sporiferous
tissue arises it is necessary to obtain very young anthers indeed,
those which are merely slight protuberances on the primordium of
the flower. The difficulties in the way of studying the anther are
decidedly increased from the fact that the inflorescence consists of
not more than four or five flowers, all usually in widely different
stages of development.

For the most part the nuclear division.s* in the sporogenous
cells of Fillbasket x Debarbieux are of the regular equational
type, quite the same as those in the somatic cells of the plant.
That is, fourteen segments are formed in the prophases and these
become arranged in the equatorial plate and in mctaphase are
already split before the migration to the poles of the spindle. In
anaphases the chromosomes are frequently L-shaped, although
not so sharply bent, and are arranged haphazard, that is, as purely
by chance. In the telophases the chromosomes unite end to end
and form a spireme ; this becomes changed into a delicate reticu-
lum in which all trace of the chromosomes as such disappears.

In some of the later sporogenous divisions a departure from
the usual appearance was noted as in the case of the pure form
Fillbasket. In Fillbasket x Debarbieux the same curious associ-
ation of chromosomes two and two was observed in a few cases.
F. 6 shows an anaphase of the last sporogenous division ; the pair-
ing was even more clearly marked in the preparation than in the
sketch. The pairing was with a single exception observed in
anaphase or early telophase only. The exception is shown by
f. J, which is an equatorial plate in probably the last division be-
fore synapsis. I have not seen the association in early anaphase
for the reason perhaps that the chromosomes are at that time
under the influence of the spindle fibers.

When the chromosomes reach the poles of the spindle they
unite by their polar ends, and these appear to fuse together, mak-
ing a more or less well-marked ring from which depend the chro-
mosomes in pairs. This is shown very imperfectly byy. 7, 8 and
p. In no case did I see a spireme which I could surely say was
continuous or which was made up of more than two elements ;
this may well be because the particular stage was wanting in the

524 Cannon : Studies in Plant Hybrids :

preparation. On the other hand instances were observed, hke/. 8,
in which a reticulum was being formed from the condition as just
described. The exact fate of the chromatin was thus not deter-
mined.

The association of chromosomes in pairs in the early history
of the reproductive cells recalls the similar condition in animals
as given by Montgomery * and Sutton, f But the significance of
the association in hybrids is not entirely clear, since with Mont-
gomery's hypothesis in mind that in pure races the two chro-
mosomes of each pair are of different parentage, it seems difficult
to conceive of Sutton's as applied to hybrids, that the members
of each pair cover the same field of development ; because in hy-
brids the latter can only be true in a broad way, depending ap-
parently upon the affinities of the pure parents of the form. It
may be, however, that the differences between the chromosomes
of any pair in hybrids capable of forming functional spores are so
slight that union is possible even if the potentials of the chromo-
somes are not identical.

After the sporogenous cells are cut off there ensues a long
period of growth during which the nucleus of the young spore
mother-cell undergoes characteristic changes. The chromatin of
the nuclear reticulum becomes gathered to one side of the nucleus
and the contracted condition known as sj-napsis results. The
preparation for synapsis takes place at an early time in the history
of the spore mother-cell, and the portions of the nuclear changes
resulting in synapsis which were seen may be briefly described.
The delicate reticulum which extends through the nuclear vacuole
of the resting nucleus becomes so changed that small masses of
chromatin of unequal form and size are found scattered along its
meshes, more especially in the peripheral part of the nucleus.
These masses are arranged in lines which often radiate from the
nucleolus as a center ; the linin threads which connect the masses
may be seen for a considerable distance from them extending into
the nuclear vacuole. From a position extending over the entire
periphery of the nucleus the chromatin becomes gathered to one

* Montgomery, T. H. The heterotypic maturation mitosis in Amphibia and its
general significance. Biol. Bull. 4: 259. 1903.

t Sutton, W. S. On the morphology of the chromosome group in Brachystola
7nag7ia. Biol. Bull. 4: 24. 1902.

The Spermatogenesis of HviiRin Peas 525

side of the vacuole ; when it enters into the synapsis stage it is
probably already in the form of a continuous spireme. It is much
bent with short turns and appears usually to be homogeneous and
not split ; the spireme at this time closely resembles the much
convoluted thread with its short, sharp turns as it emerges from
the synapsis stage. Since the chromatin assembles in a spireme
before the so-called synapsis condition the significance of the
latter is not apparent in these forms.

The cytoplasm of the spore mother-cell is reticular and it
always, from the spore mother-cell to the tetrad, retains this
structure. In this regard the pea is very different from the cotton
hybrid in which kinoplasm forms a marked character of the cyto-
plasm in the more advanced stages of spore formation, and from
some of the homologous cells of Syringa RotJioniagcnsis * in which
kinoplasm is strongly developed. The exaggerated development
of kinoplasm or filar-plasm in the cytoplasm of the spore mother-
cells is therefore not a constant or diagnostic feature of hybrids.

Although I did not study closely the origin of the achromatic
figure, I observed that it is derived from a multipolar one. The
multipolar spindle originates in groups of fibers which are placed
near the outer edge of the cytoplasm and which radiate from a
common point towards the nucleus. The place of genesis of the
achromatic figure is therefore very different from that of the cot-
ton hybrid, and the structural reason appears to lie in the presence
in the latter, and absence in the former, of a dense perinuclear zone
of cytoplasm. The achromatic figure in cotton originates inside
of this perinuclear zone, which thus appears to inhibit its more
peripheral origin. It may also be that the zone of denser cytoplasm
which in the cotton closely surrounded the nucleus not only formed
a limiting boundary for the spindle, but at the same time made
necessary for the mechanical support of the cell the formation of
an outer layer of kinoplasm, and that its absence in the pea not
only changed the structural relations of the parts of the cell but
altered the mechanical condition as well, making the formation
of the kinoplasmic layer unnecessary.

The first maturation mitosis is heterotypic. The chromosomes in
the metaphase are not different from those in the pure races Fillbasket

*Juel, H. O. See bibliography.

526 Cannon : Studies in Plant Hybrids :

and Debarbieux. Their relatively large size and small number
make the observation of them comparatively easy and sure. The
reduced number is seven ; this number was repeatedly seen in the
first division and was verified in the second mitosis and in that of
the division of the nucleus of the tetrad also.

The chromosomes in the anaphases of this division are more or
less bent and some show the splitting sometimes seen in these
stages of division. There was an equal apportionment of the
chromosomes, both as regards their number and size, to the
daughter nuclei. The single exception to this, and the only ab-
normality noticed in the hybrids, is shown in /. 2j. In the nu-
cleus from which the sketch was made one chromosome lagged
behind its companions ; fission, also, had not taken place.

The daughter nuclei undergo reconstruction to a greater de-
gree than was observed in the cotton hybrid, but on account of
the small size of the nuclei the study of them is difficult and unsat-
isfactory. The chromosomes in early telophase are much elon-
gated and appear to unite end to end to form a single spireme,
which later becomes lumpy and more or less reticular. The
identity of the chromosomes is quite lost.

In the prophases of the second division a spireme, which does
not seem to be split, may be seen, and the chromosomes formed
from it become associated in pairs in the nuclear plate, but they
may not on that account, as Mottier shows, be sister chromo-
somes. It is probable, however, that they are, and that they are
not homologous with the pairs found in the presynaptic division.
The metaphase or the anaphases present no noteworthy condition.
The reduced number of chromosomes, seven, was frequently ob-
served and the number verified.

In late anaphases of the second division the cell-plates sepa-
rating the daughter nuclei may be seen, and when the grand-
daughter nuclei are reconstructed cell-division occurs. Immediately
after the division of the cell the tetrads are angular in form, but
they soon become spherical and at length they assume the ellip-
tical form characteristic of the pollen-grain. During these changes
in form the tetrad increases much in size and the nucleus of the
young spore moves from its early central position to one end of the
cell and there prepares for the division of the nucleus of the mi-
crospore.

THK Sl'liKMATOGENESIS Ol IIVKKID PeAS 627

The first division of the nucleus of the spore (/. J2) appears
to be similar to that of the somatic nuclei except for the number
of the chromosomes and the shape of the spindle. The chromo-
somes are of course the reduced number, and the spindle, owing
perhaps to the eccentric position of the nucleus, is somewhat trun-
cated. As in the two preceding divisions, the distribution of the
chromosomes in this one was apparently perfectly equable. There-
fore in all and in each of the nuclear divisions which were directly
connected with the development of the microspores the number of
the chromosomes, and also presumably the amount of chromatin,
were equally distributed.

In late anaphase of the first division of the nucleus of the micro-
spore a cell-plate is seen separating the crescentic-shaped genera-
tive nucleus from the vegetative one. Reconstruction of the two
nuclei goes on at an unequal rate. The vegetative appears always
to form its wall in advance of the sister nucleus, and its chromo-
somes are also the first to become reticulated and lose their
identity. In the reconstruction of both nuclei excellent opportu-
nity was given to observe the beginning and extension of the nuclear
wall, which appear to occur precisely as Lawson has described.*

Serpftte, Pure Race.

The material of the pure parents of the Express x Serpette
hybrid at my disposal was not sufficient to permit an extended
study of either ; but some observations were made on the critical
stages of spore-formation in Serpette.

The first maturation division in Serpette is heterotypic and the
second homotypic. In the anaphases of the first division the
chromosomes were not seen to be split, either for part or all of
their length, as was the case in the other pure races and the hy-
brids also. In the telophases of the daughter nuclei the chromo-
somes unite end to end and form a continuous daughter spireme.
In both maturation mitoses the reduced number of chromosomes
was seven. No differences in form or size were noticed between
the chromosomes of this and those of the other pure races.

* Lawson, A. A. On the relationship of the nuclear membrane to the protoplast.
Bot. Gaz. 35 : 305. 1903.

528 Cannon : Studies in Plant Hybrids :

Express x Serpette

Like Fillbasket x Debarbieux, this is a second generation hybrid,
and it also, as has been described in a preceding paragraph, shows
the sphtting into several types after the law of Mendel. The re-
version of Express x Serpette as regards the quality of size and
form of the plants is fairly well indicated by/. 34-37, where the
pure parents are also represented. It should be noted that the
sketches are drawings from photographs and that the plants were
photographed at the same time and under precisely the same con-
ditions, so that the figures faithfully reproduce the relative forms
and sizes of all of the plants.

The sporogeny of Express x Serpette corresponds closely in
the points examined to that of Fillbasket x Debarbieux, although
in certain regards, which will be indicated, a difference between the
two was noted. Also some phases of development were repre-
sented in my preparations of Express x Serpette, not found in
those of the other hybrid, and these will be mentioned, in addition
to the preceding, rather than a more complete account. By this
method I shall avoid unnecessary duplication ; the two hybrids
are so nearly alike in the matter of spore origin and development
that a description of these processes in the one would for the most
part answer perfectly well for the other.

The series representing the nuclear divisions in young anthers
was not so complete in Express x Serpette as in Fillbasket x De-
barbieux, and possibly for this reason I have not so complete evi-
dence of the association of the chromosomes in pairs in the last
sporogenous division in the former as in the latter hybrid. It
was observed, however. An apparently good example of the
pairing was seen where perhaps one would not expect it, namely,
in the division by which the sporangial wall is separated from the
inner sporiferous tissue. The outer cell resulting from this division
by further division gives rise to three series of cells, the inner of
which forms the tapetum ; the inner cell of the division noted be-
comes converted without further division into a spore mother-cell.
The division referred to occurs rather late in the pea. Both of
the nuclei resulting from this division had their chromosomes ar-
ranged in twos. If the curious association of the chromosomes

The Spermatogenesis of HYnRin Peas 529

is an especial preparation for synapsis, the condition of the outer
one is surely paradoxical.

Tile nucleus of the spore mother-cell enters into a complete
resting condition in which it remains during much or most of the
period of growth of the cell. As a preparation for synapsis the
chromatin of the recticulum collects into masses of irregular shape
and size, more or less arranged in threads, about the periphery of
the nuclear vacuole. The chromatin-thread at length takes up a
position around or closely applied to the nucleolus and the synap-
sis stage results. When the mass of chromatin unloosens, it is
seen to be made up of a much bent ribbon. The spireme is at first
beaded in appearance, but it becomes more regular in outline and
also increases in its capacity for absorbing stains. The splitting of
the .spireme was not clearly seen, although evidences of it were no-
ticed. In a few cells I saw the chromatin-loops which are char-
acteristic of the prophases of the first maturation mitosis (/. ij-ij),
but in each case the loop was not contracted and did not show
the second fission which was ob.served in the cotton hybrid.

The two maturation divisions were quite like those seen in Ser-
pette and in Fillbasket x Debarbieux. The reduced number of
chromosomes was seven.

A difference in the origin of the multipolar spindle of the first
mitosis of Express x Serpette from that of Fillbasket x Debar-
bieux was noted (this stage was not seen in Serpette), but whether
it was an artefact or not I was unable, because of the lack of ma-
terial, definitely to determine. In several spore mother-cells the
cytoplasm was very dense in the immediate neighborhood of the
nucleus, forming what was in effect a perinuclear zone. In these
cells the multipolar spindle .seemed to have its origin wholly within
the nucleus. The linin became lumpy and very distinct, it formed
a mesh-work which gave the appearance of polarity, and had every
indication of being the genesis of the multipolar spindle. The
nuclear wall remained in some cases recognizable even when the
intranuclear material was of a pronounced fibrous character. The
departure from the extranuclear origin of the spindle observed in
Fillbasket x Debarbieux is certainly a marked one, and interesting
even if it proves in the end to be an artefact.

630 Cannon : Studies in Plant Hybrids :

Summary

The main points in this paper may be briefly stated as follows :

1. Both of the hybrids studied and reported on in this paper,
Fillbasket x Debarbieux and Express x Serpette, were of the sec-
ond generation, that is, their grandparents were pure races of
peas ; both of the hybrids exhibited variation after the Mende-
lian law.

2. The hybrids are fertile.

3. Both of the hybrids matured their spores quite as the pure
ancestral forms did. The first maturation mitosis was heterotypic,
the second homotypic.

4. The reduced number of chromosomes is seven, the somatic
number fourteen.

5. No abnormal mitoses were observed.

6. In the anaphases of the last sporogenous division of both
hybrids, and of the pure form Fillbasket, the chromosomes were
associated in pairs ; this condition was seen also in one nucleus of
Fillbasket x Debarbieux that was in the nuclear plate stage. The
paired condition is probably not the result of chance ; further
work, however, which the writer hopes to carry on, will be neces-
sary definitely to determine this point.

Conclusions

If the physical basis for the Mendelian variation of hybrid
peas was not demonstrated by a single histological study of them,
certain conclusions which touch more or less directly the causes
of the variation may nevertheless be legitimately although ten-
tatively drawn. In the first place it has been shown that the vari-
ation of the peas, in the third generation at least, is due entirely
to the union of gametes which originate and develop quite the
same as do those of the pure ancestral types, or in other words,
abnormalties and irregularities of nuclear divisions do not form the
basis for the variation of these hybrids.

What then is the probable structural basis for the Mendelian
variation of the hybrid peas ? As has already been intimated, I
think it is fairly well shown that this basis is the same as that of
the variations in the pure forms, and if so an understanding of the
latter should give the key to explain the former as well. But

The Spermatogenesis of IIvukid Peas 531

nothing is clearer than the fact that the present conception of the
essential nature of the maturation processes in pure races of plants
does not provide adequate grounds for this variation. This study
v/ill hardly profess to do this, but I may repeat what I have be-
fore suggested * and what Montgomery f and Sutton | have em-
phasized, that a study of the early history of the reproductive
cells may give some clue leading to a better understanding of the
meaning of heterokinesis, and, at the same time, a truer concep-
tion of the physical basis for variations, both in pure races and in
hybrids. With this idea in mind I have studied, as well as the ma-
terial at hand would permit, the divisions of the sporogenous
cells, more especially the last one, the division preceding the first
maturation mitosis, in the hybrids and in the pure form Fillbasket,
and I have found unmistakable evidence of an association of the
chromosomes two and two. Of course I cannot say positively
that this was not a chance association, since I could not trace the
later phases of these divisions as closely as would have been de-
sirable, but the pairing was observed in well prepared material and
had every appearance of being the normal and regular occurrence.
If further studies confirm this observation, it may prove of great
importance in explaining the structure of the heterotype rings.
There are difficulties in the way of positively identifying any
chromatin loop as having its origin in any given pair of chro-
mosomes, because the nucleus enters on a complete resting stage
at the close of the presynaptic mitosis and the identity of the
chromosomes is lost. Further studies may, however, give re-
sults that point to a connection between the two even if they do
not make the demonstration. For example, I have seen the union
of chromosomes by the polar ends only, in the telophases of the
last sporogenous division, and I failed to find, although I searched
very closely, the formation of a spireme, but in the place of it the
pairs of chromosomes appeared to be united at the polar ends. A
continuous spireme may not be formed in this division. It is con-
ceivable therefore that the chromosome pairs which enter the rest-
ing condition may be identical pair for pair with the loops which
are formed later.

* Cannon, 2.

I Montgomery, /. c.
I Sutton, /. c.

532 Cannon : Studies in Plant Hybrids :

Mottier * has described processes in the second maturation
division which should be considered in connection with the earlier
nuclear divisions as well. In the metaphase of the second division
the chromosomes are brought into the paired position in the
nuclear plate by the action of the fibers of the spindle ; that is,
the pairing here is probably due to the activity of the achromatic
elements. This seems to make unnecessary the conception of the
preliminary pairing of the chromosomes as a direct preparation
for heterokinesis, since the spindle can distribute the chromosomes
at its will, so to speak. It should be considered, however, that
the rings are bivalent, and are probably formed without the inter-
vention of the fibers, since the rings antedate the formation of the
spindle itself Also, if my observation is correct, the spindle
fibers are not instrumental in segregating the chromosomes in
the presynaptic division, because in that case, with possibly one
exception, the pairing was not observed until the chromosomes
were at the poles of the spindle and hence probably after the work
of the fibers of that dividing nucleus was done. So it may well
be that there is not only a difference in the homologies of the
chromosome-pairs of the metaphase of the second division and
those of the presynaptic one, but a difference in their affinities as
well ; consequently influences, whatever they may be, other than
those of the spindle, may cause the earlier pairing.

The foregoing considerations are based upon the idea of the
individuality of the chromosomes and upon that of their being
qualitatively unlike.f If these conceptions are valid we should

* Mottier, D. M. The behavior of the chromosomes in the spore mother-cells of
higher plants, etc. Bot. Gaz. 35 : 250. I903.

•f-The results of Rosenberg's studies (see bibliography) on the sporogenesis
of the natural hybrid between Drosera longifolia and Drosera rotundifolia seem to
throw objections in the way of too strict a conception of the individuality of chromo-
somes in hybrid plants. Rosenberg finds that the reduced number of chromosomes in
the Drosera hybrid may be 10, as in D. rotundifolia, or 20, as in D. longifolia, or 15,
which is half the somatic number in the hybrid. He also finds some evidence indicat-
ing that part of the chromosomes may be made up of four portions of chromosomes
each, and others of two halves of chromosomes each; the former kind are also larger.
In any case it is difficult to reconcile Rosenberg's results with the idea that the chro-
mosomes are distinct morphological entities ; the results point rather to the preserva-
tion by a portion of the spores of the purity of the chromatin, a fact which may prove
disconcerting to recent critics in this field (Cook; see bibliography). It should be

The Spermatogenesis of Hvurid Peas 533

have spores whicli miglit be of pure descent or might not be so
according to the manner of the orientation of the segments in the
metaphase of the first division. The chromosomes might be as-
sociated and distributed by chance as I have recognized (Cannon, i)
and as Sutton has dwelt on at length.* What the connection of
the chance distribution of the chromosomes may be with the
distribution of the characters among the hybrids is not at all
clear, since the relation of the chromosomes to the characters
is not known, but the cytologist has abiding faith that there
is some connection which will be demonstrable with the micro-
scope, and he looks to the field of hybrids and of hybridization as
a new opportunity for experimentation by which to supplement
and to substantiate the results of his studies. Experimentation
and cytological studies on hybrid forms in which the number of
chromosomes is known and in which the number of characters
has been determined should show a connection, if there is one,
between the number of characters that are transmissible and the
number of the chromosomes and their morphology. I wish here
to call attention to what very likely is merely a chance coinci-
dence, but nevertheless suggestive and interesting, namely, to the
agreement in the number of the groups of constant differentiating
characters as given by Mendel for the peas used by him in ex-
periments with the reduced number of chromosomes in the peas.
As given by Bateson (Bateson, i), the characters are (a) differences
in the length and color of the stem ; (d) in the size and form of the
leaves; (r) in the position, color and size of the flowers ; (c/) in
the length of the flower-stalk ; (i-) in the color, form and size of
the pods ; (/) in the form and size of the seeds ; (^) in the color
of the seed-coats and cotyledons. In the peas there are evidently
more appreciable characters than chromosomes, and therefore the
chromosomes bear more than one character each. If then the
chromosomes maintain their individuality certain characters might
be associated together in the same chromosome, and might be

borne in mind, however, that a difficulty in drawing conclusions at present from the
results of Rosenberg's studies of the sporogeny of the Drosera hybrid may lie in the
partial sterility of the form. (See Focke, Ptlanzenmischlinge, 155.) Further work
on this hybrid is promised and will be awaited with interest.

* Sutton, W. S. The chromosomes in heredity. Biol. Bull. 4: 231. 1903.

534 Cannon : Studies in Plant Hybrids :

coupled together in the hybrid organism, not being capable of sep-
arating from each other — correlations are not unknown among
hybrids. The idea suggests itself that the case might be very dif-
ferent if the number of the chromosomes in relation to that of the
characters were greater than in the pea ; the chance association and
distribution of chromosomes representing portions of characters
might not bring about so evident a distribution of characters in
the hybrid generations, the Mendelian reversions. May it not be
possible then that fortune favored Mendel when he selected for his
experimentation plants which had so small a number of chro-
mosomes ? *

New York Botanical Garden.

List of the More Important Recent Works on Plant Hybrids

Abbado, M. L'ibridismo nei vegetali. Nuovo Giorn. Bot. Ital. II. 5 : 76, 265. 1898.

Ames, 0. (i) Lobelia inflatay^cardinalis. Rhodora, 3 : 296. 1901 ; (2) Repro-
duction in relation to problems in hybridization. Am. Gard. 22 : 130. 1901.

Andrews, A. L. A natural hybrid between Ilabenaria lacoa and H. psychodes.
Rhodora, 3 : 245. 190 1.

Arcangeli, G. (i) Sopra alcuni Norcissiis. Bull. Soc. Bot. Ital. 2: 480. 1893;
(2) Sopra varie mostruosita dell' Ajax odorus Car., e della sua probabile origene.
Bull. Soc. Bot. Ital. 2 : 290. 1893.

Bailey, L. H. (i) Crosses and crossing of plants. Gard. and For. 5: 2. 1892.
(Reprinted in) Gard. Chron. HI. 11 : 235, 256, 298. 1892 ; (2) Cross-breeding
and hybridizing. (Rural Library, v. I, no. 6. Ap) 1892; (3) Fruit items.
Am. Gard. 14: 500. 1893; (4) Hybridisation in the United States. Jour. Roy.
Hort. Soc. 24: 209. 1900 ; (5) Plant-breeding. New York, 1895; (6) The
hybridity corollary. Bot. Gaz. 22 : 460. 1896.

Barron, L. (i) The hybrid origin of roses. Am. Gard. 22: 5. 1901 ; (2) The
varying quality of adaptability in hybrid types. Am. Gard. 24 : 200. 1903.

Bateson, W. (i) Hybridisation and cross-breeding as a method of scientific investi-
gation. Jour. Roy. Hort. Soc. 24: 59. 1900; (2) Mendel's principles of he-
redity. A defense. With a translation of Mendel's original papers on hybridisa-
tion. Cambridge, 1902; (3) Note on the resolution of compound characters by
cross-breeding. Proc. Cambridge Phil. Soc. 12 : 5°- I903-

Bateson, W., & Saunders, E. R. Experimental studies in the physiology of heredity.
Rep. Evolution Committee Roy. Soc. Report I. 1902.

*The data at hand showing the number of chromosomes in the somatic cells of
Mendelian hybrids are unfortunately not complete. According to Wilson (The Cell,
206), in wheat there are 16 chromosomes, and in cattle and rats the number is probably
the same ; the mouse has 24 chromosomes ; according to the present study the pea has
14. On the other hand, there are 56 chromosomes in the somatic nuclei of hybrid
cotton, but the nature of the variations in the cotton, whether or not Mendelian, has
not, I believe, been determined.

The Spkrmatogenesis of IIvukid Peas 635

Beck von Mannagetta, G. Ueber Mischfriichte (Xenien) und deren Enlstehung.
Wiener Illustr. (iartenzeitung. Ap 1895.

Beeby, W. H. On natural hybrids. Jour. Bot. 30 : 209. 1892.

Berg, F. Ro^Licnenziichtung 1890. Hot. Centralb. 46: 183. 189I.

Blackman, V. H. Some recent work on hybrids in plants. New Phytol. i : 73. 1902.

Bonnet, E. Remarques sur quelques hybrides, etc. Jour, de Hot. 1 1 : 239. 1897.

Briquet, J. Sur un hybride nouveau de la famille des Ombelliftires. Bull. Herb.
itoiss. 4 : 354. 1896.

Britton, E. G. (i ) A hybrid moss. Plant World, i : 138. 1898 ; (2) On a hybrid
growing with Aphanorliegma serrata SuU. Bull. Torrey Club, 22 : 65. 1895.

Brizi, A. Su Timpollinazione nel gen. Cucurbita. Bull. Soc. Bot. Ital. 7 : 217.
189S.

Camus, E. G., et Duffort, L. Orchidees hybrides ou critiques du Gers. Bull. Soc.
Hot. France, 45 : 433. 1898.

Cannon, W. A. ( i ) ^ cytological basis for the Mendelian laws. Bull. Torrey Club,
29: 657. 1902; (2) Studies in plant hybrids: The spermatogenesis of hybrid
cotton. Bull. Torrey Club, 30 : 133. 1903.

Castillon, De. Hybridation artificielle. Revue Hort. 64 : 541. 1892.

Chauveaud, G. Sur les caractdres internes de la graine des vignes et leur emploi dans
la determination des esp^ces et la distinction des hybrides. Compt. Rend. 118 :
485. 1894.

Cogniaux, A. Laelio-Cattleya Binoti hyb. nat. nov. Card. Chron. HI. 28 : 370. 1900.

Cook, 0. F. Evolution, cytology and Mendel's laws. Pop. Sci. Monthly, 63: 219.
1903.

Correns, C. (i) Bastarde zwischen Maisrassen, mit besonderer Beriicksichtigung der
Xenien. Bibliotheca Botanica, Heft 53. 1901 ; (2) Die Ergebnisse der neuesten
Bastardforschungen fiir die Vererbungslehre. Ber. Deuts. Bot. Gesells. 19: (71).
1902 ; (3) Die Merkmalspaare beim Studium der Bastarde. Ber. Deuts. Bot.
Gesells. 21 : 202. 1903; (4) Gregor Mendd's " Versuche iiber Pfianzen-Hybri-
den" und die Bestatigung ihrer Ergebnisse durch die neuesten Untersuchungen.
Bot. Zeit. 58^: 229. 1900; (5) G. Mendel's Kegel iiber das Verhalten der
Nachkommenschaft der Rassenbastarde. Ber. Deuts. Bot. Gesells. 18 : 158.
1900; (6) Neue Untersuchungen auf dem Gebiet der Bastardirungslehre. Bot.
Zeit. 61'^: 113. 1903; (7) Scheinbare Ausnahmen von der Mendel'schen
Spaltungsregel fiir Bastarde. Ber. Deuts. Bot. Gesells. 20: 159. 1902; (8)
Ueber Bastarde zwischen Rassen von Zea Mays, nebst einer Bemerkung iiber die
"faux hybrides" Millardet's und die " unechten Bastarde" de Vries'. Ber.
Deuts. Bot. Ge.sells. 19 : 211. I901 ; (9) Ueber Bastardirungsversuche mit
yJ/;;vz/5//zV-Sippen. Ber. Deuts. Bot. Gesells. 20: 594. 1902; (10) Ueber den
Modus und den Zeitpunkt der Spaltung der Anlagen bei den Bastarden vom Erbsen-
Typus. Bot. Zeit. 60^ : 65. 1902; (11) Ueber diedominierenden Merkmale der
Bastarde. Ber. Deuts. Bot. Gesells. 21: 133. 1903; '^'2) Ueber Levkojen-
bastarde. Zur Kenntnis der Grenzen der Mendel'schen Regeln. Bot. Centralb.
84: 97. 1900; (13) Untersuchungen iiber die Xenien bei Zea Mays. Ber.
Deuts. Bot. Gesells. 17: 410. 1899; (14) Weitere Beitrage zur Kenntnis der
dominierenden Merkmale und der Mosaikbildung der Bastarde. Ber. Deuts. Bot.
Gesells. 12 : 195. 1903-

Crepin, F. (l) Rosae hybridae. Etudes sur les Roses hybrides. Bull. Bot. Soc.
Roy. Belg. 33: 7. 1894; (2) Note upon a probable hybrid of Rosa Carolina
L. and Rosa nitida Willd. Rhodora, 2 : 112. 1900.

536 Cannon : Studies in Plant Hybrids :

Daniel, L. ( i ) Les variations specifiques dans la greffe, ou hybridation asexuelle,
Rapp. Congr. hybrid, de la vigne. Lyon. 1902; (2) Sur quelques applications
pratiques de la greffe herbacee. Rev. Gen. Bot. 6 : 356. 1894.

Daveau, J. Sur une Scrofulaire hybride. Bull. Soc. Bot. France, 44 : 270. 1897.

Davenport, G. E. (i) Abnormal forms, and hybridity in ferns. Linnaean Fern
Chapter, Boston Meeting, i. 1899 ; (2) A plumose variety of the ebony spleen-
wort. Rhodora, 3 : I. 19OI ; (3) Aspidiuin cristatum X marginale Daven-
port. Gard. and For. 9 : 444. 1896.

Davidson, A. Fentstemon Pa7-ishii, a hybrid. Bull. So. Gal. Acad. Sci. i : 141.
1902.

Deane, H. & Maiden, J. H. Further notes on supposed hybridisation amongst
eucalypts ; including a description of a new species. Proc. Linn. Soc. N. S. W.

36 : 339. 1901.
Devansaye, A. de la. Fertilisation of the genus Anthu7-iuin. Jour. Roy. Hort. Soc.

24 : 67. 1900.
Dollo, L. Les lois de revolution. Bull. Soc. Belg. de Geologie, 7: 164. 1893.
Dorner, F. Crossing and hybridizing. Am. Florist, 16 : 752. 1901.
Druery, C. T. (i) Crosses or hybrids? Gard. Chron. IIL 29: 32. 1901 ; (2)

Fern crossing and hybridising. Jour. Roy. Hort. Soc. 24 : 2S8. 1900.
Emerson, R. A. Preliminary account of variation in bean hybrids. Nebraska Agr.

Exp. Sta. Ann. Rep. 15 : — . 1902.
Evans, W. E. International conference on plant breeding and hybridization. Exp.

Sta. Rec. 14: 208. 1902.
Ewart, J. C. (l) Intercrossing and hybridising. Gard. Chron. III. 26: 133.

1899; (2) Intercrossing and interbreeding as causes of variation. Rep. Brit.

Assoc. Adv. Sci. 1901 : 672. 1901.
Farmer, J. B. On the structure of a hybrid fern. Ann. Bot. 11 : 533. 1897.
Fernald, M. L. (i) As/er undulattis X Novi-Belgii. Rhodora, 4: 186. 1902;

(2) Some undescribed varieties and hybrids of Carex. Rhodora, 2 : 170. 1900.
Focke, W. 0. (i) Beobachtungen an Mischlingspflanzen. Abh. Nat. Ver. Bremen,

j2 : 403. 1893. (2) Neue Beobachtungen iiber Artenkreuzung und Selbststeril-

itat. Abh. Nat. Ver. Bremen, 14 : 297. 1897 ; (3) Ueber Unfruchtbarkeit bei

Bestaubung mit eigenem Pollen. Abh. Nat. Ver. Bremen, 12 : 409, 495. 1893.
Gagnepain, F. (l) A propos d'hybrides. Actes i'^'' Congr. Internal. Bot. Paris,

359. 1900; (2) Sur le pollen des hybrides. Bull. Soc. Nat. Autun, 14: Pro-

ces-verbaux, 20. 190 1 ; (3) Sur un hybride artificiel des Lychnis diurna et

vespertina. Bull. Soc. Bot. France, 43 : 129. 1896.
Gardner, F. D. Field experiments with corn. Bull. 111. Exp. Sta. 37 : 1,20. 1895.
Gauchery, P. Notes anatomiques sur I'hybridite. Assoc. Fr. Congr. d'Ajaccio, 1902.
Gerber, C. De la fecondation directe chez quelques plantes dont les fleurs semblent

adaptees a la fecondation crois^e. Compt. Rend. 126 : 1734. 1898.
Giltay, E. Ueber den directen Einfluss des Pollens auf Frucht- und Samenbildung.

Jahrb. Wiss. Bot. 25 : 489. 1893.
Glatf elter, N. M. ( l ) A study of the relations of Salix nigra and Salix amygdal-

oides, together with the hybrids arising from them as these species exhibit them-
selves in the vicinity of St. Louis. Trans. Acad. Sci. St. Louis, 6 : 427. 1894 ;

(2) Notes on Salix longipes. Ann. Rep. Mo. Bot. Gard. 9 : 43. 1897; (3)

Relations of Salix Missouriensis Bebb to S. cordata Muhl. Trans. Acad. Sci.

St. Louis, 7: 137. 1896; (4) Salix cordata X sericea. Bot. Gaz. 22: 392.

1896.

The Spermatogenesis of IIvukih Peas 537

Guyer, BS. F. Some notes on hybridism, variation and irregularities in the division of

the germ-cell. Science, II. 15: 530. 1902.
Hansen, G. The orchid hybrids. ( With first supplement. ) London, 1 895. Second

supiilcrjient. lierkeley, Cal., 1897.
Hays, W. M. (l) lirceding staple food jiiants. Jour. Roy. llorl. Soc. 24: 257.

1900; (2) Methods of breeding wheat. liuJi. I niv. Minn. Agr. Exp. Sta. 62:

422. 1S99,
Heidemann, C. W. H. Secondary effect of foreign pollen. Exp. Sta. Rec. 5 : 659.

1.S94.
Heinz, A. Ueber Scolopendrium hybridum Milde. Her. Deuts. Hot. Gesells. 10 :

413. 1892.
Henry, L. Cro.ssings made at the Natural History Museum of Paris from 1887 to 1899.

Jour. Roy. Hort. Soc. 24 : 218. 1900.
Henslow, G. (l) Greenhouse rhododendrons. Gard. Chron. III. 9: 618. 1891;

(2) Hybridisation and its failures. Jour. Roy. Hort. Soc. 24 : 76. 1900.
Hildebrand, F. (l) Einige weitere Beobachtungen und Experimente an Oxalis-

Arten. Bot. Centralb. 79 : 1,35. 1S99 ; (2) Ueber Bastardierungsexperimente

zwischen einigen I/epa(ica-a.r\.en. Bot. Centralb. 84: 65. 1900 ; (3) Ueber

Hcterostylie und Bastardirungen bei /'()r.j)'/'//w. Bot. Zeit. 52 : 191. 1894.
Hill, E. J. (l) Early experiments in crossing plants. Gard. and For. 9 : 32. 1896 ;

(2) Two noteworthy oaks. Bot. Gaz. 26: 53. 1898.
Hitchcock, A. S. A hybrid Bapiisia. Bot. Gaz. 19 : 42. 1894.
Hooker, J. D. CattUyay^ IVhitei. Curt. Bot. Mag. III. 56: //. 772"/. 1900.
Hurst, CO. (i) Curious generic hybrids. Jour. Roy. Hort. Soc. 21 : 471. 1898;

(2) Mendel's "law" applied to orchid hybrids. Jour. Roy. Hort. Soc. 26: 688.

1902; (3) Mendel's methods of plant-breeding. Gard. Chron. HI. 33: 76.

1903; (4) Notes on some curicsities of orchid breeding. Jour. Roy. Hort. Soc.

21: 442. 1898; (5) Notes on some experiments in hybridisation and cross-
breeding. Jour. Roy. Hort. Soc. 24 : 90. 1900.
Hy, F. C. Orchis pseudo-militaris hybrid, nov. Actes I" Congr. Internal. Bot. Paris,

362. 1900.
Jack, J. G. Hybrid birches. Gard. and For. 8 : 243. 1895.
Janczewski, E. de. (l) Etudes morphologiciues surle genre Anemone L. Rev. Gdn.

Bot. 4 : 241, 289. 1892; (2) Hybridesdesgroseillersa grappes. Bull. Internal.

Acad. Sci. Cracovie, 6 : 295. 1901.
Jencic, A. Untersuchungen des Pollens hybrider Pfianzen. Oesterr. Bot. Zeitschr.

50 : I, 41, 81. 1900.
Jones, H. L. A graft hybrid. Bot. Gaz. 18: in. 1893.
Jouin, E. Can hybrids be obtained by grafting ? Jour. Roy. Hort. Soc. 24: 237.

1900.
Juel, H. 0. Beitrage zur Kentniss der Tetradentheilung. Die Tetradentheilung bei

einer hybriden Pflanze. Jahrb. Wiss. Bot. 35 : 638. 1900.
Keller, I. A. Notes on the study of cross-fertilization of flowers by insects. Proc.

Acad. Nat. Sci. Phila. 1895 : 555. 1895.
Kittel, G. Die Kreuzung der Orchideen. Gartenflora, 41 : 160. 1892.
Klinge, J. Zwei neue bigenere Orchideen-Hybride. Acta Hort. Petrop. 17': 205.

1S99.
Koehne, E. Zwei Pfropfbastarde von Crataegus monogyna und Mespilus Germanica.

Gartenflora, 50 : 624. 1901.

538 Cannon: Studies in Plant Hybrids:

Korshinsky, S. (l) Heterogenesis und Evolution. Naturwiss. Wochenschr. 14:
273. 1899; (2) Uebereineneuebigenere Hybride : Cucumis Melo \.. Y^ Citrul-
lus vulgaris Schrad. Bull. Acad. Sci. St. Petersb. V. 6 : 322. 1897.
Krasan, F. Variety, race, modification. Actes I" Congr. Internat. Bot. Paris, 366.

1900.
Kiikenthal, G. (i) Die Benennung der Hybriden. Allg. Bot. Zeitschr. i: 60.
1895 ; (2) Die Hybriden der Carex caespitosa L. und der Carex stricta Good.
Oesterr. Bot. Zeitschr. 46 : 161. 1896.
Kiister, E. Die Mendel' schen Regeln, ihre urspriingliche Fassung und ihre modernen

Erganzungen. Biol. Centralb. 22 : 129. 1902.
Kuhn, H. Die Baumwolle, ihre Cultur, Structur und Verbreitung. Wien, 1892.
Kusnezow, N. Uber den Polymorphismus der Veronica Teturiwn (L. ) Wallr.

Bull. Acad. Sci. St. Petersb. V. 6^: I75- 1897.
Laubert, R. Anatomische und morphologische Studien am Bastard Laburnum

AdatniVmx. Beih. Bot. Centralb. 10 : 144. 1901.
Lemoine, E. Hybrids between the common lilac and the laciniated Persian lilac.

Jour. Roy. Hort. Soc. 24 : 299. 1900.
Leveille, H, (i) Les hybrides en general et les Epilobes hybrides de la France.
Bull. Acad. Internat. Geogr. Bot. 8 : I33- 1899; (2) Nouvelle classification des
hybrides. Actes i""' Congr. Internat. Bot. Paris, 355. 1900.
Levier, E. N^otulipes et Paleotulipes. Malpighia, 8 : 401. 1894.
Lewton-Brain, L. Hybridization of the sugar-cane. West Indian Bull. 4 : 63. 1903.
Linton E. F. (i) Experiments in cross- fertilisation of 5a/?V^j. Jour. Bot. 36 : 122.

1898 ; (2) Two new willow-hybrids. Jour. Bot. 32 : 201. 1894.
Lowe, E. J. (l) Ferngrowing. London, 1895; (2) Hybrid ferns and crossed va-
rieties. Jour. Roy. Hort. Soc. 12 : 505. 1890.
Macfarlane, J. M. ( i ) A comparison of the minute structure of plant hybrids with that
of their parents, and its bearing on biological problems. Trans. Roy. Soc. Edinb.
37I : 203. 1892 ; (2) Hybridisation : its benefits and results to ornamental horticul-
ture. Gard. Chron. III. 14 : 361, 395. 1893 ; (3) Illustrations of heredity m plant
hybrids. Bot. Gaz. 16 : 255. 1891 ; (4) Observations on pitchered insectivor-
ous plants. VIII. On hybridity, etc. Ann. Bot. 7 : 445- 1893; (5) Observa-
tions on some hybrids between Droserafilijormis and D. intermedia. Contr. Bot.
Lab. Univ. Penn. 2 : 87. 1898 ; (6) The colour, flowering period and constitu-
tional vigour of hybrids. Gard. Chron. III. 9 : 753. 1891.
Malinvaud, E. The species and hybrids of ^.fw^'/^a. Jour. Bot. 38 : 171. 1900.
Marggraff, G. Anatomic der Ca7'ex-hxK.(tVi. (Inaug. Diss. Univ. Erlangen. ) Leip-
zig. 1896.
Marshall, E. S. Do natural hybrids exist ? Jour. Bot. 31 : 20. 1893.
Masters, M. T. Hybrid conifers. Jour. Roy. Hort. Soc. 26 : 97. 1901.
Maxcn, W. R. Notes on the validity of Asplenium ebenoides as a species. Bot. Gaz.

30 : 410. 1900.
Mayer, A. Salix Ratisbonensis, nov. hybr. Denies. Bot. Gesells. Regensb. 7 : 58.

1898.
Meehan, T. (i) Notes on hybrids. Jour. Roy. Hort. Soc. 24 : 337. 1900; (2)

The occasional cross. Bot. Gaz. 17 : 420. 1892.
Mell, P. H. Hybrids from American and foreign cottons. Bull. Ala. Exp. Sta. 83 :
380. 1897.

The Spermatogenesis of Hybrid Peas 539

Mendel, G. ( i ) Experiments in plant liyliridisalion (translation). Jour. Roy. Ilort.
Soc. 26 : 1. 1901 ; (2) Versuche iiber Pllanzenhybriden (reprint). Flora, 89:
364. 1901 ; Id. Ostwald's Klassiker der exakten Wiss. No. 121. I901.

Metcalf, H. Certain problems relating to the individuality of chromosomes. Proc.
Neb. Acad. Sci. 7: 109. 190I.

Millardet, A. (l) Essai sur I'hybridation de la vigne. Mem. Soc. Sci. Bordeaux,
IV. 2: 299. 1891 ; (2) Importance de I'hybridation pour la r^constitution des
vignobles. Compt. Rend. 119: 1176. 1894; (3) Note sur I'hybridation sans
croisement ou fausse hybridation. Mem. Soc. Sci. Bordeaux, IV. 4 : 347.
1894.

Moenkhaus, W. J. Some stages in hybrid development. Science, II. 11 : 788. 1900.

Molliard, M. Note sur les particularites t[ue presentent les fleurs doubles du Petunia
Jiybrida. Bull. Soc. Bot. France, 40 : 330. 1893.

MuUer, F. (l) Ein Versuch mit Doppelbestaubung. Flora, 83: 474. 1897; (2)
Mischlinge von Ruellia formosa und sylvaccola. Abh. Nat. Ver. Bremen, 12 :

379- i89;v
MuDSOn, W. M. (l) Preliminary notes on the secondary effects of pollination. Maine

Agric. Exp. Sta. Rep. 1892: 29. 1892; (2) The development of a tomato

hybrid. Proc. Soc. Prom. Agric. Sci. 21 : 1900.
Murbeck, S. Neue oder vi'enig bekannte Hybriden. Acta Horti Bergiani, 2* : 3.

1S94.
Naudin, C. Variation desordonnee chez les hybrides. Revue Hort. 70 : 509. 1898.
Oliver, G. W. Hybridization. Am. Gard. 20 : 397. 1899.
Pammel, L. H. (l) Experience in crossing cucurbits. Am. Gard. 14 : 459. 1893;

(2) On the crossing of cucurbits. Bull. Iowa Agric. Exp. Sta. 19: 595- 1892;

(3) Results of crossing cucurbits. Bull. Iowa Agric. Exp. Sta. 23 : 906.
1894; (4) The effects of cross-fertilization in plants. Trans. Iowa Hort. Soc.
26: 439- 1891.

Parmentier, P. Une nouvelle fougere hybride. Bull. Acad. Internal. Geogr. Bot. 9 :

40. 1900.
Paulesco, P. R6cherches sur la structure anatomique des hybrides. (Th^se) Univ.

(leneve Lab. Bot. IV. 11 : 103. I900.
Petrucci, G. B. Ricerche anatomiche sopra la Chamaerops humilis (Linn.), la

Phoenix dactylif era (Linn.) ed i loro pretesi ibridi {^Alicrophoenix). Malpighia,

14 : 306. 1900.
Poeverlein, H. Die bayerischen Arten, Formen und Bastarde der Gattung Potentilla.

Denks. Bot. Gesells. Regensb. 7: 147. 189S.
Preda, A. Contributo alio studio delle narcissee italiane : Ibridi. Nuovo Giorn.

Bot. Ital. II. 3 : 404. 1896.
Reed, M. (l) Cross and self-fertilization. Bot. Gaz. 17: 330. 1892; (2) Cross-
fertilization of petunias. Bot. Gaz. 19: 336. 1894.
Rehder, A. Notes on liybrids of Quercus ilicifolia. Rhodora, 3 : 137. 1901.
Rimpau, W. Kreuzungsprodukte landwirthschaftlichen Kulturpflanzen. Landw.

Jalirb. 20 : — . 189I.
Rodigas, E. Injudicious hybridisation. Gard. Chron. III. 17 : 16. 1895.
Rolfe. R. A. Hybridisation viewed from the standpoint of systematic botany. Jour.

Roy. Hort. Soc. 24 : 181. 1900.
Rosenberg, 0. Das Verhalten der Chromosomen in einer hybriden Pflanze. Ber.

Deuts. Bot. Gesells. 21 : 110. 1903.

540 Cannon : Studies in Plant Hybrids :

Rowlee, W. W. & Wiegand, K. M. Salix Candida Willd. and its hybrids. Bull.

Torrey Club, 23 : 194. 1896.
Roze, E. Le Geum Kivaii-urbanum. Bull. Soc. Bot. France, 43 : 273. 1896.
Sargent, C. S. A hybrid walnut-tree. Gard. and For. 7 : 434. 1894.
Saunders, E. R. On a discontinuous variation occurring in Biscutella laevigata. Proc.

Roy. Soc. 62 : II. 1897.
Schweitzer, K. 0. Die Baumwolle nebst Notizen iiber die Cultur und Verbreitung

in Amerika. Ausstellungsbericht. Bern. 1894.
Scott, W. B. On variations and mutations. Am. Jour. Sci. III. 48 : 355. 1894.
Slosson, M. (l) Experiments in hybridizing ferns. Fernwort Papers, 19. I900 ;

(2) The origin oi Aspletiiuni ebenoides. Bull. Torrrey Club, 29: 487. 1902.
Small, J. K. Some new hybrid oaks from the southern states. Bull. Torrey Club,

22 : 74. 1895.
Spillman, W. J. (i) Exceptions to Mendel's law. Science, II. 16 : 709, 794. 1902 ;

(2) Quantitative studies on the transmission of parental characters in hybrid off-
spring. Bull. Exp. Sta. U. S. Uept. Agric. 115: igoi.
Starnes, H. N. Cotton crosses and hybrids. Bull. Ga. Exp. Sta. 24: 124. 1894.
Strasburger, E. (i) Ueber Befruchtung. Bot. Zeit. 59^: 353. 1901 ; (2) Ver-

suche mit diocischen Pflanzen in Riicksicht auf Geschlechtsvertheilung. Biol.

Centralb. 20: 766. 1900.
Stuart, C. A few notes on reproduction in hardy plants by means of hybridising

species and crossing varieties. Jour. Roy. Hort. Soc. 24 : 280. 1 900.
Swingle, W. T. Some theories of heredity and of the origin of species considered in

relation to the phenomena cif hybridization. Bot. Gaz. 25 : III. 1898.
Swingle, W. T. & Webber, H. J. Hybrids and their utilization in plant breeding.

Yearb. U. S. Dept. Agric. 1897 : 383. 1898.
Thiselton-Dyer, W. T. A Aa/^wc/iOd'' hybrid. Ann. Bot. 17: 435. 1903.
Tobler, F. Fortschritte der pflanzlichen Bastardforschung. Naturw. Rundschau,

17: 640. 1902.
Trabut, L. Eucalyptus hybrids in the Mediterranean region. Jour. Roy. Hort. Soc.

24 : 250. 1900.
Tschermak, E. (i) Der gegenwartige Stand der Mendel'schen Lehre und die Ar-

beiten von W. Bateson. Zeits. Landw. Versuchsw. Oesterr. 5 : — . 1902 ; (2) Me-

thoden und Gesetze der kiinstlichen Kreuzung. Wiener Illustr. Garten-Zeitung.

4 : — . 1903 ; (3) Ueber Correlation zwischen vegetativen und sexualen Merkmalen

an Erbsenmischlingen. Ber. Deuts. Bot. Gesells. 20 : 17. 1902; (4) Ueber die

Gesetzmassige Gestaltungsweise der Mischlinge. Zeitschr. Landw. Versuchsw.

Oesterr. 5: 807. 1902; (5) Ueber kunstliche Kreuzung bei Pisttm sativum.

Zeits. Landw. Versuchsw. Oesterr. 3 : 465. 1900 ; (6) Ueber Ziichtung neuer Getrei-

derassen mittelst kiinstlicher Kreuzung. Zeits. Landw. Versuchsw. Oesterr. 4 :

1057. 1901 ; (7) Ueber Veredelung und Neuziichtung landwirthschaftlicher

und gartnerische Gewachse. Zeitschr. Naturw. 71: I. 1898; (8) Weitere

Beitnige iiber Verschiedenwerthigkeit der Merkmale bei Kreuzung von Erbsen

und Bohnen. Ber. Deuts. Bot. Gesells. 19: 35. 1901.
Ule, E. (l) Ueber einen experimentell erzeugten Aristolochienbastard. Ber. Deuts.

Bot. Gesells. 17 : 35. 1899 ; (2) Ueber spontan entstandene Bastarde von

Bromeliaceen. Ber. Deuts. Bot. Gesells. 17: 51. 1899.
Veitch & Sons. Manual of orchids, 10 : 45. 1894.
Vestergren, T. Om individbildingen hos slaktet Mentha, samt om hybriden M.

aquatica X arvensis L., etc. Ofvers. Vet. Akad. Forh. 1898I : 33. 1898.

The Spermatogenesis of Hybrid Peas 541

Vilcoq. A. i.liyhridation des HliJs. La Nature, 30 : 250. 1902.

VilmoriD, H. L. de. (1) Cross Breeding. Gard. Chron. III. 14: 332. 1S93 ;
(2) On some hybrid poppies. Jour. Roy. Hort. Hoc. 24 : 203. 1900; (3) Per-
petual strawberries. Jour. Roy. Hort. Soc. 22: 311. 1^991 (4) Selection and
its eflects on cultivated plants. Exp. Sta. Rec. 11 : 3. 1899.

Vries, H. de. (l) Anwcndunjj dcr Mutationslehre auf die Bastardierungsgesetze.
Ber. Deuts. Bot. Gesells. 21 : 45. 1903; (2) Das Spaltungsgesetz der Bastarde.
lier. Deuts. Bot. Gesells. 18: 83. 1900 ; (3) Die Mutationstheorie. Leipzig.
I901 ; (4) Hybridising of monstrosities. Jour. Roy. Hort. Soc. 24: 69. igoo;

(5) Sur la fecondation hybride de I'albumen. Compt. Rend. 129: 973. 1899;

(6) Sur la fecondation hybride de I'endosperme chez le Mais. Rev. Gen. Bot.
12: 129. 1900; (7) Sur la loi de disjoinction des hybrides. Compt. Rend.
130 : 845. 1900 ; (8) Sur les unites des caracteres specifiques et leur application
a I'etude des hybrides. Rev. G6n. Bot. 12: 257. 1900; (9) Ueber erbung-
leiche Kreuzungen. Ber. Deuts. Bot. Gesells. 18 : 435. 1900.

Ward, C. W. Hybridizing and the introduction of new carnations. Am. Gard. 24 :
71. 1903.

Waugh, F. A. Hybrid plums. Bull. Vermont Agric. Exp. Sta. 67 : 3. 1898.

Webber, H. J. ( i ) Complications in Citrus hybridization caused by polyembryony.
Bot. Gaz. 29: 141. 1900; (2) Hybridism in Citrus. Gard. Chron. HL 27:
186. 1900; (3) On a supposed immediate effect of pollen. Science, H. 4:
498. 1896 ; (4) Work of the United States Department of Agriculture on plant
hybridisation. Jour. Roy. Hort. Soc. 24: 128. 1900; (5) Xenia, or the im-
mediate efl'ect of pollen, in maize. Bull. Veg. Phys. & Path. U. S. Dept. Agric.
22 : 7. 1900.

Webber, H. J. & Bessey, E. A. Progress of plant breeding in the United States.
Vearb. Dept. Agric. 1899 : 465. 1900.

Weldon, W. F. R. (i) Mendel's laws of alternative inheritance in peas. Biomet-
rika, i : 228. 1902; {2) On the ambiguity of Mendel's categories. Biomet-
rika, 2: 44. I902 ; (3) Professor de Vries on the origin of species. Biomet-
rika, I : 365. 1902.

Wilson, J. H. (I) Hybridization. Am. Gard. 20: 413. 1899; (2) Observation,
on the fertilisation and hybridisation of some species of Albuca. Bot. Jaarboeks
3: 233. 1891 ; (3) The structure of certain new hybrids. Jour. Roy. Hort. Soc.
24 : 146. 1900.

Witte, E. T. Hybrides des Bromeliacees cultives en Europe. Leide. 1894.

Wittmack, L. On the particular intluence of each parent in hybrids. Jour. Roy.
Hort. .Soc. 24 : 252. 1900.

Wood, J. [Hybrid Newtown pippin apples.] Am. Gard. 22 : 134. 1901.

WoolsoD, G. A. A third New England station for Aiplenium ebenoiUes. Rhodora,
3 : 248. 190I.

Wunderlich, J. Beitrage zur anatomischen Charakteristik der On/ww-Bastarde. Bot.
Centralb. 65 : 352. 1896.

Yasuda, A. On the artificial cross- fertilization between some garden varieties of
Pliarbitis heJeracea L. Bot. Magaz. Tokyo, 11 : (l). 1897,

Explanation of Plates

Note. — A Zeiss microscope with a 2.00 mm. homogeneous oil-immersion objec-
tive and a number 18 eye-piece were u.sed, together with a camera lucida, in making all
of the sketches in plates 17 and 18.

542 Cannon : Studies in Plant Hybrids :

Plate 17

Figs. 1-3. Fillbasket, pure race. Anaphase of last sporogenous division ; pairing
of the chromosomes not well marked. The three sketches were made from one nucleus.

Fig. 4. Fillbasket, pure race. Late anaphase of the last sporogenous division.
The chromosomes are uniting in pairs. One chromosome is in the section next to the
one from which the sketch was made.

Fig. 5. Fillbasket X Debarbieux. Nuclear plate of the last sporogenous division.
Six chromosomes are associated in pairs.

Fig. 6. Fillbasket X Debarbieux. Anaphase of last sporogenous division showing
a pairing of the chromosomes.

Fig. 7. Fillbasket X Debarbieux. Early reconstruction stage in last sporogenous
division, showing the fusing of the pairs of chromosomes at their polar ends ; three
chromosomes are in the adjoining section. The connecting chromatin is exaggerated in

the sketch.

Fig. 8. Fillbasket X Debarbieux. Same division as the preceding ; the connect-
ing ring of chromatin does not become reticulated as early as the remainder of the
chromosomes.

Fig. 9. Fillbasket X Debarbieux. Same division as the preceding sketch ; shows
fusing of chromosomes at the polar ends.

Figs. 10-14. Fillbasket X Debarbieux. Figs. 10-12 are sketches of metaphase
of sporogenous division, />ro^ad/y noi the last one, and figs. 13 and 14 are anaphases
of two similar divisions. These do not show pairing of the chromosomes.

Plate 18

Figs. 15-17. Express X Serpette. Prophase of the first maturation mitosis, het-

erotype rings.

Figs. 18, 19. Fillbasket X Debarbieux. Metaphase of the first maturation

division.

Fig. 20. Fillbasket X Debarbieux. Chromosomes from the metaphase of the first

division.

Fig. 21. Fillbasket X Debarbieux. Nuclear plate, first division, showing the re-
duced number, seven, of chromosomes.

Figs. 22, 23. Fillbasket X Debarbieux. Anaphase of the first maturation mitosis ;
in the latter sketch one of the chromosomes has failed to divide ; this was the only
abnormality which was seen in the sporogeny of the pea hybrids.

Figs. 24, 25. Fillbasket X Debarbieux. All of the chromosomes of one daughter
nucleus are shown ; the number of chromosomes is seven.

Fig. 26. Fillbasket X Debarbieux. Early reconstruction stage of a daughter
nucleus ; one chromosome is in the adjoining section. No splitting of the segments is
to be seen.

Fig. 27. Fillbasket X Debarbieux. Metaphase of the second maturation di-
vision.

Fig. 28. Fillbasket X Debarbieux. Typical chromosomes from the metaphase of
the second division.

Figs. 29, 30. Fillbasket X Debarbieux. Metaphase of the'second division show-
ing the reduced number of chromosomes to be even ; the nuclei are sister nuclei.

Fig. 31. Fillbasket X Debarbieux. Resting nucleus of tetrad; shows complete
resting condition.

The SpEK\fATOGENi:sis of Hybrid Peas o43

Fig.- 32. Fillbasket X Debarbieux. Metaphase of the first nuclear division of tet-
rad ; germination of microspore.

Fig. ;i;i. Fillbasket X Debarbieux. Reconstruction of the vegetative and the
generative nuclei in first mito.sis of tetrad nucleus.

Pl,ATK 19

Fig. 34. Express, pure race. (Reduced.)

Fig. 35. Serpette, pure race. (Reduced.)

Fig. 36. Express X Serpette, showing Express reversion. (Reduced.)

Fic. 37. Express X Serpette, showing Serpette reversion. Both of the figures of
Express X Serpette were drawn from plants selected at random. They are representa-
tive second generation forms. (Reduced.)

Bull. Torrey Ch'b, 30.

I'L.^m I'

/

^s.'

M

*3p

8

t1

a^

9

12 gr^

SPERMATOGENESIS OF HYBRID PEAS.

Bui.L. ToRREY Club, 30.

Pi..^H 18

15

^

') ^

18

24

29

30

•»

20

21

25

>'.

28

26

ill

TTl

SI

32

33

SPERMATOGENESIS OF HYBRID PEAS.

o

o

CO

eT

D

.J
U

>«

Ed

a!
O

J

.J
a
»

The Plant World.

Vol. VI. Plate 32.

0

C3

f/"r

—>

^

f/-i

a.

n>

m

0

f^

"•^

00

ij

^

1^

0

TJ O 3

' O i/>

- J" t^

rt i^. -

• X o

00 .= z

— t^, i^'

c/5 X '^

r- ^

1 .

-u

~^

rt

c/

--_,

0^

Ov

CQ

ZJ

nj

T>

>

-*-•

.

T^

1—

CO

1J

a

^^

[Reprinted from The Plant World, Vol. VI, No. 11, November, 1903.]

Some Aspects of Desert Vegetation.

By D. T, MacDougal.

A SERIOUS investigation of some of the numerous problems presented
by the xerophytic vegetation of arid regions has recently been made
possible by the establishment of the Desert Botanical Laboratory of the
Carnegie Institution, which has been located at Tucson, Arizona. While
any of the more important phases of botany might be the subject of
investigation by means of the facilities offered by this laboratory, yet its
special function consists of an inquiry into the morphology, physiology,
habit and general life-history of the species indigenous to the deserts of
North America, an area which amounts to more than a million square
miles.

The conditions afforded vegetation in these districts show such wide
departures from those of humid temperate, and those of tropical regions ;
the living flora is accessible to so few workers and the entailed investigations
are necessarily so wide in scope, so extensive and difficult in execution,
that the advance of knowledge of the life of desert plants has been com-
paratively slow. Nearly a decade has elapsed since any notable investi-
gations have been made upon xerophytic vegetation as such, and in that
period morphology and physiology have made a marked general advance,
while heredity and the origin of species have taken on a renewed interest
because of the results brought to light within the last few years. It may
be expected therefore that a searching study of the vegetation of arid
regions will afford information concerning the fundamental processes of

250 THE PIvANT WORLD

plants that will materially modify some of the most important general-
izations current in botany at the present time.

A brief sketch of investigations dealing with desert vegetation has
recently been published in the report of the Advisory Board of the
Desert Laboratory , together with a bibliographical appendix w^hich in-
cludes the principal papers bearing upon the subject.*

It may be profitable to point out some of the striking features of the
more prominent types of plants indigenous to arid regions, and to call
attention to phases of their life-history needing investigation, at the
outset of the activity of the Laboratory.

A general . inspection of the southwestern deserts during the active
vegetative season and during the resting period shows that the following
groups of plants may be distinguished upon the basis of seasonal habit
and general composition of the shoot and root. In citing these groups
it is by no means intended that the divisions in question are to be taken
to include the entire flora. The ideas set forth in this crude classification
however may serve as a point of departure for an exhaustive analysis of
the flora upon a more exact basis of habit and habitat :

I. Herbaceous annuals which start into activity from seeds soon after
the beginning of the rainy or favorable season, and quickly develop a
complete flowering shoot. — The formation of flowers and the maturation
of seeds ensues under conditions not widely different from those en-
countered by species living in the same latitude in moister regions. The
roots of such plants are not distinguishable by any special characteristics,
and do not exhibit any unusual capacity for the penetration of the soil,
horizontally or vertically. The greater number of forms of this group
however show a comparatively greater development of the lateral roots
through the upper layers of the soil in a manner that places these organs
in a position to make use of the scant rainfall before it sinks deeply into
the soil. The shoots do not present marked xerophytic adaptations
beyond a hea\'y cuticle, and the usual regulatory devices for controlling
the action of stomata. The total leaf surface of the shoot may be very
great, and the whole plant so delicate that it quickly wilts when uprooted.
The more marked structural features are to be found in the seeds, which
are provided with extremely resistant coats ; for it is the seeds of such
species that must endure the rigors of drought and extremes of temper-
ature of the desert. The seeds of many forms are so well protected by
impervious coatings that they may be soaked in strong acids and other
corrosive solutions for some time without injury to the embryo or the
storage material within.

II. Perennials with woody, bulbous or tuberous stems, which are
chiefly subterranean, or which lie closely on the surface of the soil. —

* Desert Botanical Laboratory of the Carnegie Institution. Origin and Site. F. V. Coville and
D. T. MacDougal, Publication 6, Carnegie Institution. Washington, 1903.

Thl Pi. am Wuwin.

Vol. VI. Plate 33.

ifeWJ.-aCir o

MacDougal on Desert Vei;etatiun. Yikyj t\iL/io^a : youns plant. Tularosa Desert. New Mexico.
B3' permission. Desert Botanical Laboratory. Carneiiie Institution.

THE PLANT WORLD

251

Plants of this group are almost entirely dormant during the season
unfavorable for growth, and produce a rosette of leaves with a shoot of
greater or less extension on which are borne the flowers during the rainy
or favorable season. IJmonuim limbatum Small, which has recently come
under the notice of the author, may serve as an example. This plant
seems to be known only from near the White Sands in the Tularosa
Desert in New Mexico, and during February of this year the numerous
shallow " washes " that extend across the arid plain west of Alamagordo
were seen to bear many thousands of the short stems of these plants
bearing the grayish-brown scales and dead leaves of the previous season
and having every appearance of being lifeless. When specimens of this
kind were brought under suitable cultural conditions and supplied with
water an irregular open cluster of elongated leaves was formed, from one
of the axils of which a widely branching inflorescence arose that reached
a height of 70 cm and bore numerous flowers. The maturation of the seeds
was soon followed by the death of the aerial shoot and basal leaves, the plant
quickly resuming the inert appearance of the previous season. (Fig. 1.)

III. Perennial shrubs or
trees which bear deciduous
leaves. — The leaves are
formed during the rainy
season or under favorable
conditions of temperature
and are discarded during
the periods of most intense
aridity. As a compensa-
tion for the comparatively
brief season during which
the plant has the benefit of
the photosynthetic activity
of the leaves the stems are
are generally rich in chloro-
phyl and carry on a limited
amount of this work during
the greater part of the year.
This is still further supple-
mented by the petioles and
midribs of some species
which are retained while
the leaflets are discarded,
as may be seen in some of

Fig. I. fJmonhim limbalum SmzU. After a photograph of a the LcgUminOSaC. Ipomoea
plant from White Sands, N. M., cultivated in the New , , ,

York Botanical Garden. aroorescetis, the tree morn-

252

THE PLANT WORLD

ing glory, may be included within this group. The trunk of this tree
does not appear to be formed by the successive and progressive activity
of a distinct cambium layer as in ordinary hard-wood trees, but is of a
herbaceous type, and a new cambium is formed every year external to the
bast, which endures only during the season in which it is formed. The
entire trunk is soft, generally rich in water and storage material, while
the external layers including the bark are succulent and rich in chloro-
phyl. The leaves are cast off during the dry season, but flowers are pro-
duced in some profusion by the aid of the water and food material stored
in the trunk. Acacia greggii, and the mesquite {Prosopis) may be
included in this group as well as Parkinsonia. The last-named tree is
the "palo verde" of the Mexicans, and the amount of chlorophyl and
material constructed by its activity present in the branches is so great
that the tree forms a most excellent forage for cattle, horses, and deer,
which also eat the branches of Ipomcea arbor escens. The leaflets of Park-
insonia are often discarded while the petiole remains and carries on the
leaf-functions in diminished measure. (Fig. 2.)

.if/'*"'

^ 1

Fig. 2. Plantlet oi Parkinsonia , some of the leaves of which have discarded the leaflets and retained

the green petioles.

Another interesting variation in the method of casting of the leaves
has been found in the recently discovered Fouquieria macdougalii. The
branches of this tree are very rich in chlorophyl and continue the func-
tions of photosynthesis during almost the entire year. The casting of
the simple leaves is carried out in such manner however that the petiole
is not cut off at its base near the trunk. On the other hand the line of
separation begins on the lower (outer side) of the petiole near the basal

The Flam Wuklu.

Vol. VI, Plate 34.

MacDougal on Desert Vegetation. Echinocadm emoryi- near Torres. Sonora.
By permission, Desert Botanical Laboratory. Carnet;ie lnstitutir>n.

THE PLANT WORLD

253

end of the laminae and cuts slantingly down through the petiole, reaching
the upper (inner) surface of the petiole at its base. The portion of the
petiole remaining is sharp-pointed and remains for a long period, making
a most effective spine. (Fig. 3.)

It is among trees of this group that some
very extensive root-systems have been noted,
the tips of the smaller roots being found
many yards from the base of the trunk.

IV. Perennials of a spinescent habit with
reduced leaves, which do not exhibit any
marked seasonal alterations in activity, but
continue development more or less slowly
throughout the entire year. — The leaf-surfaces
of plants of this character are reduced and
practicall}' non-existent except upon young
plants, the stems and branches being highly

Fig. 3. Leafy hranchof Fouquifrt'a
macdougalii , showing manner
of casting of leaves, and forma-
tion of spines.

chlorophyllose, with many
marked transpiratory adap-
tations of the epidermis.

The cortical tissues of
the stems are sometimes
palisaded after the manner
of leaves . The spines of this group are generally atrophied branches . Not
all plants which share the above characteristics form spines however.

Koeberlinia, Holacantha enioryi, and Ephedra may be offered as
representatives of the group. (Fig. 5.)

V. Perennials with regular foliar organs protected by varnish, wax.

Fig. 4. Ephedra sp. growing in gypsum, White Sands,
New Mexico.

254

'tHE PI.AN'r WORLD

resin, or other water-proof material, or furnished with a volatile oil. — One
of the more common and widespread examples of this large group
is the creosote bush {Covillea trideyitata) , which is a characteristic plant
of extensive desert areas of North America. The endurance of the leaves
varies with the locality and the conditions, but in some instances it is
known to produce two crops of leaves and flowers during the year. In
general it may be said that the reddish-brown resin secreted by the
leaves is most abundant in the drier periods immediately following the
rainy seasons. Covillea is not furnished with spines, but the resin is so
strong in odor and taste that it is but little attacked by animals. (PL 35.)

VI. Perennials with succulent stems or leaves, or with other special
devices for the storage of water. — This group includes a wide diversity of
morphological types, and the species showing greatest capacity for
storage of water are most abundant in the regions in which the rainfall
%, occurs within a comparatively brief period. The total surface

exposed from which transpiration may take place is generally
comparatively small in proportion to the volume of the body
of the plant. The reduction of the shoot and the adaptation
of organs to water storage show the greatest diversity in
different species.

In Cereus greggii the main root and the base of the shoot
appear to be converted into a storage organ containing a
large amount of w^ater, while the shoot shows an extreme
reduction . The tuberous underground portions of this plant
are beet-shaped, a foot or two in length, and may weigh as
much as ten or fifteen pounds. This means that an amount
of w^ater sufficient to meet the ordinary needs of the plant
for a year is held in reserve. (Fig. 5.)

Ibervillea sonorae, the guarequi of the Mexicans, is an
example of a plant that has converted the basal perennial
portion of the stem into a storage organ which holds a
supply of water sufficient to keep the plant alive for years.
The large compressed tubers which lie on the surface of the
soil may attain a diameter of 40 cm across and half that
amount vertically. The woody pulp contains in addition to
water, reserve food-material and some poisonous substance,
the nature of which has not yet been ascertained. It is
%• from Tuf son; evident, however, that it is sufficiently potent to secure the
va''/ld'in the^New tubcrs against the attacks of grazing animals. The tubers
Garden '^"'^"''^^^ ™^y ^^ sccu lying ou the sand under arborescent opuntias
and acacias during the dry season, and quickly send up the characteristic
leafy stems of the Cucurbitaceae upon the approach of rains. Seeds are
produced within two months from the beginning of activity and the

The Plant World.

Vol. VI, Plate 35-

MacDougal on Desert Vegetation. Covillea iiideittata, near Tucson, Arizona. By permission,
Desert Botanical Laboratory, Carnegie Institution.

THE PLANT WORLD 255

climbing stems die back as the rains cease and the plant enters into
another resting period. When in this inert condition it may be thrown
about like so many pieces of wood without injury. Tubers of guarequi,
which were obtained from Torres, Sonora, in February, 1902, have lain
about on the metal benches in the greenhouses in the New York Botanical
Garden, and occasionally start into activit}'. A number placed in the
exhibition cases in the Museum sent out vines in 1902, and in August,
1903, eighteen months after being lifted from the soil, a second
crop of tendril-bearing stems were produced in the museum cases. (See
Plate XVIL Publication 6, Carnegie Institution, Washington, 1903.)

A large number of examples of this general tj'^pe is offered by the
flattened and cylindrical opuntias and cereuses, in which the leav^es are
reduced and quickly fall off, while the much reduced shoot is made of
stems and branches containing a comparatively large proportion of suc-
culent tissue. The amount of water which may be stored in such tissues
may be sufficient for all of the needs of the plant for many months and
perhaps more than one season. It is characteristic of this type that the
water-storage tissue is found throughout the entire shoot, so that when
a single section or "joint" of such a stem is detached it carries with it
a supply that may enable it to propagate the plant under tl;e most arid
conditions. The presence of spines on nearl)' all of the plants of this
type prevents their extermination by animals, which would otherwise
soon destroy them for the water to be obtained from them. During
seasons of extreme drought, ranchers sometimes cut great ninnbers of
opuntias and burn off the spines by means of fires of brushwood and then
feed the denuded branches to cattle and other animals. (PI. 36.)

The large barrel-shaped echinocactuses consist of a great swollen
stem and a root-system penetrating the rocky and sand}'- soil in all direc-
tions. It is impossible to calculate the practical storage capacity of these
plants, but even the most casual inspection would tend to show that a
supply of water equal to the entire transpiration of the plant for years is
kept in reserve. The Indians of southern California, Arizona, and
Mexico habitually make use of this supply when traveling in arid regions
away from pools and streams. (PL 34.)

The yuccas maj' be used to illustrate a different type of plant exhibi-
ting water storage. In many of the species of this family huge rosettes,
or crowns of long, tapering, fleshy leaves are formed, which in some
species have a capacity for the storage of large quantities of water and
food-material. The arrangement of the sharp-pointed leaves, and the
further devices of cutting edges and spines prevent plants of this type
also from being plundered by animals for the sake of the water. (PI. 33.)
Water storage in leaves is also exhibited by Lyciiim fremoyilii and other
species of the genus, the small obovoid leaves of these shrubs being rich in

256 THE PLANT WORLD

water and easily detachable from the plant. In this instance a large
number of animals make use of the leaves as a water supply, eating both
leaves attached to the branches and those which have fallen to the ground.

VII. Species adapted to soils containing large proportions of soluble
salts. — Great areas in deserts have soils from which the soluble salts are
not washed by reason of lack of rainfall, and in certain places around saline
springs and alkaline pools the concentration of the salts is extremely high.
Many of the plants capable of living in soils of this character show the
same structural adaptations as those of the saline districts near the sea-
shore, being truly halcphytic in character. The stems and the leaves of
some species are succulent and are brittle, being easily broken into frag-
ments. A good illustration of the features of halophytes may be obtained
by a comparison of the two eastern species of Tissa, one of which,
T. mariiia, is an inhabitant of sandy salt-saturated soils and the other is
found on arid ridges and on dry hills. Quite a number of forms are
characteristic of the saline and alkaline regions in western deserts, among
which may be mentioned the salt-bushes {.Atriplex) and Allenrol/ea,
while /uncus Cooperi may be cited as the case of a plant which finds
suitable environment only in the briny wet soils around saline springs and
has been collected only in the Mohave and Colorado Deserts, and in
the Death Valley region. (PI. 32.)

The most important features then of the environment encountered by
desert vegetation consist in scanty and unequally distributed water sup-
ply, coupled with high concentration of soil salts in most instances; an
extremely dry atmosphere, with the occurrence of high air and soil-sur-
face temperatures. The actual difference of temperature between the
root and shoot is quite unlike that of plants in moist regions, and as I
have previously pointed out, must be of great influence in all of the vege-
tative processes.^

The general features of desert vegetation offer some most alluring
problems in the study of the origin of species. Here a general interpre-
tation of the face of nature might lead one directly to the conclusion that
xerophytic species are examples of a direct adaptation to environmental
factors and the consequent development of forms adapted to these con-
ditions. This alluring generalization has the academic advantage and dis-
advantage of being most difficult of proof and disproof and may be supported
only on theoretical grounds, for no one has ever actually seen a species arise
in this manner, and for every argument brought forward to support the
theory, the advocate of natural selection will adduce equally cogent, and
quite as theoretical proof, that desert forms arose otherwise. The study
of the extreme types presented by the vegetation of arid regions might

* MacDougal, D. T. Soil Temperatures aud Vegetation. Monthly Weather Review, 31:375. 1903-
Washington.

The Plant World.

Vol. VI, Plate

THE PLANT WORLD 257

well lead to some positive evidence upon the general method of origin of
species, and if the entire effort of the Desert Laboratory inquiry resulted
in bringing to light positive evidence as to the actual origin of two species
of plants its energy would have been well expended by reason of the
value of such results to evolutionary science in general.

The long standing questions as to the causal facts in regard to the
various protective structures of xerophyles have advanced but little
toward their final solution in the last decade. Are the spines, thorns,
prickles and poisons of desert plants really the results of efforts on the
part of the plant for self-protection ? So many of the special features of
xerophytic vegetation have been interpreted in the spirit of an enduring
optimism that leans on the future for a confirmation of its conclusions,
which have been received with patient credulity by the botanical world,
that it is necessary to ask ourselves the most elemental questions about the
best known and most apparent features of the vegetation of the desert.

Not the least interesting of the results to be obtained from studies of
the plants of arid regions, are those which may be expected as to the
physiological and mechanical causes accountable for the remarkably low
densit)' of distribution, one of the most highly characteristic features of
the flora of the desert.

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 47

ANATOMY AND PHYSIOLOGY OF BACCHARIS

GENISTELLOIDES

By ELSIE M. KUPFER

NEW YORK
1903

[Reprinted from Bulletin of the Torkey Botanical Club, 30 : 686-C96. December, 1903]

Anatomy and Physiology of Baccharis genistelloides

I>V I-'.I.SIK M. Ktl'lKK

Baci/taris gciiistcl/ouiis Pcr.s. is an inhabitant of the mountains
of Peru, growing in dry places at an altitude of about 1200 me-
ters. It has several well-marked varieties which apparently spread
north into Ecuador, and south and east into Brazil. The typical
form is a slightl)- shrubby plant of extremely peculiar appearance.
The leaves are atrophied, being represented by minute scales
scarcely more than 2 mm. in length ; and the work of photosyn-
thesis is assumed by conspicuous wings which stand at three
equal angles from each other. The scales are arranged in three
orthostichies, and two wings originate from the base of each .scale.
One of these wings runs through only one internode and termi-
nates at the margin of the scale below ; the other, in each case,
runs through two internodes until it is similarly stopped. This
alternation of a long and a short wing in each orthostichy, as
well as the fact that the wings (following the phyllotaxy) twist
spirally, accounts for the curious appearance of the plant repre-
sented in /". /.

While the shoot is almost uniformly leafless in the upper por-
tion, usually one, two, or even several normal leaves are to be
found near its base. They are of moderate size, obovate and
somewhat thickened (/. 2 and ?). The genus Baccharis is a
large, and as far as leaf-form is concerned, a heterogeneous one.
The leaves of many species, e.g., B.floribunda and B. trincrvis,
are large, thin and smooth ; others, as those of B. alpina and
/). inicrop/iy//a, are minute and thickened ; B. apJiylla and B. gra-
cilis have only narrow phyllodes ; and, finally, B. genistelloides
and B.fastigiata are devoid of foliar leaves. The leaves of sev-
eral of the species, however (and it is significant that these are
generally South American species), particularly B. dioica and
B. cassinaefolia, closely resemble in shape, size and texture the
juvenile leaves of B. genistelloides. There is every reason, then.

685

686

KuPFER : Anatomy and Physiology

it seems to me, for considering this form as that of the ancestral

type.

The seeds, unnamed, were brought from
Peru by Mr. de Lautreppe in 1900 and planted

Fig. 2.

Shoot from base of plant of Bacchaiis genistel-
loides, showing leaves.

in the New York Botanical (jarden. Because
of the ignorance as to their identity, little
attention was given to the plants until they
had attained a considerable size ; so that unfor-
tunately none of the early stages has been ab-

FiG. 3. Single \e2ii oi Bdcc/iaris geiiistel/oiJes, natural size.

served. At the present time, in the beginning
of their third year, the plants are about i meter
high. The stems near the base are secondarily
thickened, and the wings there are either much
thickened or have been entirely lost. Several
of the plants flowered in February, 1902, and
again in December, 1 903 ; but although the
flowers were pollinated, the seeds produced
were exceedingly minute and failed to germinate.

Noteworthy adaptations to an environment both of strong light

Fig. 1. Shoot of j5«(
charts genistelloides.

OI' IVXCCIIAKIS CKM.srKM.OIOES ()87

and excessive dryness come to li^lit on an examination of the
anatoni}' and physiology of the plant.

TllK I'J'I DERMAL SySTEM

The protecti\e system of the plant is very well developed.
There is a considerable covering of wax over the whole plant
(cf. /'. g). In addition, the epidermal cells are cuticularized
both without and within ; and particularly in the younger por-
tions, the surface is covered with peculiar hairs, of which usually
only the end-cells project above the wax. The stomata are exserted
and are apparently sensitive. The wax is laid down in the form
of irregular plates fitted together into a sort of mosaic. In the
younger portions, to about '^-y cm. back of the tip, the waxy
covering is very thick, while the outer layers are shed in the form
of powdery scales. Its very ready solubility in ether indicates its
fatty nature. The ciitifi is present in a layer over the entire plant
even up to the very tip of the growing point, though, naturally, it
has less thickness here than in the older parts of the plant. The
surfaces of the leaves present on the lower parts of the plant are
more heavily cutinized than either the wings or the stem.

The /lairs almost always arise in pairs from two adjacent epider-
mal cells. In most of the members of this family, as Vesque,* whose
observations have been confirmed by more recent investigators,
points out, two kinds of hairs are to be found : glandular hairs and
mechanical hairs. Both kinds are present in this plant. The gland-
hairs, which are composed of double series of from 3-5 cells each,
differ from those mentioned by Vesque and Solereder t in that all
but the end-cells contain chlorophyl. The cutin is but slightly
raised by the not very abundant secretion, the nature of which has
not been any more definitely determined than that it is odorless
and soluble in ether (/. 7). That the secretion of wax is to be
assigned to the epidermal cells in general, and not particularly to
these hairs, seems probable from the fact that few or none of the
hairs are to be found on the older portions of the plant, where,
nevertheless, the wax is abundant.

* Vesque, T. Caract^res des * ■•" * Gamop6tales. Ann. Sci. Nat. Bot. VII. i :
183. 1885.

t Solereder, H. Systematische Anatomic der Dikotyledonen. 51511'. 1899.

088

KuPFER : Anatomy and Physiology

The whip- hairs, too, seem to differ from the types of Vesque
in that they are also, as a rule, bicollateral. They are thus ap-
parently homologous with the gland-hairs. The early condition
of the two kinds of hairs is very similar, the only difference being
that the whip-hairs are from the beginning devoid of chlorophyl.
In the adult condition, however, they are entirely separated from
each other. The enormously enlarged end-cells have walls so
highly thickened that the cell-lumen is almost obliterated (/. j).
The developing whip-hairs arc cutinized almost from the start.
When fully formed they are also uniformly covered with cutin,
but there is a stage in their growth in which the base only of the
end-cells is surrounded by a collar of cutin, while the thickened

Hairs of Baccharis geniUelloides. Fig. 4. (lland-bair. FiG. 5. Two whip-haiis.
Fig. 6, a, b. Developing whip hairs.

cellulose wall of the upper part shows prominently above and
through it (/. 6 a and b). Probably the cell in its period of hy-
pertrophy breaks through its cutin covering, which is reestablished
after growth has ceased. The two whip-hairs often fail to
develop simultaneously, so that a pair is frequently found, one
of which is still in its prime, while the other consists of only
shriveled empty cells. As far as has been determined, new hairs
of either kind are not developed much below 5-6 cm. from the
tip. As some also drop entirely off, the hairs on the older por-
tions are very scattered and frequently consist of only the shriveled
cells.

OF Hacchakis genist?:lloides OSO

It is difficult to determine the function of cither kind of hairs,
if tiiey have anw In the younger parts, where alone the hairs are
plentiful, they are often entirely embedded in the wax. The gland-
cells are usually thickly covered, while occasionally part of the
end-cell of the whip-hair protrudes beyond. Of what use can a
secretion be, the escape of which is entirely prevented? In the
cases among the Compositae in which the whip-hairs are found,
they are usually present as Vesque shows as a tomentum. Prob-
ably here, as in Chrvsoiiia paiuiJJoscnlosa described by Professor
Lloyd, they are functionless.*

The stojuata are evidently of a motile tvpe (/. 7). They are
provided with distinct accessory cells, so that this plant adds
another to the list of compo.sites so pro-
vided, although Benecke f declares these
cases rare. Both upper and lower walls

taper into very finely pointed ridges. No Fif'-- 7- Stoma from

/•!• LL J- ju Stem of Baccharis geni-

trace of a hmge has been discovered by me, „ . , *

=* . . stelloides.

but the cutin seems thinner at the junction

of the guard and accessory cell. The accessory cells raise the
stomata somewhat above the level of the epidermis. They are
more exserted on the stem than on the wing.

The Stem
In the plants examined, the stem varies in thickness from i to 4
mm. In the younger portions, to about 12 cm. from the tip, the
supporting tissue consists largely of sclerenchyma, the wood proper
being confined to a very narrow ring about 25// in diameter.
The masses of bast-fibers are found at intervals around the stem,
but the six largest masses are opposite the three wings and half way
between them respectively (/. 8). These bast-fibers are strongly
lignified, so that they give a much more pronounced reaction with
phloroglucin than the young wood-ring itself. The lignification
of the bast-fibers begins very near the growing tip ; signs of it
can be found at the distance of 1.5 cm. therefrom. This early
acquisition of lignin would seem to indicate a very slow rate of
growth — a point which will be recurred to presently.

* Lloyd, F. E. Anatomy of Chrysoma pauciflosculosa. Bull. Torrey Club, 28:
445. Au 1901.

fBot. Zeit. 50: 570. 1892.

690

KupFER : Anatomy and Physiology

The bast-bundles lie so close to the circle of wood — a feature
according to Solereder characteristic of the Compositae — that no
cambium was distinguishable as such in any portion of the stem of
the growing plants. There are a number of thin- walled irregular
cells between wood and bast, most of which are sieve-tubes ; but
some must be of a meristematic nature, because the wood-cylin-
der gradually increases in diameter so as to be about 60 mm. at a

Fig. 8.

Cross -section of stem o^ Bai'i^iaris i^enis/i'iloides.

distance of 15 cm. from the tip, 125 mm. at 19 cm., and i56 mm.
at the base of the plant examined.

In an older stem taken from a herbarium specimen there were
present three definite wood-rings and a very clearly established
cambium layer. The wood on maceration was found to contain
spiral, pitted and annular vessels. Wood parenchyma is abundant,
but the medullary rays are obliterated. Around the primary bast-
masses when they occur, otherwise external to the sieve tissue, an
endoderm of one layer of rather large clear cells extends. Out-
side of this lies the cortex, of three or more rows of chlorophyl-
bearing cells. Outside of the bast-bundles, however, these cells
lose their chlorophyl and become collenchymatic. The angles of
the older stems from which the wings have been lost become heavily
suberized.

OF Bacciiakis (.KNISTELI.OIOES fj!)]

Al'OGEOTKOIMSM

The stems of /hui/iaris gcnisfi/loidts are markedly apogeo-
tropic. A number of the branches of phints which have been
grown in a moist greenhouse have fallen down from their own
weight. In all cases, however, the tips have turned perpendicularly
upwards, and the plants present a very striking appearance in con-
sequence. They are markedly different from the diageotropic
shoots of /). luiliinifolia, which is growing in the same place. That
this falling down of the branch is also characteristic of the plant
in its natural environment is seen from herbarium specimens.
There are often as many as si.x upright branches from one pros-
trate one.

The Wing

The wings are the conspicuous portion of the plant ; they vary
in width from a few millimeters at the youngest portions to 2 cm. at
the older. The length depends upon their relation to the internodes.
As shown above (/. /) some wings go through only one, others
through two internodes, the longer ones being from 10-18 cm.,
the shorter from 3-8 cm. A transverse section (/. g) shows a
very compact structure. There is no definite pali.sade or spongy
tissue. Several rows of chlorophyl-bearing cells packed more or
less clo.sely extend inward from both surfaces. The air spaces,
which are comparatively few, are chiefly .situated in the middle of
the wing. The stomata are appro.ximately equal in number on
both sides of the wing, averaging 70-80 to the square millimeter.

The wings are notably rigid ; this is due to the occurrence of
-Strands of mechanical tissue which are developed in connection
with the vascular bundles. The conducting tissue plays but an
unimportant role in the support of the wing. There is a particu-
larly large stereome ma.ss on the extreme edge, which is readily
removable as a thick thread.

In order to determine the morphology of the wing, serial sec-
tions of the growing point were prepared. From these it appeared
that the wings are directly continuous with the margins of the
scales. On comparison with shoots bearing true leaves this view
was substantiated. In the terms of many of the text-books, then,
these wings would stand as " decurrent leaves" — as elaborations

(;9^!

Kui'i'KK : Anatomy and Physiology

of the leaf-base, named b}^ Vines the " hypopodium." Their pe-
culiar phyllotactic arrangement would thus obviously be explained.
The question was raised however as to whether these facts alone
/. c, the connection with the leaf-base and their phyllotactic rela-
tion, would be sufficient to establish the homology with leaf-
structures. The point was decided in the negative for the follow-
ng reasons :

I. Their main axes, even at the growing point, are parallel to
the stem, instead of at a greater or less angle as in other developed
leaf-bases, /. c, stipules.

Fk;. 9. Cross-section of wing
of Baccharis genistelloides.

Fig. 10. Cross-section of leaf
of Baccharis gaiistclloides.

2. In many plants, Latliynis latifoiia, L. sylvcstris and L.
grmidiflonis among others, both stipules and wings occur. In
such cases the wings, although just as obviously continuous with
the wings of the petiole, evidently could not be hypopodia.

3. Other organs have been found that follow the leaf arrange-
ment, but may not therefore be supposed to be leaf-structures —
for example the thorns in some species of Riibus and Siiii/ax, the
ridges in the stems of Urtica gracilis^, McntJia crispa and Lcptandra
Virginica.

For these reasons the wings are, it seems to me, rather to be
regarded as lateral-vertical expansions of the stem than as " decur-
rent leaf-margins."

Phototropism

To determine the action of the plant towards light a growing
shoot was fastened in a dark chamber illuminated from one side
only. After fourteen days there was found to be a decided photo-

OF BaCCIIAKIS C.KMSTI-.l.I.OIDES <»{)'•;

tropic reaction. The stem had turned towards the li^ht, the bend-
ing region extending to about 3 cm. from the tip. The two wings
nearest the aperture had turned at right angles to each other and
in the direction of the rays, the third wing had moved to a po.sition
just behind and parallel to one of the others. In this way only its
very edge was illuminated. Although at first it seemed as if the
wings must have played a positive part in the movement towards
the light, probably a purely mechanical explanation is to be as-
signed for their position. By experiment it was found that these
would be the positions assumed by any rigid objects — such as
strips of cardboard — fastened to a bending axis, so that probably
only the .stem was directly concerned in the reaction.

TuF. Leaf
The leaves themselves vary in size. The largest found measured
2.2 X 1.3 cm., the smallest 5x3 mm. (f.j). They appear, as
above pointed out, at the basal portion, but also occasionally on
new shoots, developing from buds on these regions. A transverse
section of the leaf (/. 10) shows that here also we have a fairly
compact mesophyl structure. The intercellular spaces are, as in
the wings, comparatively small. In the leaf, however, sharply
distinguishing it from the wings in structure, we find a noticeable
dor.si-ventral differentiation ; for not only has the leaf a definite
palisade parenchyma of two layers of cells, but a difference is also
to be observed in the number of stomata present on the two sur-
faces. There are 106 stomata to the square millimeter on the
lower surface to 50 on the upper. It is to be noted as significant
that the whole number of stomata per square millimeter would aver-
age almost exactly the same in the case of the wing and the leaf.

Reversion to the Juvenile Leaf-form
Several experiments were started with a view to inducing the
re-formation on adult shoots of the juvenile leaf-form. Some of
the growing shoots were kept in a moist chamber, others in a dry
and still others in a dark chamber for several weeks, but no results
were obtained. This may have been due to one or more of several
causes. First, of course, these particular agents may not induce
leaf-formation in this plant ; or again the failure may have been

<>94

KuPFER : Anatomy and I'uvsioiam.v

due to the fiict tliat growing shoots (and therefore shoots very far
from the usual basal region of leaf-production) were used ; or
finallv, it mav have been due to the fact that all the shoots which
were experimented with formed flower-buds on being released, as
it is believed that when a shoot has once laid down the i)nn1or-
dium of a flower it is thereafter unchangeable.

To avoid the repetition of
these conditions it was deter-
mined to defer the experiments
until such a time as separate
plants would be available for the
different experiments. For this
purpose, but also more particu-
larly to determine the effect upon
the leaf- reversion, a large num-
ber of cuttings were made, con-
sisting, of course, of the youngest
shoots, /. i\, those farthest from
the juvenile character. The cut-
tings were difficult to root ; less
than half of those planted at
three different times succeeded.
At present, however, there are
about 1 5 more or less healthy
young plants. '"

On three of the cuttings which
were planted earliest, reversion
shoots of a remarkable character
developed. The first one to be
affected produced from buds two branches, on each of which ap-
peared three reversion leaves, and on which the wings were re-
duced almost to a minimum. Also (and this seems extraordinar)-)
the main shoot produced near its apex several smaller leaves. This
is the plant represented in /. //.

On the second plant a .similar set of phenomena was ob.servable.
Three shoots bearing some of the largest leaves that have been
found and the greatly reduced wings appeared, and at the same
time the tip of the shoot continued to produce leaves as it grew.

Fig. 1 1 . Cutting of Bacc/iaris genistc/-
loides, showing production of juvenile
aves.

Ol HaCCHAKIS GENISTELLOIUES OI>0

On the third the main shoot withered ; and one of the buds
produced a branch on which the win^s are apparently normal,
but on which three small leaves have been found. Four other
plants produced leaves on branches that retained their wings.

The apogeotropic habit of the secondary branches made anotlicr
experiment possible. A plant on which the main branch had fallen
from its own weight and on which there were a number of such
upright shoots was layered — that is, had its main s«:em covered
with earth. After allowing the branches to rest in this condition
for .some time, the connection of the vertical shoots with one
another and with the main stem was severed. However, although
all the shoots flourished and grew into separate plants, none pro-
duced the leaves.

To sum up briefly the mo.st important points brought out in
this investigation :

J. Baccharis genistelloidcsx'^ 2l plant remarkably well adapted
by the loss of leaves, by the po.sition of the wings, and by the
coverings of the cutin, wax and hairs, to withstand a high degree
of both dryness and insolation.

2. The glandular hairs differ from the hairs previously de-
.scribed in the Compositae in the possession of chlorophyl, and the
whip-hairs in being biserial.

3. The early acquisition of lignin in the bast of the stem, and
of mechanical tissue in the wings, gives to the plant its character-
istic rigidity, and at the same time necessitates a short period of
elongation.

4. The leaves found in any given portion of the stem on the
older parts of the plant resemble closely the leaves of other species
of Baccliaris.

5. The leaf differs from the wing in structure in showing
marked dorsiventrality, which is absent in the latter.

6. While the wings are directly continuous with the margins of
the scales and of the leaves when present, and while they follow
the phyllotaxy, they are to be considered morphologically as
lateral-vertical expansions of the stem and not as " decurren
leaves."

7. The shoots have been shown to be markedly apogeotropic
and positively phototropic.

69^ KuPFER : Baccharis genistelloides

8. Cuttings of growing shoots were made, a number of which
after rooting produced branches which bore reversionary leaves
and greatly reduced wings ; the apices of the old shoots also pro-
duced the leaves in two cases.

The thanks of the author are due to Dr. D. T. MacDouo-al
under whose direction the work was carried on, and to Prof F. E.
Lloyd, for many valuable suggestions.

New York Botanical Garden.

r*TJBLIO.A.TIOIVS

OF

The New York Botanical Garden

Journal of the New York Botanical Garden, monthly, illustrated, con-
taining notes, news and non-technical articles of general interest. Free to all mem-
bers of the Garden. Toothers, lo cents a copy; ^l.oo a year. [Not offered in
exchange.] Vol. I, 1900, viii -f- 213 pp. Vol. II, 1901, viii -(- 204 pp. Vol. Ill,
1902, viii -\- 244 pp.

Bulletin of the New York Botanical Garden, containing the annual reports
of the Director-in-Chief and other official documents, and technical articles embodying
the results of investigations carried out in the Garden. Free to all members of the
Garden; toothers, $3.00 per volume. Vol. I, Nos. 1-5, 449 pp., 3 maps, and 12
plates, 1896-1900. Vol. II, Nos. 6-8, 518 pp., 30 plates, 1901-1903.

Memoirs of the New York Botanical Garden. Price to members of the
Garden, $1.00 per volume. To others, ^2.00. [Not offered in exchange.]

Vol. I. An Annotated Catalogue of the Flora of Montana and the Yellowstone
Park, by Dr. Per Axel Rydberg, assistant curator of the museums. An arrangement
and critical discussion of the Pteridophytes and Phanerogams of the region with
notes from the author's field book and including descriptions of 163 new species.
ix -f- 492 pp. Roy. 8vo, with detailed map.

Vol. II. The Influence of Light and Darkness upon Growth and Development,
by Dr. D. T. MacDougal, first assistant and director of the laboratories. An ac-
count of the author's extensive researches together with a general consideration of
the relation of light to plants. The principal morphological features are illustrated.
xvi -[- 320 pp. Roy. 8vo, with 176 figures.

Contributions from the New York Botanical G-arden. A series of tech-
nical papers written by students or members of the staff, and reprinted from journals
other than above. Price, 25 cents each.

Vol. I. Inclusive of Nos. 1-25, vi -(- 400 pp. 35 figures in the text and 34
plates. I5.00.

CURRENT NUMBERS 25 CENTS EACH.

No. 41. The Polyporaceae of North America. — V. The genera CryptoporuSy
Piptoporus, Scutiger and Porodiscus, by Dr. W. A. Murrill.

No. 42. A revision of the family Fouquieriaceae, by George V. Nash.

No. 43. Some correlations of leaves, by Dr. Daniel Trembly MacDougal.

No. 44. Soil -temperatures and vegetation, by Dr. Daniel Trembly MacDougal.

No. 45. Studies in plant-hybrids, by Dr. William Austin Cannon.

No. 46. Some aspects of desert vegetation, by Dr. Daniel Trembly MacDougal.

All subscriptions and remittances should be sent to

New York Botanical Garden

Bronx Park. New York City

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN — No. 48

MUTATION IN PLANTS

By DANIEL TREMBLY MACDOUGAL

NEW YORK
1903

[Reprinte<l from The American Natcraltst, No. 443; Noyember, 1903]

MUTATION IN PLANTS.'

I). T. MACDOUC'.AI..

Ir is presumably safe to say that all stiulciits ol natural his-
tory a^ec in the opinion that livinj; matter has qualities at the
present time that it did not ori;(inally, or always i>ossess, and
furthermore it is universally conceded that protoplasm is under-
going such development that it is constantly accjuiring new proj>-
erties, and taking form in an increasing number of tyjK's, kinds,
or species of organisms as a consequence. In other words living
matter is increasing the number of its qualities, multiplying the
number of forms in which these qualities are variously gr(»uj)ctl,
and at the same time undergoing such differentiation that an
increasing complexity is the general tendency of the organic
world. These facts once realized the biologist finds himself
confronted with two stupendous interrogatories. Hy what
method is the general development and differentiation of
organisms brought about as exjjressed in the formation or
origin of new species, and secondly what are the general fac-
tors which shape this progression ? The amount of mere dis-
cussion ensuing fn^m the presentation of conflicting views
brought out by these questions, in comparison with the total
scientific effort to obtain positive evidence upon the ix)ints
involved is appalling to contemi)late. Happily the biological
world is becoming intolerant of wrangling and siK-cuIative con-
tentions, and has earnestly set about finding the facts that will
afford an adequate and satisfactory solution to the main i)robIems.
The cult of the study of statistical variations may be regardetl as
one expression of this newly assumed attitude, while the devious,
intricate and oft-times labyrinthine ways of cytological invest i-

iThe general disciis.sion of the mutation theory embodied in this paper,
together with an exhibition of the seedlings of CKnothera wxs given before the
Zoological Seminar of Columbia University, April 23, 1903. The comparisons
between the mutants were not completed until August, 1903.

737

738 THE AMERICAN NATURALIST. [Vol. XXXVII.

gations have, or should have, their chief purpose in the discovery
of the physical mechanism of heredity.

The terms discontimioiis variation, or mutation in connection
with the study of inheritance, descent, and the origin of species
may be taken to mean the autonomous physiological processes
by which one or more individuals of a species give rise to off-
spring which exhibit qualities, or groupings of qualities not pos-
sessed by their immediate ancestors and not previously exhibited
by the individuals comprised in the parent species (progressive
mutation), or by which one or more individuals give rise to indi-
viduals lacking qualities or groupings of qualities exhibited by
the ancestral forms (retrogressive and degressive mutation).
These aberrant individuals or mutants may transmit their char-
acters to their offspring in such a manner as to give rise to a
new line of descent constituting the origin of a new type by
mutation.

The number of freaks, sports, bud-variations, and specimens
of plants with abnormal forms and sizes of leaves, stems, and
flowers, some of them highly teratological, to which attention
has been called by various writers in botanical periodicals under
the designation of mutants makes necessary the emphasis of
the fact that observations on a single individual, or a single gen-
eration of individuals are of but little value in distinguishing
fluctuating variations from mutations. Results worth a moment's
consideration may be obtained only by the most careful exclusion
of the possible effects of disease, of animal or plant parasites,
of hybridization, and by a careful analysis of the phylogenetic
value of the divergences as tested by observations on successive
generations of living forms. It is in this manner, and in Ibis
manner alone, that discontinuous, saltatory variations may be
distinguished from the results of common, fluctuating and indi-
vidual variability. Mutation rests in the main upon such sub-
stantive, discontinuous variations as the acquisition of new
characters, or the loss of old ones hitherto transmitted by the
parent type, or upon simultaneous alterations of both kinds.
These changes may be accompanied by, or may result in, the
masking of current qualities, or the unmasking and energizing
of latent quahties of the parent type.

No. 443] MUTATION L\ PLANTS. 739

The essential diffcremcs between the two processes appeiir to
have been orii^inally set forth by (harks Darwin,' and are treatetl
at lenj^h by deVries. The formal distinctions drawn bydeVries
appear to need some sli<^ht m<Klification and elaboration in order
to make them universally applicable. Thus he holds that con-
tinuous, or fluctuating variability occurs only in accordance with
Ouetelet's laws, and that it involves only the number, size and
weis^ht of organs, and does not include differences in (|ualities.
Cultural experiments of various kinds during the last few years
have given results in which the qualities as well as the numlK-r,
size and structure of organs have been materially altered, but
such induced variations or divergences were not transmissible.
This jjarticular factor in distinguishing between fluctuating antl
mutating variability therefore becomes a safe one, when it is
modified to make mutating variability include <Mily newly
acquired and transmissible qualities. The |)resence of a plant
or an animal parasite may not only change the mechanical fea-
tures of an organ but may also cause most radical alterations
in its physiological properties. A single example of the latter
may be cited in the case of the common species of luiphorbia in
which the affected leaves alter their geotropic sensibility in such
manner that they change from diageotroi^ism to ai)ogeotropism.
Such variations are not transmissible however, and in this lies
the true test between mutation and fluctuating variation A
still further distinction consists in the fact that mutations ensue
in the rudimentary state of the individual, while the alterations
in qualities induced by any of the above factors in fluctuating
variability may be caused in various stages of the development
of the individual, but in a rudimentary stage of the organs
concerned. Mutative alterations arise with the individual, are
not the direct result of external factors, and are jK-rfectly
transmissible, while fluctuating variatit)ns may arise by the influ-
ence of external factors at various stages in the individual
development, and are not transmissible in their entirety.

Much of the confusion inevitable to any discussion of the
subject may be avoided if it is borne in mind that we h:i'>i"""^

lUeVries. .\futationsthcorii, IJd., 2, 1903.

738 THE AMERICAN NATURALIST. [Vol. XXXVII.

gations have, or should have, their chief purpose in the discovery
of the physical mechanism of heredity.

The terms discontimioiis variation, or mutation in connection
with the study of inheritance, descent, and the origin of species
may be taken to mean the autonomous physiological processes
by which one or more individuals of a species give rise to off-
spring which exhibit qualities, or groupings of qualities not pos-
sessed by their immediate ancestors and not previously exhibited
by the individuals comprised in the parent species (progressive
mutation), or by which one or more individuals give rise to indi-
viduals lacking qualities or groupings of qualities exhibited by
the ancestral forms (retrogressive and degressive mutation).
These aberrant individuals or mutants may transmit their char-
acters to their offspring in such a manner as to give rise to a
new line of descent constituting the origin of a new type by
mutation.

The number of freaks, sports, bud-variations, and specimens
of plants with abnormal forms and sizes of leaves, stems, and
flowers, some of them highly teratological, to which attention
has been called by various writers in botanical periodicals under
the designation of mutants makes necessary the emphasis of
the fact that observations on a single individual, or a single gen-
eration of individuals are of but little value in distinguishing
fluctuating variations from mutations. Results worth a moment's
consideration may be obtained only by the most careful exclusion
of the possible effects of disease, of animal or plant parasites,
of hybridization, and by a careful analysis of the phylogenetic
value of the divergences as tested by observations on successive
generations of living forms. It is in this manner, and in this
manner alone, that discontinuous, saltatory variations may be
distinguished from the results of common, fluctuating and indi-
vidual variability. Mutation rests in the main upon such sub-
stantive, discontinuous variations as the acquisition of new
characters, or the loss of old ones hitherto transmitted by the
parent type, or upon simultaneous alterations of both kinds.
These changes may be accompanied by, or may result in, the
masking of current qualities, or the unmasking and energizing
of latent qualities of the parent type.

No. 443-] MUTATION IN PLANTS. 739

The essential differences between the two processes appear to
have been orig^inally set forth by Charles Darwin,' and are treated
at lenjj^h by deVries. The formal distinctions drawn by deVries
appear to need some slight modification and elaboration in order
to make them universally applicable. Thus he holds that con-
tinuous, or fluctuating variability occurs only in accordance with
Ouetelet's laws, and that it involves only the number, size and
weight of organs, and does not include differences in qualities.
Cultural experiments of various kinds during the last few years
have given results in which the qualities as well as the number,
size and structure of organs have been materially altered, but
such induced variations or divergences were not transmissible.
This particular factor in distinguishing between fluctuating and
mutating variability therefore becomes a safe one, when it is
modified to make mutating variability include only newly
acquired and transmissible qualities. The presence of a plant
or an animal parasite may not only change the mechanical fea-
tures of an organ but may also cause most radical alterations
in its physiological properties. A single example of the latter
may be cited in the case of the common species of Euphorbia in
which the affected leaves alter their geotropic sensibility in such
manner that they change from diageotropism to apogeotropism.
Such variations are not transmissible however, and in this lies
the true test between mutation and fluctuating variation. A
still further distinction consists in the fact that mutations ensue
in the rudimentary state of the individual, while the alterations
in qualities induced by any of the above factors in fluctuating
variability may be caused in various stages of the development
of the individual, but in a rudimentary stage of the organs
concerned. Mutative alterations arise with the individual, are
not the direct result of external factors, and are perfectly
transmissible, while fluctuating variations may arise by the influ-
ence of external factors at various stages in the indi\idual
development, and are not transmissible in their entirety.

Much of the confusion inevitable to any discussion of the
subject may be avoided if it is borne in mind that we habitually

1 DeVries. Mutationstheorie., Bd., 2, 1903.

742

THE AMERICAN NATURALIST. [Vol. XXXVII.

within the last three years and it will not be profitable to
rehearse the details at this time. Briefly stated deVries's inves-
tigations may be embodied in the following paragraphs.

1. Observations were chiefly concerned with a large num-
ber of plants growing wild and under cultivation, of the type of
CEjiotlicra lamarckiana. The identity of the parent form was
found by comparison with the original description of the plant
made a century earlier, and by comparison with a type specimen
in the Museum d'Histoire Naturehe in Paris collected in 1788.
The actual name of this plant in the revised nomenclature is a
matter of minor importance in the present connection.

2. raim^-:rs of individuals of the parent type, as a result of
cross- and self-pollination indifferently, constructed seeds which
developed into independent forms, constant and self-maintenant,
which differed in habit, structure, stature, appearance and prop-
erties from the parent type.

3. The aberrant or mutant forms might be divided by
characters as sharp and numerous as most of the so-called
minor species of the systematist.

4. No forms intermediate between the mutants, or between
the mutants and the parent type were found.

5. That the mutant forms were really groups of phylogenetic
value was proven by their behavior when crossed with one another,
with the parent form, and with other species in the same genus.
The hybridization experiments with these forms has yielded
some exact evidence as to the preponderance of phylogenetically
older characters by reason of the fact that the mutants are
forms the exact ages of which are known. Of the crosses of
(E. lata and ffi. nanella with the parent form, from a half to
three-fourths were found to be of the parent type, and the
remainder of the mutant type form. The crossing of mutants
with each other produces a generation many of which show
reversionary characters. The mutation hybrids are constant in
succeeding generations. The separation of antagonistic charac-
ters in the first generations is weighty evidence in support of
the theory of elementary characters, and for the mutation theory.

6. The new types were either constant from the beginning,
or if weak, inconstant or perishing, showed no tendency to revert

No. 443] MUTATION IN PLANTS. 743

to the parent type, and their constancy or fixity might not be
increased by artificial selection.

7. More than one mutant might arise simultaneously from
the parent individuals.

8. Any one of the several mutants observed might originate
from several parent individuals simultaneously.

9. The mutant forms might arise from successive genera-
tions of the parent types.

10. The mutant forms might in turn give rise to new types
after their separation from the parent type.

The above statements rest directly upon observations of care-
fully conducted experimental cultures and admit of but little
argument as to interpretation. With this positive evidence at
hand questions at once arise as to the frequency, occurrence,
prevalence, e.xclusiveness, and as to the mechanism of discon-
tinuous variation as a method of origin of new species. When
we take up these points we at once enter a field of speculation
in which it may be seen there is opportunity for unlimited
argument, and in which with the bias to which most of us are
subject as a result of our training and investigations, it is diffi-
cult to maintain a purely judicial attitude. It will be profitable
to recall some of the more important facts bearing upon these
matters however.

First, as to the occurrence of discontinuous variations in plants
the following examples cited by Korschinsky will be illustrative :
Erjthriua crista gain \\2i^'mXxod\xcQQ[ into cultivation in 1771 and
no aberrant forms were seen until seventy-three years later: Bego-
nia sevipcrflorcns showed deviating forms only after fifty years :
Cyclamen persicnni gave no unusual forms until after one hun-
dred and twenty years of observation : no mutations were
obser\'ed in Ipomoea purpurea in one hundred and twenty years.

De Vries observed many thousands of individuals of a hundred
species growing in the vicinity of Amsterdam in 1886 and 1887
and found mutations in only one, that one Ginothera laniarck-
iana. He points out that remains of plants of various species
found in mummy cases four thousand years old have been found
identical with living species in all recognizable characters. As
a result of a rough examination he also concludes that the

744 THE AMERICAN NATURALIST. [Vol. XXXVIl.

elementary characters of any species of a higher plant may be
reckoned at a few thousand — about 6000 in CEnothera. If
Lord Kelvin's estimate of the period during which life has
existed on the earth is accepted it might be concluded that in a
general way the average interval separating mutable periods of
any plant must be several thousand years, although nothing
in the nature of the question may be taken to indicate anything
like uniformity in the matter. Some writers have put forward
the conclusion that at least ten times the above named period,
or twenty-five hundred million years, would be necessary for the
derivation of the existing forms of plants and animals by natural
selection. It must be admitted that both ideas are valuable
chiefly as attractive examples of imaginative grasp rather than
as affording any real evidence in the matter.

It will be recalled that the various theories which have been
put forward to account for the origin of species have been held
by their authors and advocates to be mutually exclusive, and it
seems to have been, and is still taken for granted by the major-
ity of writers, that all organic forms, both plants and animals,
have arisen in the main by one simple method of biological pro-
cedure. The development of biological science has certainly
reached a stage where this a priori generalization may well be
abandoned. I can not say that a candid review of the mechan-
ism of protoplasm, or of the pertinent evidence, from any point
of view compels adherence to this ancient assumption.

The gi-eat amount of critical study that is being directed to
the study of hybrids and hybridization is widening the horizon of
this subject momentarily, and the result of our recently acquired
information leads us to conclude that species may originate by
crossing. In such instances the new types are due either to new
combinations of unit characters or to reversionary qualities, it
being necessary to keep in mind the fact that by such union of
two types no new characters are brought into existence. It must
be regarded as unsafe moreover to declare any plant a hybrid of
any other given forms unless the process of origin has been care-
fully followed. The fact remains that hybridization is a demon-
strated source of origin of species however, and it is becoming
more and more generally recognized that more than one method

No. 443] MUTATION IN PLANTS. 745

of procedure may have been followed in the development of the
prevalent types of vegetal organism. This view of the subject
has been thoroughly discussed by von Wettstein and need not
occupy our attention further at the present time {Bcric/it, dent.
Bot. GiSilL, Hd. 13, p. 303, 1895).

DeVries concedes that species might originate b)- more than
one metht)d. l)ut he holds that natural selection may account tor
neither the origin, noi' the prcscrxation and continuance of spe-
cies. He furthermore calls attention to the fact that Darwin
repeatedly asserted that characteristics or qualities were formed
very slowly but might disappear suddenly, or in other words that
retrogressive and digressive species formation might ensue by
discontinuous \ariation or mutation. {MntationstJicorie. Bd. 2:
661. 1903.)

It is necessary to point out that the use of the term natural
selection as applying in any sense to the origin of species b)- mu-
tation is wrong in view of the special meaning long attached to
that phrase. Natural selection implies constant and progressive
variation in one or many directions, the individuals distinguished
by the greatest improvements constituting the fittest and surviv-
ing from successive generations. The constant and repeated
survival of the fittest and most improved effecting in time such
an amount of departure from the original as to constitute a new
type. The mutants which arise in discontinuous variability are
seen to depart in all directions from the original, but none of
these may be fitter than the parent type and may perish. It is
prob-^.ble that many thousands of mutants come into existence
for every one that is capable of existence in competition with the
parent type. The repeated failure of the successive series of
mutants can in nowise affect the character of the later crops
of discontinuous derivatives, and hence the failure of the non-fit
and the endurance of the improved form are \\(A dependent upon
natural selection. Every mutant that survives must not only be
suitable for its environment but must be of a structure and habit
that will enable it to compete successfully with existing types,
in comparison with which it is enormously weaker in numerical
strength. It must therefore gain a foothold at once, with but
little opportunity for adaptations of any kind. Every mutant is

746 THE AMERICAN NATURALIST. [Vol. XXXVII.

a possible species and the only selection which might be said to
act is that which determines the type able to live : this selec-
tion has nothing to do with the origin of the surviving form
however.

Thus of the sixteen mutants discovered by DeVries one had
already established itself when found, although seen to arise anew
from the parent type subsequently. Perhaps one or two of the
others might have succeeded in gaining a foothold, but the major-
ity of the new forms must have inevitably perished if subjected
to the ordinary competition of the prevailing meadow species.

As to the cause of mutation, and the mechanism of the proc-
ess but little except of a speculative nature may be offered.
Korschinsky assumes that heredity and variability are opposing
forces or tendencies which are ordinarily balanced. External
agencies such as successive seasons of good nutrition might
allow the tendency to variation to overcome the hereditary
stabiUty and allow the origination of a new form as a result
of the unloosed, superfluous unbalanced energy. He supposes
that whatever the agencies may be that cooperate to bring
about the mutative condition, these forces act upon the develop-
ing embryo in the seed, although he hazards no guess at the
manner in which this might be accomplished {Flora, 89 : 240,
1 901). The above it may be noted is in direct contrast with
the proposal of Darwin that the development of new types is
more rapid when species are competing under adverse condi-
tions, or when the struggle for existence is fiercest.

So far as DeVries's theory of mutation is concerned it may be
said to be the logical outcome of, and to rest upon his hypothesis
of intracellular pangenesis. By this, protoplasm is taken to con-
sist of ideally minute pangens, which make up the living sub-
stance. The pangens and aggregations of pangens are the
bearers of the elementary characters of the species. Altera-
tions in the numerical relations of pangens are made to account
for fluctuating variability. The inactivity of pangens and groups
of these units would cause degressive or retrogressive mutation.
The formation of new characters in progi-essive mutations would
depend upon the development of new pangens, this process con-
stituting premutation. The formation of identical pangens in
separate species would account for parallel mutations.

No. 443] MUTATION IN PLANTS. 747

But little definite evidence is at hand as to the time at which
the changes antecedent to mutation, constituting pre-mutation
occur, although certain stages of development may be designated,
previously to which they must come about. Mutations of the
higher plants are first apparent in the seedling but the actual
alterations or departure from the hereditary behavior must have
taken place at least as far back as the formation of the se.xual
elements the union of which produced the embryo, and may
have occurred even earlier. In any case the mutants are per-
fectly formed in the embr)'o and influence of any kind upon the
germinating seed may not alter their nature (see page 746). It
may be seen from the foregoing that the mutative processes may
be connected with either the vegetative body or the sexual ele-
ments, and may be found within the sporophyte, or be confined
to the gametophyte.

If the pre-mutative alterations occur in the vegetative proto-
plasts of a self-fertilized individual both gametes would presuma-
bly carry the same characters to the union. If, on the other
hand, premutation occur in one of the sexual elements, or if it
occur in the vegetative cells of species which are cross fertilized
only, the embryos formed would be the result of the union of one
mutant gamete and one of the regular inherited form. In a sense
such mutants might be considered as hybrids. This theoretical
aspect of the question seems to find a reflection in the behavior
of Gi. lata, one of the mutants with pistillate flowers only. When
pollinated by the parent form, Gi. lamarckiana, it produces CE.
lata and CE. lamarckiana .

DeVries conjectures that the causes inducing mutation are
partly internal, and partly external to the organism. The state
of external factors necessary to the process probably occur only
at uncertain intervals, and is supposed to embrace a combination
of extremely favorable and unfavorable conditions.

Probably no more profitable subject for research in the whole
realm of natural history offers itself to the investigator than the
problem of the causes which produce new species. The above
supposition deserves early attention from the experimentalist
since it is one that is comparatively easily capable of proof and
disproof.

748

THE AMERICAN NATURALIST. [Vol. XXXV II

0«' 19 l«0>

Fig. I. — Ginotheya lamarckiaita. Rosettes of seedlings two montlis, and five months of age.
Photographed from herbarium sheets. (See Figs. 2 and 3.)

No.

443]

MUTATION IN PLANTS.

749

F.,.. .. a^.nolhcra ru',riuervis. Seedlings two months, and five months of age. Photo
graphed from herbarium sheets. (See Figs, i and 3.)

750 THE AMERICAN NATURALIST. [Vol. XXXVII.

Being desirous of testing the general facts of mutation as
illustrated by the behavior of the Oenotheras under environ-
mental conditions different from those at Amsterdam, seeds of
(E. lamarckiana, CE. rnbiinervis, CE. lata, Gi. nanclla, Gi.
brevistylis, and (E. gigas were procured from Professor DeVries
and these were placed in soil in the propagating houses of the
New York Botanical Garden May 15th, 1902. Germination fol-
lowed in a few days, and a number of individuals ranging from
fifteen to forty of every species were pricked out and suitably
repotted from time to time. The cultures were examined three
to seven times per week except during February, 1902, and July,
1903. The amount of work necessary to make minute and exact
observations on all of the above forms being too great a demand
upon my time, chief attention was devoted to a comparison of
the parent type with riibrinervis and najiella, two mutant forms.

In order to systematize the results general notes were made
continuously upon the habits of the growing plants and formal
comparisons were made at successive stages as follows :

First stage. — July nth, 1902. The plantlets were nearly
two months old and still retained the cotyledons.

Second stage. — October 15th, 1902. A distinct tap root
had been formed and a rosette of leaves had been developed.

Third stage. — June ist-ioth, 1903. Adult rosettes had
been formed, and the smaller leaves which appear around the
base of the stems were apparent. Some flowering stems were
beginning to push up.

Fourth stage. — Augu.st ioth-i5th, 1903. A number of
inflorescences had been produced and flowers were opening daily
in great profusion on some of the forms. Some of the inflores-
cences were enclosed in paper bags in order to secure pure seeds
by means of artificial transfer of pollen.

The more apparent anatomical differences among the forms
examined are shown quite strikingly by the series of photo-
graphs and drawings which illustrate this article.

The main fact to be kept in mind in regard to the parent
form is that it is a recognized and constant species, which has
not undergone noticeable alteration during the long period it
has been under exact observation. The seeds from which the

No. 443.]

MUTATION IN PLANTS.

1ST-

Fu;. 3 — (Enotliera nanella. Rosettes of seedlings two months, and five months of age.
Photographed from herbarium sheets. (See Figs, i and 2.)

752 THE AMERICAN NATURALIST. [Vol. XXXV II.

experimental material was derived were sown in a bed at s' Grave-
land near Amsterdam in 1875 and had been allowed to spread
over an adjoining neglected field until in 1884 an area of 2800
square meters was covered. This material showed the presence
of a form so different from the parent type, when examined by
deVries in 1886, as to lead him to consider it as a new species,
and this mutant, CE. bre%nstylis, which did not arise again during
the observations, maintained itself in the same locality during a
period of twelve years, records of it having been made as late as
1898, and it is still cultivated among the other mutants grown
by deVries and myself. Other forms appeared during the
course of the next fourteen years as has been described in detail.
It was deemed advisable to make independent comparisons of
the plants grown in my own cultures with the type specimen
with which deVries identified his parent form, and to this end
Miss A. M. Vail made a visit to the herbarium of the Museum
d' Histoire Naturelle in Paris, in May, 1903, at my request, and
also later a journey to Amsterdam and inspected the cultures
of OiiwtJicm under Professor deVries's own guidance. Miss
Vail has kindly prepared the following report on the matter :

"The parent form, CEnothera lamarckiana Ser. was found by
deVries to agree in every particular with two specimens in the
Museum d' Histoire Naturelle in Paris. These specimens con-
sist of, first : a plant cultivated in the Paris Garden that had
formed the basis of the original description of CEnothera gmnd-
iflora Lam. It bears a label indicating it as having been
included in the herbarium of Lamarck which was acquired by
the Museum in 1850. On the margin of the sheet in the hand-
writing of Poiret (the author of the section dealing with CEno-
thera in Lamarck's Encyclopedia) is the following inscription
'CEnothera — (grandiflora) — nova spec, flores magni lutei, odore
grato, caulis 3 pedalis.' This specimen is in flower only and
consists merely of the branched upper portion of the shoot with
numerous rather small leaves and conspicuously large typical
flowers. The second specimen comes from the collection of
Abbe Pourret that was contained in the collections of Dr.

1 For a brief general account of the e.xperimental cultures, see MacDougal, The
Original of Species by Mutation. Torreya, Vol 2, pp. 65-6S, 81-84, 97-100. 1902.

No. 443] AfLTTAT/ON IN /'LINTS. 753

Barbier inherited by the Museum in 1847. It is filed in a cover
with 6E". biennis L., and bears that name on the sheet, a small
label inscribed with a series of prelinnean names, and another
with ' Onnj^ra vn/j^aris Spach ' and ^ Ginothcra biennis Linne,'
both apparently in Spach's handwriting. This is the plant
referred to by de Vries as having been collected presumably by
Abbe Pourret in the Paris Garden during his visit in 1788.
The specimen represents an unbranchcd upper portion of a
shoot with numerous large well-developed leaves, partly mature
capsules and several flowers that are somewhat smaller than
those of the previously mentioned specimen. These two speci-
mens differ in no important particular. Tracings of them com-
pared with living i)lants grown in the New York Botanical
Garden from seeds sent by de Vries agree quite perfectly.

"A search through the herbarium of the Museum d'Histoire
Naturelle and that of the New York Botanical Garden does not
bring to light any specimen of a wild North American plant that
can be referred to Gi. laviarckiana as it is now known and cul-
tivated in Europe, nor does it seem to be known to collectors in
North America at the present day.

" Several specimens were found however, which might be con-
jectured as representing a North American plant from which
Gi. lamarekiana might have been derived. One of them is a
plant collected by Michaux now preserved in the Museum at
Paris, and cited by de Vries in the Mutationsteorie (Bd. i : p.
316) and referred by him to a plant frequently cultivated in
Europe under the name of GLnotJiera grandijiora Ait Gi. suave-
olens Desf. but which he considers different from GL. lamarek-
iana. A tracing was also made of this plant which consists of
two specimens fastened on the same sheet upon which numerous
inscriptions bear witness to much diversity of opinion as to its
real identity. A small slip of paper bears in Michaux's hand-
writing ' Ginothera grandijiora,' another (the customary label
of the Michauxian specimens) the inscription ^Gi,notJiera grandi-
jiora Poiret Encycl.,' in the writing of that author of the
section dealing with Ginothera in Lamarck's Encyclopedia ;
beneath that ' GinotJiera siiaveolens Hort. par.' in the writing
of Desfontaines, and lastly ' Onagra vulgaris grandijiora Spach.'

754 THE AMERICAN NATURALIST. [Vol. XXXVII.

in the writing of Spach. The larger of the two specimens con-
sists of a simple entire plant not fully developed, showing root,
leaves, flowers, and capsules, but no basal leaves. The other
specimen, which is smaller, is incomplete and fragmentary. A
comparison of the tracing of the larger specimen with material
in the herbarium of the New York Botanical Garden shows that
it is identical with a specimen under the name of Onagra biennis
grandiflora (Ait) Lindl., collected by E. S. and Mrs. Steel on
Stony Man Mountain, Luray, Virginia, August 15 th, 1901.
The comparison also shows that the wild plant has undergone
no change of any kind during a period of over a century.

" The following memoranda and citations may be of interest
as throwing some light on the history of CE. lamarckiana previ-
ous to 1788.

Linnaeus in his Species Plantarum says that Ginothcm biennis
was brought from Virginia in 16 14 and was then (1753) common
in Europe. In Hortus Cliffortianus (1737) he states on p. 144,
that it is a native of Virginia, having been brought from there
to Europe 120 years before and was at the time he wrote spon-
taneous and plentiful in the fields of Holland. In Hortus
Upsaliensis (p. 94. 1748) he gives the date of its introduction
as 1620, then declared it to be spontaneous in Belgium, Italy,
'Gallia and Germania.' So that from the middle of the 17th
century it was generally in cultivation in the botanical and horti-
cultural establishments of Europe.

Referring to some of the prelinnean writers we find that
Tournefort in Inst. rei. herb., on p. 302 (1700) enumerates nine
species of Onagra, the first four of which only are of interest
here, as follows :
(i.) Otiagra latifolia. LysimacJiia Intea, cornicidata. C. B.

Pin. 245.
(2.) Onagra latifolia, flore dilntiore. LysimaeJiia corniculata
non papposa, Vijginiana, major, flore snlpJinreo. H. L. Bat.
(3.) Onagra latifolia, floribns ampliis. LysimacJiia Virginiana,
altera, foliis latioribus, floribns biteis, majoribns. Cat. Alt-
doif.
(4.) Onagra angustifolia. LysimacJiia angustifolia. Canadensis,
corniculata H. R. Par. LysimacJiia corniculata, lutea. Cana-
densis minor, sen angustifolia Mor. H. R. Bles.

No. 443-] MUTATION IN PLANTS. 755

In the first of these references Caspar Bauhin in Pinax on p.
245 (1671) writes of an American evening primrose under the
name of LysimacJiia Intea cortiicn/ata, as being a Virginian Lysi-
machia growing in the Garden at Padua in 16 19 and adds that
it was a pleasing plant and easy to propagate from seed. The
second reference goes back to Hermann's Catalogus, 1687, where
on p. 396 he records a species of Virginian Lysimacliia with
sulj^hur colored flowers as growing in the Garden at Ley den.
The third reference is to a plant with larger leaves and larger
flowers from the Altdorf Garden. In Jungermann's Catalogus
plantanivi quae in horto Medico Altdorpliino rcpcrinntur we read
that a Lysiviachia Intea Fl. inajoiibiis, adore Tabaci. and a (Lysi-
machia) Vitginiaiia Intea Delphininni qnonindnvi, were known in
the old Bavarian garden at Altdorf in 1635 and the statement is
again repeated in another Catalogns in 1640. It was a suffi-
ciently remarkable plant for Tournefort to note especially in his
Institutiones, and it might be inferred that this large flowered
plant from Altdorf was the ancestor of Ginothera lainarckiana.
It would appear as if a form of what is generally claimed to be
Qinothem biennis L. with delicate sulphureous flowers grew in
the Leyden Garden and another with larger flowers in the gar-
den at Altdorf. Under the same name, LysimacJiia corniculata,
an American evening primrose is said to have been growing in
the Messina Garden in 1640 and it was known in the Paris Gar-
den at about the same time or a little earlier and in 1653 in the
Copenhagen Garden. Morison also records it as occurring in
the Hortus Blesensis in 1669. This last reference is the one
quoted by Tournefort as his fourth species. Again under the
same name of L. coniicnlata Sherard speaks of it on p. 44 of his
Schola Botanica as growing in the Paris Garden in 1689 and,
presumably, descendents of the plants he saw were those col-
lected by Abbe Pourret a century or so after and later made the
type of the much discussed Qi. grandiflora Lem. = Qi. lamarck-
iana Sen. The plant described by Linnaeus in the Species
Plantarum was doubtless a composite species and it would be
particularly interesting in this connection to know just what he
meant by the plant described in the Hortns Cliffortianns as
being plentiful in the fields of Holland. A tracing of the speci-

756

THE AMERICAN NATURALIST. [Vol. XXXVII.

men which could be considered as the type of the plant described
by him in the Hortus Cliffortianus has been kindly furnished by
Dr. A. B. Rendle of the British Museum, and although the flow-
ers are somewhat smaller than those of the living plants of ffi.
lamarckiana as grown in the New York Botanical Garden nurs-
eries, yet the general characters are identical, notably that of
the entire or slightly emarginate petals. This character is cer-
tainly not typical of the wild weed-like CE. biennis of waste lands

Fig. 4. — CEtwthera lamarckiana. Adult rosette immediately preceding development of flow-
ering stem. Photograph of living plant taken from directly above. (See Figs. 5 and 6.)

in North America to-day. In any case it seems extremely doubt-
ful that all these cultivated evening primroses should be referred
to so ungainly and unornamental a plant as CE. biennis.

Prof. deVries in an article on the introduction of OZ. lamarck-
iana in Holland (Ned. Kruidk. Arch. ser. 2, Vol. 6, p. 579,
1895) gives a long and detailed history of the ancestors of the
plants taken into cultivation for his experiments. They were
traced to plants escaped from cultivation and originally raised
from seed received from a seedsman of Erfurt, Germany. Prof.

No. 443.]

MUTATION IN PLANTS.

757

deVries also states that (E. biennis and Ui. umricata are found
in Holland, notably on the dunes.

It seems well established that a large flowered CEnothera was
seen in the Altdorf Garden in 1635, which is probably referable
to none other than lauiarckiana. Later notes of its occurrence
are in existence, but the first definite record of the species was

1

p^'l^^^^^^l

^^^^^H

^^B ^^^^^^^^^^H

Fig. 5. — CEnotlura riibriiiervis. Adult rosette immediately preceding the development of the flowering stem.
Photograph of living plant taken from directly above. (See Kigs. 4 and 6.)

in 1788. It has been found constant since this date, both in
gardens and when running wild : its evolutionary procedure is
therefore none the less valuable as scientific evidence than as if
it were an indigenous wild growing species.

CEnothera lauiarckiana is a species which, so far as present

758 THE AMERICAN NATURALIST. [Vol. XXXVII.

knowledge is concerned, has no exact duplicate in the native
flora of any region, and two probabilities are suggested as to its
origin : It may have been a native of a restricted range in " Vir-
ginia" in which it has been exterminated by agricultural opera-
tions, and hence cannot be found at the present time : or it may
have arisen by some such sudden, and abrupt, discontinous vari-
ation, as that by which deVries' mutants came into existence,
from 65". biennis in the gardens, at Padua, Altdorf or else-
where : at least no intermediate forms are known."

Only eleven specimens of hybrid seedlings derived from
CEnothera lata were brought to the adult stage, in my cultures
and of these but two conformed to the type of Qi. lata, the
remainder being the O. lamarckiana form. CE. lata does not
perfect its stamens but it is capable of being pollinated from the
parent. The offspring followed the laws governing parent and
mutant hybrids, which with deVries were found to consist of
i8% to 20% of the mutant type and the remainder of the par-
ent. My own results agree with this. It is clear that this form
would not have survived beyond the season of its appearance as
it does not display any marked propagative capacity.

QinotJiera nanella originated in deVries' s cultures in 1888
and has since been followed by him through fifteen seasons.
The qualities of this form separate it from the parent in such
manner that it might be considered as a variety by some sys-
tematists, although its behavior and physiological properties are
constant and very clearly distinguishable. In following out
the development of the plant during the eighteen months over
which my own observations extended it became evident that it
differs most widely from the parent in its earlier, and also in its
adult stages, being most like it in the full rosette stage. The
most apparent feature is its diminutive size, both in the young
plant and in the mature flowering shoot. The stem shows but
little capacity for branching and did not reach a height of more
than 20 to 25 cm. in my cultures, or about one fourth that of the
parent, which sends out numerous vigorous branches. The first
few leaves have very broad laminae with irregular apical portions,
and are short petioled. Later leaves are more nearly like the
parent type but remain shorter petioled which has the effect of

No. 443.]

MUTATION IN PLANTS.

759

niakin<r a denser more crowded rosette. The bases of the lam-
ina are almost cordate in some instances, and vary from oblong
ovate to ovate in outline, bein^- sparingly toothed. The plants
established in the soil in the open air did '!()t bloom until about
three weeks later than the parent and (E. rubrincrvis. No
noticeable departure from the characteristics assigned this form
by deVries was found.

Seedlings of Uinothcra ntbrincnns were .seen to ha\c nar-
rower leaves throughout from the earliest stages. The rosettes
were very closely appressed to the soil, and in this stage the
margins of the long petiolate leaves were inrollcd, thus decreas-

FiG. 6. — (Knoihera nanella. Adult rosette immediately preceding the formation of flowering
stem. Photograpli of living plant taken from directly above. (See Figs. 4 and 5.)

ing their apparent width. Attention is to be called here to the
fact that comparisons of leaf forms in plants of this kind
are permissible only between organs on corresponding por-
tions of shoots. The laminae were more bluntly toothed than
those of the parent type, and the midribs occasionally bore a
tinge of red, while the entire shoot including the leaves of the
upper part of the stem showed a tendency to the formation of
anthocyan. The physical qualities of the leaf were strikingly
different from those of the parent, perhaps the most noticeable
feature being the great brittleness of the leaves and stems of

760

THE AMERICAN NATURALIST. [Vol. XXXVII.

young plants, indicative of high turgidity and weak development
of mechanical and supporting tissues. Both of these characters
have been observed by deVries, who notes that the bundles of

Fig. 7. — Oenothera lamarckiana. Adult plant two weeks after beginning of opening of flow-
ers. Photograph of living plant grown in the soil in the open air, and temporarily fixed
in a pot. (See Fig. 8.)

bast fibers of the flowering stems were composed of elements
with thinner walls than those of the parent type.

The leaves of the full rosettes, were silvery white owing to

No. 443.]

MUTATION AV PLANTS.

761

the fact that the hairs on both surfaces were both Vnv^QV and
more numerous than on the parent type. The average length
of the hairs on the upi)er surfaces was 35 as compared to 28
in the parent type, and on the lower surfaces 42 as compared to
30. The average number of stomata on a unit of area of the
upper surface of the leaves of nibrincnns was 37 as compared

Fig. %.—<Enotheta rubrinervis. Adult plant two weeks after beginning of opening of flowers. Photograph of living plant
grown in the soil in the open air, and temporarily fixed in a pot. (See Fig. 7.)

with 34 in Uiiiotlicra lamankiaiia. The brittleness character-
istic of the tissues of nibrincivis may be seen to extend even to
the hairs, since these structures are easily detachable from the
dried specimens, and hence giving rise to the conclusion that
lubrincn'is is less densely pubescent than lamankiaiia as given
in the systematic description below.

762 THE AMERICAN NATURALIST. [Vol. XXXVII.

Still another major difference between the forms in general
habit is that of the method of branching and the growth of the
branches. In Qi. lamarckiana, the branches from the basal por-
tion of the shoot were of a length amounting to more than half
that of the shoot which is also true of GL. riibrinervis. The
upper branches of the former remain short and stout however,
while those of rubrinervis attain greater lengths which decrease
upwardly so that a plant may have a roughly globular outline.

The majority of the features in which the mutant departs
from the parent, as described above, are of a nature that would
equip the^ new form for living under more arid conditions than
the parent, although the actual endurance of rubrinervis to
decreased supply of moisture was not tested. So far as this
single observation goes then, it is to be seen that the new char-
acters of mutants are harmonious in their adaptive relations.

(Enotliera rubriiiervis originated in deVries' cultures in 1899,
and has also appeared by independent mutations since that time.
It has been found to be independent and self-maintenant in
competition with the parent form.

A large number of flower buds in both rnbiincrvis and
lamarckiana were pierced by some insect, and the larvae coming
from the eggs deposited made great destruction, and also caused
the abnormal enlargement of the buds and capsules, which failed
to perfect seeds.

De Vries has continued to find the recurrence of some of the
mutants in the successive crops of seedlings of (Enothera
lamarckiana indicative of the fact that the mutating period of
the parent has not yet been passed. No departures from the
parent type were found among the individuals which have come
into bloom up to this time in the New York Botanical Garden.

The leaves of the seedlings of ffi. lamarckiana are easily dis-
tinguishable from those of lata, nanella, and rubrinervis even in
. the earlier stages, although not so easily separable from some of
the other forms such as brevistylis and leptocarpa according to
deVries. The earliest leaves were ovate, or round-ovate with
rounded apices, or sometimes slightly pointed. These leaves as
well as those formed at the age of five months were distinctly
petiolate but with the laminae relatively narrower. Adult basal

No. 443] • MUTATION IN PLANTS. 763

leaves of the rosette in the period immediately preceding flower-
ing were petiolate with the apices bluntly pointed and with broad
laminiE. TIk- margins of all of the earlier leaves were sparingly
but sharply toothed.

Plants set out early in May were blooming profusely early in
August. The basal branches coming out from the axils in or
near the rosettes were strong and vigorous but the upper branches
of the stem were short and offered a distinct contrast to the
longer, more slender branches of iiibiincnns, with which it was
also contrasted by its denser foliage and larger more showy
flowers. Both stems and branches were thicker and heavier
than in rubrincrvis.

After noting the great variance in behavior and appearance of
the parent and two mutants as described above, mature plants in
bloom, the dried material of the younger plants, and photographs
were submitted to Dr. J. K. Small, who had previously published
an arrangement of the American species, and who is familiar
with them in (Small, J. K. CEnothera and its Segregates. Bull.
Torr. Bot. Club. 23: 167-194, 1896.) the herbarium and in
the field. Dr. Small has kindly prepared the following statement
concerning three forms, which is given in full below :

The characteristics of Qinothcra lamarckiana and (2t. nibri-
nervis as given by Dr. Small are set in parallel columns for con-
venience of comparison :

764

THE AMERICAN NATURALIST. [Vol. XXXVII.

Oenothera lamarckiana Sen

I. Seedling about two months
old. — Leaves sparingly pubescent ;
blades ovate to suborbicular, the
larger about 2 cm. wide, obtuse or
rounded at the apex, each abruptly
narrowed into a petiole.

II. Seedlings 3 inoiiths old. —
Rosettes relatively dense : leaves
copiously fine-pubescent ; blades
typically oblong, the larger ones
fully 3 cm. wide, quite approxi-
mately denticulate, obtuse, or some-
what apiculate at the apex, much
longer than the petioles.

III. Adult plant. ~V\-&x\X very
stout and luxuriant, 0.5 to i m.
tall. Stem markedly channeled,
sparingly hirsute with rather
spreading hairs, nearly simple, or
with several relatively short as-
cending branches near the base,
and few very short ones above :
leaves very numerous, 2-2.5 dm.
long about the base of the stem ;
blades shallowly and often irregu-
larly toothed, those of the lower
cauline leaves broadly spatulate to
oblong, rather acute, each nar-
rowed into a nearly semi-terete
petiole, those of the upper cau-
line leaves oblong to oblong lanceo-
late, acute, or somewhat acuminate,
short-petioled : bracts subcordate
at the base: hypanthium 4.5-5.5
cm. long, about 8 mm. wide at the
mouth, prominently ridged : sepals
4-5 cm. long, longer than the tu-
bular portion of the hypanthium,
the free tips 8-10 mm. long: pet-
als firm 4-5 cm. long, emarginate :
anthers 13-15 mm. long: stigmas
5-6.5 mm. long. (See Figs, i, 4,
7 and g.)

Oenothera rubrinervis deVries.

I . Seedlings about 2 months old.
— Leaves manifestly less pubes-
cent than those of Oe. Lamarck-
iana ; blades elliptic, the larger
ones about 1.5 cm. wide, acute or
acutish at the apex, each gradually
narrowed into a petiole.

I I . Seedlings 5 months old. —
Rosettes lax : leaves less densely
pubescent than in Oe. Lamarck-
iana ; blades spatulate to elliptic-
spatulate or oblong-spatulate, the
larger ones about 2.5. cm. wide,
remotely denticulate, acute, or ab-
ruptly pointed at the apex, about
as long as the petioles or shorter.

III. Adult plant. ~V\?cni rel-
atively stout, less luxuriant than
Oe. Lamarckiana. Stem scarcely
channeled, hirsute, with rather as-
cending hairs, typically branched
throughout, the branches near the
base elongated, decumbent, the up-
per ones gradually shorter : leaves
numerous ; blades less prominently
toothed than in Oe. Lamarckiana.,
those of the lower cauline leaves
spatulate to broadly oblong, obtuse
or acutish, each narrowed into a
relatively long petiole, those of
the upper caviline leaves elliptic-
oblong to oblong or oblong-lanceo-
late, acuminate, short-petioled :
bracts rounded or round-truncate
at the base: hypanthium 5.5 to 6
cm. long, about 4 mm. wide at the
mouth, obscurely ridged : sepals
3.3 to 3.5 cm. long, shorter than
the tubular portion of the hypan-
thium, the free tips 5-6 mm. long :
petals tender, 3-3.5 cm. long,
notched: anthers 6-10 mm. long,
stigmas 7.5-10 mm. long. (See
Figs. 2, 5, 8 and 10.)

No. 443] MUTATION IX PLANTS. 765

Ginotlicra nanc/la was taken by deVries to have a degree of
separation from the parent type that would lead it to be consid-
ered as a variety, a conclusion which is borne out by Dr. Small's
description as given below :

I. Sccdlins^ about two months old. — Resembles that of (Jino-
t/iera linnarckiana ; but the leaf-blades are less uniform, some of
them ovate or oval, others ovate and somewhat lobed near the
apex, others broadly ovate, or prominently apiculate.

II. Seedling § months old. — Nearly like that of Ui. lamarck-
iana\ but leaves inclined to have longer petioles.

III. Adult plant. — Plant, stout and stocky in all [)arts, resem-
bling iK. lamarckiana, but smaller, less than 3 dm. tall.
Stem obscurely channeled, hirsute with somewhat ascending
hairs, simple: leaves approximate, 7-12, 5 cm. long near the
base of the stem ; blades shallowly, often rather remotely, but
quite evenly toothed, those of the lower cauline leaves spatulate
to oblong, acute, or acutish, each narrowed into a semi-terete
petiole, those of the upper cauline leaves broadly oblong to
oblong-ovate, acute or slightly acuminate, nearly sessile : bracts
subcordate at the base : hypanthium 3-3.5 cm. long, about 5
mm. wide at the mouth, obscurely ridged : sepals 3-3.5 cm.
long, longer than the tubular portion of the hypanthium, the
free tips 5-6 mm. long: petals 3.5-4 cm. long, emarginate :
anthers 11-12 mm. long: stigmas 4-5 mm. long.

General Summary.

Discontinuous variation as a possible method of origin of
species was considered by Charles Darwin in his studies of
plants and animals under domestication, and he concluded that
if new forms did arise in this way that they were not self-main-
tenant (1868). On the other hand Galton took the position that
the evolution of species is not necessarily by minute steps (1889),
but Dollo (according to deVries's, Mntationstheorie, Bd. i, p. 46,
1 901) was the first to accept discontinuous variation as the prev-
alent method of origin of species (1893). Bateson (1894)
brought together a large amount of evidence as to types which
have arisen in this manner, and a comprehensive summary of

Fig. 9. — Qinotliera laniarckiatia. A , leaf from basal portion of adult rosette ; B, leaf from
middle, and C, leaf from upper portion of rosette ; D, leaf from middle of flowering
stem; E, bract from lower part of inflorescence; F, flower with petals removed; G,
petal. (See Fig. 10.)

Fig. io. — (Eiwthera rubrinervis. A , leaf from lower part of adult rosette ; B, leaf from
middle portion, and C, leaf from upper portion of rosette ; D, leaf from middle of flow-
ering stem ; E, bract from lower part of inflorescence ; E, flower with petals removed ;
G, petal. (See Fig. 9.)

768 THE AMERICAN NATURALIST. [Vol. XXXVII.

the principal evidence furnished by plants was made by Kors-
chinsky in 1 899. Systematic observations upon the subject were
begun by deVries in 1886 and have been continued until the
present time. As a result of his investigations deVries formu-
lated his "Mutationstheorie," which has appeared in book form,
the separate parts of which have been published in the period of
1 90 1 - 1 903 . This hypothesis rests upon the theory of pangenesis
previously formulated by him.

The parent type, O. lamarckimia, from which deVries saw
mutant forms arise has been found constant in its characters in
cultivation in Europe and America and also when running wild.
This type is not identical with any known member of the Ameri-
can flora, and is most nearly allied to Onagra biennis grandiflora
{Oenothera biennis grandiflora) from which it is suggested it
might have arisen by mutation.

The mutant derivatives of the parent form are found to be
constant in their characters, with no connecting or intergrading
forms, as illustrated by the cultures of the parent, (E. nanella
and (E. rubrinervis, in the New York Botanical Garden during
1 902- 1 903. The mutants are clearly separable from the parent
and from each other both by physiological and taxonomic stand-
ards. Furthermore the specific character of the mutants was
borne out by their behavior when hybridized with one another.

It has become evident from the results so far accomplished
that the testing, study, proof or disproof of the theory of the
origin of species by mutation involves an actual examination of
hnes of descent, and observations upon successive generations
of organisms of known genesis. In this manner only may
mutant forms be distinguished from hybrids, individuals with
aberrant non-transmissible characters and teratological formations.
The nature of the questions involved, and the essentially mate-
rial character of the evidence to be considered is such that all
controversial discussions not supported by facts of this character
must be viewed with distrust. In no instance is this more
plainly apparent than in the recent treatment of the subject by
Vernon (Vernon, H. M. Variation in Animals and Plants. 1903).
This author says "Hence it {CEnothem lamarekiana) is probably
a garden variety of CEnothein biennis (Evening Primrose), and

No. 443] MUr.lHiKX l.\ I'LAMS. ybg

may be a hybrid plant, whilst the mutations obtained by deXries
may be merely partial or complete reversions to the orij^inal
ancestors ot the plant."' It is quite possible, and even probable
th It CE. LunAixkiaw.x may have been ori^inall) derived from the
same type as iE. biennis as noted above, but to designate it as
a "garden variety." and as such ineli^nble as research material
is simple evasion. The plant in question has been under more
or less continuous observation tor a lumdietl and fifteen years
during which period it has been constant in its characters, and
has shown no evidence b\- anatomical similarity or physiological
behavior of being anything but an indej^endent species. With
what species could biinuis hybridize t(» jDroduce lamarckianal
The genus comprises a comparatively small number of types, all
natives of America, and none of which were available as a
hybrid mate to biennis at the time of the origin of lamarckiana.
The conjecture in question is totally unsupported after the most
rigid searcii for evidence upon the matter.

Again to consider the mutants as reversions to the original
ancestors of lamarekiana is impossible, since the mutant forms
exhibit qualities not possessed by any other known members of
the genus, including biennis.

Tin- pninl raised bv Bateson and Saunders (Reports to the
Kvolution Committee. Royal Society. I. p. 153, London.
1902) that the pollen of lamarckiana izon\.2i\x\':-, deformed grains,
which points to its origin by crossing, is without significance.
since the author has found that the stamens t)f plants of biennis
growing in the vicinity of New York exhibit a much larger pro-
portion of deformed pollen than that of the specimens of lamar-
ekiana cultivated in the New York Botanical Garden.

It has been impossible so far to assign mutations to definite
causes, or to forecast the frequency, or occurrence of the phe-
nomenon. 'Ihese phases of the subject constitute the most
important problems of the subject, which await investigation.
Theoretical evidence upon such a subject can have but limited
value, and conclusions of any satisfactory degree of finality may
be expected only from direct experimental research under cir-
cumstances in which the probabilit\ of error is reduced to a
minimum.

770 THE AMERICAN NATURALIST. [Vol. XXXV II.

So far as the origin of mutations is concerned, it seems well
decided that the premutative alterations in seed-plants ensue in
the vegetative and sexual cells previously to the formation of the
embryo in which they first appear, and that no environmental
disturbances may bring about the alterations in question bv
direct action on the seedling.

It is not the purpose of this paper to discuss the various theo-
ries which have been put forward from time t(j time to account
for the origin of species, but to bring under consideration the
facts upon which the conclusions as to the origin of species by
discontinuous variation have been based by deX'ries. These
facts make inevitable the conclusion that new types of specific
rank, taxonomically separable, and physiologically distinct and
constant, without intergrading and connecting forms, have arisen
in GEnothera by discontinuous variation. That mutation is
the principal method of evolutionary procedure is not proven.
That natural selection is universally prevalent is certainly dis-
proven : that natural selection or any other method is capable of
accounting for the existence of any single species has not been
proven with the finality offered by the evidence of discontinuous
variation. It may be said, therefore, that species have actually
been demon.strated to have arisen by mutation, some are known
to have arisen as the result of hybridization, and that evidence
has been accumulated which has been interpreted to demonstrate
the origin of species by natural selection, and by adaptation.
Nothing in the nature of living organisms demands that all spe-
cies should have originated in the same manner, or that one
simple, or single method of procedure should have been followed.

NkU \()RK HoTANKAI. (JARDEN,

August, 1903.

CONTRIIiUTIONS FROM THE NEW YORK BOTANICAL

GARDKN—No. 4<)

THi: POLVPORACEAE OF NORTH AMERICA-VI.
THE GENUS POLYPORUS

By WILLIAM ALPHONSO MURRILL

NEW YOKK
1904

[Keprlated tiom Bulletin of the Torrky Botanical Club, 81 : 29-44. Janoary, 1904]

IFrom the Bulletin op thb Tokubv Botanical Cub, 31 ; 29-44 January, 1904.]

The Polyporaceae of North America — VI. The genus Polyporus

WlLUAM Al.rHONSO MURRILL

The genus Polyporus, as established by Micheli (Nov. PI. Gen.
129. //. 70-ji. 1729), was such a natural division and so clearly
distinguished that it remained intact for over a century. Its nom-
enclatorial type was J\ leptoccphalus (Jacq.) Fr. and associated with
this species were some of the most common and well-known mem-
bers of the family. Unfortunately, however, Linnaeus retained
the name Boletus for all pore-bearing fungi, and those mycologists
who adopted Micheli's genus failed to establish it according to
modern ideas. Adanson, for example, only cited Micheli's figures
and listed no properly named species ; Haller used only poly-
nomials ; and Scopoli in his Introductio listed no species at all
under Polyporus. It was thus left to Paulet (Icon. Champ. />/. /j.
1793) to securely establish the genus. Paulet's work, written
twenty or more years before its publication, contains descriptions
and figures of si.x species of Polyporus ; P. Uhni, P.frondosus, P.
u)idnlicatus, P. carbon arius, P. fasciatus and P. Tuheraster, four of
which belong to Micheli's genus in the strictest sense. The first
species, P. Ulmi, is the very common one well known as P. squa-
mosus (Huds.) Fr. and must be considered the nomenclatorial type
of Polyporus according to principles now in vogue. The general
use of Polyporus instead of Boletus is chiefly due to Fries, who,
without knowledge of Paulet's work, " restored " the name in 1815
and made it popular in spite of the influence of Linnaeus.

In recent systems of classification the original significance of
the term Polyporus has been somewhat perverted. Karsten, for
example, assigned Polyporus to the terrestrial central-stemmed
forms and placed the wood-loving species under the new genus
Polyporellus (Medd. Soc. Faun. et. Fl. Fenn. 5 : 37. 1879). Quelet
adopted new names for both of these groups, Caloporus for the
first and Leucoporus for the second, and erected the monotypic
genus Cerioporus on Polyporus caudicinus (Enchiridion Fungorum,
164-167. 1886). Patouillard followed Quelet in the main, but

29

30 Murrill: Polyporaceae of North America

used Polyporiis in the Karstenian sense instead of Caloporus and
proposed the new name Melanopiis for the group to which Karsten
had assigned the name Polyporellns (Hymenom. Eur. 137. 1887).
One finds, therefore, four modern generic names, Polyporellns,
Leiicoporus, Ccriopoms and Melanopiis, associated either with the
type or with a near ally of the type of the genus Polyporiis and
hence synonymous with it.

The species of this genus are in general very similar in appear-
ance and habit, most of them being small dark-colored plants
attached to fallen branches and other decaying wood on or near
the ground. One species, however, P. caudiciniis, is very large
and does considerable damage to living trees, especially in Europe
where it is abundant. The smallest plant of the genus is P. Acic-
ula, a tropical form only two millimeters in diameter represented
by a single specimen in the Kew herbarium. Many other species
of this group are based upon very scanty collections, some well
preserved, others now in poor condition, and most of them inade-
quately described. The task of the monographer is, therefore,
in this case unusually difficult and his results more or less unsatis-
factory.

Synopsis of the North American species]

1. Stipe pallid or light brown, centrally attached, not darker than the pileus. 2.
Stipe wholly or partly black or fuliginous, variously attached, usually darker than

the pileus. I7-

2. Margin of pileus not ciliate. 3-
Margin of pileus ornamented with cilia, which often disappear with age. I2.

3. Pileus beset near the margin with hydnoid processes. I. P. hydniceps.
Pileus plainly villose, tomentose or scabrous, often becoming glabrous with age. 4.
Pileus minutely tomentose or glabrous from the first. 8.

4. Pileus scabrous, irregular, umbrinous, margin involute ; stipe scabrous, tubes small,

4 to a mm., dissepiments dentate. 2. P. scabriceps.

Pileus villose or tomentose. 5-

5. Pileus becoming virgate from the rupture of the cuticle, tubes 2 to a mm. , decur-

rent, dissepiments dentate. Z- ^- virgatus.

Pileus not becoming virgate. "•

6. Pileus less than 2 cm. in diameter, ochraceous, tubes 2 to a mm., decurrent to the

base of the stipe. 4- P- delicatus.

Pileus m.ore than 2 cm. in diameter. 7-

Murrill: Poi.vporackak of North America 31

7. Tubes decurrent, very short, entire, pileus dark purple, ornamented here and there

with paler radiating lines, surface finely toinentose, becoming glabrous.

5. P. dihaphus.

Tubes not decidedly decurrent, denticulate when mature, pileus yellowish to smoky
black, villose, at length glabrous, spores oblong, curved, 6X2//.

6. P. Polyportts.

8. Sporophore goblet-shaped, pileus less than 3 cm. broad, shallowly depressed at the

center, stipe long, striate, expanding into the pileus. 7. P. Tuba.

Sporophore trumpet-shaped, pileus '^-lo cm. broad, deeply infundibuliforni, stipe

3 cm. long, pallid, pulverulent. 8. /'. cralcrellus.

Sporophore not as above. 9.

9. Pileus minute, 2 mm. in diameter, umbilicate, margin involute, pores alveolar.

9. P. Aciiu/a.

Pileus large, 10 cm. or more in diameter, umbrinous, stipe short, thick, hispid,

pores at length sinuous, dissepiments dentate. 10. P. discoidens.

Pileus of medium size, 2-5 cm. in diameter. 10.

10. Context whitish or brownish in color, extremely thin. II.
Context golden-yellow, not extremely thin, tubes remote from the stipe.

II. /'. phaeoxanthus.

11. Pileus brown, polished, context light brown, tubes decurrent. 12. P. Cohwthiensis.
Pileus white or pallid, more or less translucent, context white, tubes adnate, ex-
ceedingly minute, 8 to a mm. 13. P. ohohis.

12. Tubes alveolar. 1 3.
Tubes not alveolar. 1 5.

13. Margin of pileus finely hispid, broadly sterile below, surface ochraceous, radiate-

striate, stipe brown, pulverulent. 14. P. aemulans.

Margin of pileus strigose, fertile below. 14.

14. Pileus very thin, smooth, pellucid, fragile, stipe thicker below, setulose.

15- P- arculariellus.

Pileus not very thin, fuscous-squamulose to glabrous, stipe equal, grooved, squam-

ulose, grayish-fuscous. 16. P. arcii/arius.

15. Tubes fairly regular, stipe slender, not polished, plants rather delicate, cilia vari-

able in form and persistence. 16.

Tubes very irregular, stipe usually thick and polished, pileus tough, umbilicate,
yellowish-white with brown marginal band, cilia short, fugacious.

17. P. variiporus.

16. Pileus opacjue, not translucent, I-4 cm. in diameter, cilia long, of uncertain dura-

tion, plants mostly cespitose. 1 8. P. Tricholoma.

Pileus very thin, more or less translucent, 1-2 cm. in diameter, cilia short, slender

fugacious, plants not cespitose. 19. P. Cowellii.

17. Pileus squamose, very large, flabelliform, tubes large, alveolar. 20. P. caudicinus.
Pileus finely tomentose, drab- colored, with reddish-brown spots, small, circular,

tubes rounded, minute. 21. P. maculosus.

Pileus glabrous, uniform in color, variable in form, tubes punctiform. 18.

32 Murrill: Polyporaceae of North America

l8. Stipe ivory-black below, pileus usually ochraceous, surface scarcely depressed, mar-
gin even, not becoming extremely thin. 22. P. elegans.
Stipe smoky-black below, pileus usually chestnut-colored, depressed at the center
or behind, margin very thin and irregular. 23. P. fissiis.

1. PoLYPORUS HYDNiCEPS B. & C. Jour. Linii. Soc. Bot. lo :

305. 1868
This is one of Wright's plants collected in Cuba. Berkeley's
description is characteristically brief, but, fortunately, the type
still exists in fairly good condition in the Kew herbarium. It is
readily distinguished from all other species in the genus by the
short cylindrical or subpyramidal hydnoid processes which it
bears on the surface of the pileus near the margin. In shape it
is variable and quite irregular, often breaking into fan-shaped lobes
as it develops and folding inward or outward at the margin as
circumstances determine. The stipe is short, thick, usually
blackish and often reticulate. The fruit bodies occur at times in
clusters with their stipes closely united at the base. In general
appearance this species resembles Sciitiger grisens and its near
allies, but it seems hardly fleshy enough for that group and is
moreover so much like species of Polyporns in habit that I have
retained it in the latter genus as here restricted.

2. PoLYPORUS scABRiCEPS B. & C. Jour. Linn. Soc. Bot. lo :

305. 1868
This species is well named. The type at Kew, collected by
Wright in Cuba on decaying wood, is well preserved and still
shows the characteristic scabrous covering. Other characters are
the dark brown surface, involute margin, decurrent, dentate tubes
and short brown stipe. Like most of the Cuban species, it is rare
and very imperfectly known as regards distribution and variation.

3. PoLYPORUS viRGATUS B. & C. Jour. Linn. Soc. Bot. lo :

304. 1868
Six type plants of this species are at Kew, collected in Cuba
by Wright. They are of the typical Polyporus form and habit
with rather delicate, fragile tubes and thin cuticle, which at length
ruptures in a way quite characteristic of the species. This plant
resembles P. discoideus, but is tougher, firmer and somewhat
smaller in addition to being virgate.

Murrill: Poi.yporaceae of North America 33

4. PoLVPORUS DELicATUS B. & C. Grevillca, i : ij . 1872

Only one small plant of this species is to be found in the Kew
herbarium. It was collected in Alabama by Peters, growing on
decaying wood. It is uniformly ochraccous in color, tomentose,
of soft elastic substance, with a thin undulate revolute margin.
The tubes are angular, 2 to a mm., decurrent even to the base and
quite collapsed when dry. The stipe is central and radicate, and
the buried portion is darker in color than the rest. At first
sight, the surface suggests Polyporus fractipcs, the color being
very similar in both, but the central stipe, firmer substance and
much larger tubes q{ P. delicahis readily distinguish it from that
species.

5. Polyporus dibaphus B. & C. Grevillea, i : 36. 1872
This plant resembles P. Polyporus in many respects, but its
pore structure appears to be different. It was collected by
Peters in Alabama on trunks of Ilex. The type at Kew is the
larger share of a single specimen cut in two. A better de-
veloped plant might show closer resemblance to P. Polyponcs.
One can never be entirely free from the suspicion that species
resting upon a slight material basis and closely resembling species
that are common and variable may possibly be only undeveloped
or depauperate or abnormal forms of the more abundant species.
It seems best, however, in the present instance to consider P.
dibaphus as distinct until its relationships are more clearly proved.

6. Polyporus Polyporus (Retz)

Boletus Polyporus Retz, Vet. Ac. Handl. 253. 1769.

Boletus bruuialis Pers. Neues Mag. Bot. i: 107. 1794.
Batsch, Elench. Fung.//. 10, f. 42a. 1783.

Boletus fasciculatus Schvdid. S'pxc. 154. 1794.

Polyporus bruuialis Fr. Obs. Myc. 2 : 255. 181 8.

Polyporus luridus B. & C. Grevillea, i : n . 1872.

Polyporcllus brunialis Karst. Medd. Soc. P^aun. et. Fl. Fenn. 5 :
37. 1879.

Polyporellus Polyporus Murrill, Jour. Myc. 9: 93. 1903.

There are two forms of this widely distributed plant, both oc-
curring throughout Europe and North America. It was the yel-

34 Murrill: Polyporaceae of North America

low, or vernal, form which Retz described, while Schrader was
dealing with the darker autumnal form of the plant. In his synop-
sis, Persoon separates the two forms as varieties vermis and fas-
ciculatus. P. hiriihts of Berkeley and Curtis is to be referred to
the latter variety.

This species is found on various kinds of decaying wood in
forests and groves, usually upon branches lying on the ground.
Its persistence far into the winter in the fresh state led Persoon to
give it the name by which it is generally known. Among the
specimens examined are the following : Tyrol, Bresadola ; Ber-
lin, Sydoiv ; Finland, Karstcn ; Sweden, Murrill ; Canada, Dear-
ness, Macoim ; Maine, Richer; New Hampshire, Miss Minns;
Connecticut, Wright; New York, Shear, Underwood, Overacker ;
Delaware, Commons ; Ohio, Morgan; Wisconsin, Lapham, Sey-
mour; Iowa, Hokvay ; Montana, Anderson.

7. PoLYPORUS Tuba B. & C. Jour. Linn. Soc. Bot. 10 :

305. 1868
This species is founded upon a single collectiqj^ by Wright in
Cuba, the types being at Kew. It is readily recognized by its
peculiar goblet-shaped form, resembling a long-stemmed PerAza
or a young stage of Cantharelliis cibariiis. Its substance is rela-
tively quite thick, the depression at the center being shallow in-
stead of deep as might be expected from its shape. The margin,
if not straight, is rolled inward instead of outward, and the stipe is
long, slender and undulate.

8. PoLYPORUS craterellus B. & C. Jour. Linn. Soc. Bot.

10 : 305. 1868
The type collection of this species was made by Wright in
Cuba. The name is well chosen and refers to a character by which
the plant is easily distinguished from its near allies. Plants col-
lected in Louisiana by Langlois are even more deeply infundibuli-
form than the types and are also somewhat larger. The species
occurs on decaying wood and appears usually in small clusters.

9. PoLYPORUS Acicula B. & C. Jour. Linn. Soc. Bot. 10 :

304. 1868
This minute species is represented by a single specimen, two
millimeters in diameter, collected on decayed wood in Cuba.

Mukkill: Polvporaceae of North America 35

The stipe is long and slender, centrally attached and somewhat
hairy at the base. The pileus is umbilicate with involute margin
resembling a minute species of OnipJialia in shape. In general
appearance the plant resembles P. TricJiolonia, but the margin is
without cilia and the pores are alveolar.

10. PoLYPORUS discoideus B. & C. Jour. Linn. Soc. lO :

303. 1868
This species was collected by Wright in Cuba. It is rather
larger than most of the members of the genus, but resembles
them closely in habit and structure. Its nearest ally is perhaps
P. virgatns. Being large, it is rather fleshy, the context becoming
soft, corky and elastic when dry. The tubes are rather large, at
length sinuose, and become collapsed on drying, indicating a soft
condition when fresh.

11. Poi.vpoRUS phaeoxanthus B. & Mont. Sylloge Crypt.

154. 1856
This rare species was collected at Columbus, Ohio, by SuUi-
vant. It grew on fallen oak wood. The type at Paris is in frag-
ments, but these are well preserved. The character by which the
species is at once recognized is the deep yellow' color of the con-
text. The pileus is convex, reddish-brown, glabrous, scarcely a
millimeter thick and about two centimeters broad ; the stipe cen-
tral and concolorous, the tubes minute and remote from the stipe.

12. PoLYPORUS Columbiensis Berk. Lond. Jour. Bot. i :

454. 1842

This is one of the thinnest species of the family, resembling a
brown cuticle stripped from some fruit having a smooth, waxy,
polished coat. It is orbicular in shape with a dark central stipe
and small decurrent tubes. The type was sent to Berkeley from
the Columbia River region ot South Carolina. There is also in
the Berkeley herbarium a specimen from Chicalahi, Mexico, bear-
ing the same name, which may be the same species.

13. PoLYPORUS OBOLUS Ell. & Macbr. Bull. Iowa Univ. Lab.
Nat. Hist. 4: 68. 1896

A small plant with very thin partially translucent pileus, brown
central stipe and exceedingly minute pores. Pileus orbicular,

36 Murrill: Polyporaceae of North America

plane, 1.5-2.5 x 0.03-0.08 cm.; surface minutely tomentose, radi-
ate-rugose, isabelline, fulvous at the center ; margin straight or
repand, even, glabrous, entirely devoid of teeth or cilia : context
0.2-0.7 mm. thick, tough, white, translucent, especially near the
margin; tubes o. i mm. long, 8 to a mm., adnate, white, cylin-
drical, regular, edges thick, entire ; spores ovoid, smooth, hyaline,
3.5-4x4.5-5/^: stipe central, tough, elastic, slender, equal,
chestnut-colored, glabrous, smooth, much compressed in drying,
2—4 cm. long, 1—2 mm. thick.

The above description of this species is made from dried plants
now in the herbarium of the New York Botanical Garden, col-
lected by C. L. Smith in Nicaragua.

Type plants kindly furnished me by Macbride agree in all re-
spects with these, being a part of the same collection. The spe-
cies is nearly related to P. TricJwloma, but the pileus is very thin
and translucent, the margin entirely glabrous and the pores
scarcely one eighth of a millimeter in diameter.

14. Polyporus aemulans B. & C. Jour. Linn. Soc.
Bot. 10: 304. 1868.
Very little is known of this species beyond the small type col-
lection from Cuba and Berkeley's rather meager description. The
types are well preserved, however, and show decided characters.
They resemble/', arcidaruis in having alveolar tubes, but these tubes
are broad and shallow and disappear near the margin, leaving a
sterile marginal band one or two millimeters in diameter. In
habit, the species resembles P. Polyporus. The whole plant is thin
and tough, with brown central stipe.

15. Polyporus arculariellus nom. nov.

Favoliis Qertisii Berk. Grevillea, i: 68. 1872.

One specimen only of this plant seems to have been sent to
Kew by Curtis from his North Carolina collections. This is well
preserved, however, and shows the very thin pellucid pileus orna-
mented around the margin with long cilia, the oblong favoloid
tubes and the centrally attached tapering stipe that characterize
the species. It is a near ally of Polyp07'iis aradarins, though
smaller and much more delicate, and also closely resembles such
ciliated forms as P. TriclioloDia and its near allies ; so that its nat-
ural affinities appear to be with Polyporus rather than with Favo-

MUKRII.I.: Por.YPORACEAE OF NORTH AMERICA 37

/us. It is unfortunate that the existence of a Po/yponis Ciirtisii
renders it necessary to change the species name in the transfer
from one genus to the other.

1 6. PoLYPORUS ARCULARius (Batsch) Fr.
Boletus arcularius Batsch, Elench. Fung. 97. 1783. (Mich.

//. 70. f- 5- 1729-)

Boletus exaspcratus Schrdid. S^'\c. 153. 1794.

Poly poms arciitarins Fr. Syst. I : 342. 1821.

This species was described by Batsch as follows : " Stipitatus ;
stipite subgracili, subclavato ; pileo membranaceo convexo, sub-
ulato-fimbriato ; stipiteque concoloribus, spadiceo-rufis ; cellulis
latissimis, rhombeis, aequalibus albis." Fries placed it in the
genus Favolus, which he treated as a subgenus under Polyporus.
Its tubes are certainly favoloid, but its close relationship to P.
Polyporus has kept it near this species rather than with species of
Favolus. Wright's specimen of P. lentus from Connecticut seems
nothing more than P. arcularius. There is at Kevv also the
remains of a specimen from Ohio, which very probably belonged
in the same category.

P. arcularius shows little variation except in size throughout
its wide range. It occurs on decaying wood and shows much the
same habit as P. elegans and P. Polyporus, but differs from these
species in occurring more abundantly in the south. Specimens
have been examined from the following localities : Tyrol, Bresa-
dola ; Connecticut, £^rt;;'/£' ; Pennsylvania, 75?r/'//rt;7/ New Jersey,
Ellis; G^ovgxdi, Raven el ; 0\\\o, James, Morgan ; Missouri, Z^^'w-
ctrio ; Nebraska, Bates; Colorado, Craiulall ; New Mexico,
Earle ; Michigan, Longyear ; Kansas, Bartholoinetv ; Kentucky,
Price; Mississippi, Richer ; Alabama, Earle, Baker; Florida,
Rau, Calkins ; Mexico, C. L. Smith.

1 7. Polyporus variiporus sp. nov.

A small plant resembling P. Tricholoma, but firmer and tougher
with thicker stipe and more irregular pores. Pileus orbicular,
convex to depressed, 1—2.5 X O. i— 0.2 cm., surface glabrous, more
or less radiately striate, somewhat concentrically rugose, straw-
colored to isabelline, fulvous to chestnut-colored around the mar-
gin, which is thin, inflexed, undulate, finely ciliate, the cilia being

38 Murrill: Polyporaceae of North America

fugacious : context 1-1.5 mm. thick, tough, white ; tubes 0.5 mm.
in length, 2—4 mm. in diameter, very irregular, much elongated
radially near the stipe or in marginal folds, not noticeably decur-
rent, yellowish, polygonal to lamelloid, edges firm, entire, becom-
ing denticulate or fimbriate with age ; spores ellipsoid, smooth,
hyaline, 4 x 7 />« : stipe central, hollow, increasing above, lighter
than the pileus, subglabrous with a silky luster, 2 cm. long, 2—6
mm. thick, the buried base enlarged, tomentose and frequently
black.

This species was collected by Earle on his recent trip to Porto
Rico. It grew on sticks buried in sandy soil in woods. The
thick stipe and irregular tubes suggest P. pacliypiis of Montagne,
but it is evidently not that species. It differs from nearly related
ciliated forms in being conspicuously tougher and thicker with
tubes that are in one place small, regular and polygonal and
in another transformed by confluence into long sinuate furrows
resembling the gills of an agaric. This transformation commonly
occurs near the stipe or in pockets made by the partial folding
of the pileus.

18. PoLYPORUS Tricholoma Mont. Ann. Sc. Nat. Bot. II. 8:
365. 1837. PL Cell. Cuba, 249. //. /7,/. /. 1842

This species was originally well described and figured by Mon-
tagne and several specimens of typical plants are still in existence,
so that no doubt exists concerning its identity. In addition to
this, a large number of fresh specimens have recently been col-
lected in Cuba and Jamaica by Earle and Underwood and some
study made of the habits and variations of the species. It is found
to grow in abundance throughout the West Indies and Central
America, varying but slightly in color, but more in size and sur-
face markings, occurring usually on dead sticks in woods, but
sometimes upon logs and even on cocoanut husks. When seen
in the fresh state it is commonly pure white and easily mistaken for
some small agaric, but often with age and always in drying the
color changes to pallid or light yellowish-brown, or even to a pale
reddish-brown. The type plants were, of course, dried specimens
and happened to be darker than is usual with the species.

Another variation still more marked and the cause of consid-
erable confusion is in the size and persistency of the cilia around

MURRII.I. : Poi.VPORACEAE OF NoKTH AmKRICA 39

the mart,nn of the pileus. Sometimes they are long and rigid,
sometimes short and flexible ; they usually disappear at early ma-
turity, but occasionally persist until the plant has passed its prime.
It is doubtful if Montague himself knew of these variations, and it
is certain that Berkeley was much confused by them. In dealing
with the types of this plant it is best to consider the Cuban spec-
imens only, since plants from elsewhere were later included in the
species concept, which are in reality different things. Likewise, in
examining Berkeley's specimens of Pol y poms similis, one must con-
fine himself to the Brazilian plants, since the Cuban ones are only
P. TricJioloma with the cilia gone. In the same way P. flcxipcs,
P. stipitaruis and a few other species have undoubtedly become
confused with P. Tricliolouia in the minds of not a few mycologists
and many collectors.

The following collections are in the herbarium of the New
York Botanical Garden : Jamaica, Earle ^02, ^g2. Underwood
^953 ; Cuba, Undeitvood and Earle 2^6, ^yg, J44, 8ji ; Mexico,
C. L. Smith; Nicaragua, C. L. Smith; Cozumel Island, J////-
spangh 1356.

19. Polyporus Cowellii sp. nov.

A small light-colored plant resembling an agaric, having a
very thin translucent context, minute pores and a slender central
stipe. Pileus orbicular, convex to plane, umbilicate, 1-2 x 0.02
—0.05 cm.; surface nearly glabrous, minutely concentrically ru-
gose, straw-colored to isabelline, becoming darker and hygropha-
nous around the margin or in blotches or even over the entire
surface, often faintly radiate-striate about the center with delicate
white or brown lines ; margin very thin, straight or inflexed,
somewhat irregular in outline, finely denticulate, the teeth pro-
longed into short fugacious cilia : context extremely thin, mem-
branous, 0.1-0.2 mm. thick, white or pallid, partially or entirely
translucent; tubes 0.2-0.4 mm. long, 3-6 to a mm., larger by
confluence, adnate, pallid, polygonal, edges very thin, subcntire,
becoming fimbriate with age ; .spores ellipsoid, smooth, hyaline,
binucleate, 2.5 x 5//: stipe central, solid, very slender, equal,
concolorous or slightly darker, pruinose to glabrous, smooth,
longitudinally striate, 2-3 cm. long, 0.3-1 mm. thick, soft and
milk-white at the center.

This species is described from dried plants collected in the
island of St. Kitts by Britton and Cowell during September and

40 MURRILL : POLYPORACEAE OF NoRTH AMERICA

October, 1901. It was found in considerable quantity growing
on decaying wood. There is also in the herbarium of the New
York Botanical Garden another good collection of this species
from Porto Rico made by Mr. and Mrs. A. A. Heller in March,
1899. It also grew on dead wood. This latter collection, how-
ever, is not considered typical.

20. PoLYPORUS CAUDiciNUS (Scop.) Murrill

Boletus caudicinus Scop. Fl. Carn. ed. 2, 2 : 469. 1772.

Boletus Ji(gla7tdis Sc\\2i&^. Fung. 3: 75.//. 1 01-102. 1774-

Boletus sqiiamosiis Huds. Fl. Angl. 614. 1778.

Polyporus ^7//// Paul. Icon. Champ.//, /j. 1793.

Polyporus squauiosus Fr. Syst. i : 343. 1821.

Polyporus caudicinus Murrill, Jour. Myc. 9: 89. 1903.

Exsicc. : France, Roumeguere 2'/o6, 2J0J, J4-0J ; Germany,
Krieger 860, Sydow 212, Magnus, Murrill ; Sweden, Murrill;
England, Murrill ; Canada, Dearness ; Connecticut, Underwood.

This species is the largest of the genus and occurs in very con-
spicuous clusters on the trunks of injured deciduous trees in
Europe and America. In London and in some of the cities of
Germany I have found it especially abundant on elms, maples,
horse chestnuts and other shade trees. It is at present compara-
tively rare in America and need not here be seriously considered
from an economic standpoint for some years to come.

Scopoli's description is not so clear and definite as might be
desired, but in the first variety described there can be no question
as to the plant he had in mind, for he speaks of it as multiple,
lobed, fleshy, large, growing on the trunks of trees, variegated
above with darker spots, light-colored beneath, with large polyg-
onal pores. The large pores connect it with the genus Hexa-
gona, but its general structure and habit are those of Polyporus.
Young plants collected ni Connecticut, by Underwood, in May,
showed very short favoloid tubes with thin toothed dissepiments,
a milk-white fleshy-tough context and a scaly pileus differing
little in color from that of the mature plant. The odor of the
fresh plant is strong and somewhat mealy ; the base of the stipe
is clothed with short dark-brown or black velvety tomentum. It
is quite possible that in this species we have modifications pro-

MUKRII.L : POLVFORACEAE OF NoRTH AMERICA 41

duced by an abundance of nutritious food such as is not supplied
to the common species growing on dead wood. Large size, very
rapid growth, ample pores and a somewhat softer context are
probably connected with a better food supply and a more vigorous
mycelium.

In addition to Polyporus, the present species has served as the
nomenclatorial type of two recent genera, /. r., Ccriopoi'us Quel.
(Ench. Fung. 167. 1886) and McUmopus Pat. (Hymenom. PLur.
137. 1887).

2 I . Polyporus maculosus sp. nov.

A small tough plant nearly related to P. fissus, but minutely
tomentose, with rounded nearly glabrous spots of a reddish-brown
color. Pileus irregularly orbicular, deeply depressed, 3-3.5 X O. i-
0.2 cm., surface finely and densely tomentose, radiately striate,
drab-colored, ornamented with several light-bay or chestnut spots
mostly situated about midway between the center and the margin,
which is very thin, striate, irregular and somewhat fissured but
entirely devoid of teeth or cilia: context 0.5-1.5 mm. thick,
tough to corky, pallid ; tubes 0.2-0.4 mm. long, 6-7 to a mm.,
decurrent, pallid or yellowish, subcylindrical, edges entire ; spores
ovoid, hyaline, smooth, 3 x 4//, immature in these specimens:
stipe central, solid, woody, tapering upward, chestnut-black at the
base with blotches of light-yellow, uneven, subglabrous, 2 cm.
long, 2-4 mm. thick.

This plant was collected in Central America by C. L. Smith.
The hymenium is not quite mature, but the other characters are
well marked. The peculiar reddish spots seem to be caused by
the disappearance of the thick tomentum at certain points. The
species occurs on wood and has the habit of representative plants
of this genus.

22. POLVPORUS ELEGANS (Bull.) Fr.

Boletus elegans Bull. Herb. France,//. ^6. 1780.

Boletus iminDudarius Bull. Herb. P"" ranee, //. 12^. 1782.

Polyponts elegans Fr. Epicr. 440. 1 836-1 838. Pat. Tab.
Fung. 137. 1 883-1 886.

Abundant throughout Europe and North America on decay-
ing branches and trunks of various trees. Being exceedingly vari-
able in form, it is not strange that in Europe it is confused with
Its variable near relative, Polyporus Calceolus {^Boletus Cakeoliis

42 MURRILL : POLYPORACEAE OF NORTH AMERICA

Bull.) \Polyporus 7.'aj'ius(Fers.) Fr.], and some European mycolo-
gists go so far as to say that P. clegans and P. Calceolus are the
same thing ; but field studies in Sweden have convinced me that
we have nothing in America to correspond to B. Calceolus of
Bulliard, whose description and figure of this species may be easily
matched with European specimens from either field or laboratory,
while even our most closely allied Canadian forms of P. elcgans
fail to connect with the European relative.

The variety first described as B. nummulariiis by Bulliard is
quite distinct from the typical form when seen in the herbarium,
but when both are picked from the same branch in the woods the
differences speedily become of minor importance. In the follow-
ing partial list of specimens examined, no distinction is made be-
tween the two : Tyrol, Bresadola ; Finland, Karsteii ; Canada,
Macoiin, Dearncss ; Vermont, Burt ; Maine, Richer, Miss White ;
Connecticut, Earle ; New York, Underwood, Cooke, Lobeiistine ;
New Jersey, Ellis ; Ohio, Morgan ; West Virginia, Nnttall ; Michi-
gan, Ward ; Iowa, Hokvay ; Colorado, Bethel, Crandall ; Wash-
ington, Piper, Parker.

23. PoLYPORUS Fissus Berk. Lond. Jour. Bot. 6: 318. 1847
Poly poms trachypus B. & Mont. Syll. Crypt. 154. 1856.

Both of the above names were assigned to plants sent from
Ohio. Those sent by Lea were small and undeveloped and one
or two were accidentally lobed. The tubes were so small as to
be overlooked and he labelled them Thelephora. The collection
made by SuUivant at Columbus was evidently described by Mon-
tagne. The description is accurate and quite complete. The
reason why neither of these names has come into common use is
due to the fact that our plant was at first miscalled P. picipes, a
name given by Fries without good reason to plants of P. calceolus
having particularly black stems. This latter species, so far as we
are aware, does not occur in America, but American specimens at
Kew are labeled P. picipes Fr. In the herbarium of Fries at
Upsala there is only one specimen of " jP. picipes" from North
America and that is determined by Ellis. For our common
species, then, which has been so generally known as P. picipes,
we must make use of Berkeley's name.

MUKKII.I. : POI-YPOKACEAE OF NORTH AMERICA 43

The plant grows upon dead wood and has the liabit of P.
clegans, but is larger and darker in color. It often jiersists, how-
ever, until bleached nearly white. Exsiccati have been studied
from Maine, Ilarvcy, Miss White; Connecticut, Umicriuood ;
Massachusetts, Seymour ; New York, Clinton, Ovcrackcr ; Penn-
sylvania, Everhart ; Vermont, Far/oza ; Kentucky, Price ; Wash-
ington, Parker; Michigan, Wood, Miss Minns.

Species IsguiRENDAE

Polyporus aviygdalinus B. & Rav. Grcvillea, l: 49. 1872.
This species is said to differ from P. caudicinus in having smaller
tubes, but it is probably only a form of that species in an unde-
veloped stage. I have not been able to find a type specimen.

Polyporus cyathifornus Lev. Ann. Sci. Nat. Hot. III. 2 : 181.
1844. The type of this species was probably burned with most
of Leveille's types during the occupation of Paris by the Germans.
The description corresponds closely with P. cratercllus, but it is
difficult to determine that the two species are identical.

Polyporus pachypus Mont. PI. Cell. Cuba, 421. 1842. This
species is described as caespitose, with thick excentric stipe and
membranaceous tubes, which are small and rounded near the
margin and large and favoloid near the stipe. No types have
been found in foreign herbaria and Montagne himself said in his
Sylloge that the species needed further investigation.

Polyporus stipitarius B. & C. Jour. Linn. Soc. Bot. 10 : 304.
1868. The original description just cited does not materially as-
sist one in interpreting the type plants in the Berkeley herbarium.
Most of these types so closely resemble P. Tricholoma that a new
description of them seems superfluous ; and the one card of speci-
mens which appears to be somewhat different from the rest was
listed under P. Tricholoma by Berkeley at the time that P. stipitarius
was described. Judging from the Kew collections, P. stipitarius
appears to differ from P. Tricholoma chiefly in possessing a longer
stipe, yet the description calls for a stipe shorter than that of /-*.
Tricholoma. New material may possibly throw light on this prob-
lem, but I seriously doubt if Berkeley's species can ever be en-
tirely disentangled from the earlier one of Montagne.

Polyporus gracilis Kl. Ann. Nat. Hist. 3: 384. 1839. This

44 MURRILL : POLYPORACEAE OF NORTH AMERICA

is a caespitose species with slender pruinose stipe, small pellucid
pileus, ciliate margin and very minute polygonal or sinuose tubes.
Saccardo considers it the same as P. fiexipes Fr. and P. liapalus
Berk., but these two species do not, so far as I know, occur in the
West Indies or even in Central America.

Polyponis Hiimplireyi P. Henn. Hedwigia, 37 : 280. 1 898.
From the description of this species it appears to resemble P. Tri-
choloma, but has adnate instead of decurrent tubes with whitish
fimbriatulate edges. This and the next species are placed here
because I have not yet examined the type specimens.

Melanopits marasmioides Pat. Bull. Soc. Myc. France, 18 : 173.
1902. This species is found in Guadeloupe and Martinique and
is related to Favolus inelanopus. It is deeply umbilicate, ochra-
ceous, with translucent denticulate margin, which is sterile below,
large shallow radiating tubes and a black central stipe. Undevel-
oped plants resemble certain species of Xylaria. An examination
of the plant is necessary to determine whether or not it properly
belongs with Polyponis.

New York City.

X^UBLICATIOIVS

OF

The New York Botanical Garden

Journal of the New York Botanical Garden, monthly, illustrated, con-
laining notes, news and non-technical articles of general interest. Free to all mem-
bers of the Garden. Toothers, lo cents a copy; $i.oo a year. [Not offered in
exchange.] Vol. I, 1900, viii -f- 213 pp. Vol. II, 1901, viii + 204 pp. Vol. Ill,
1902, viii + 244 pp. Vol. IV, 1903, viii -j- 238 pp.

Bulletin of the New York Botanical Garden, containing the annual reports
of the Director-in- Chief and other official documents, and technical articles embodying
the results of investigations carried out in the Garden. Free to all members of the
Garden; to others, $3.00 per volume. Vol. I, Nos. 1-5, 449 pp., 3 maps, and 12
plates, 1896-1900. Vol. II, Nos. 6-8, 518 pp., 30 plates, 1901-1903. Vol. Ill,
No. 9, 174 pp., 15 plates, 1903.

Memoirs of the New York Botanical Garden. Price to members of the
Garden, $1.00 per volume. To others, $2.00. [Not offered in exchange.]

Vol. I. An Annotated Catalogue of the Flora of Montana and the Yellowstone
Park, by Dr. Per Axel Rydberg, assistant curator of the museums. An arrangement
and critical discussion of the Pteridophytes and Phanerogams of the region with
notes from the author's field book and including descriptions of 163 new species.
ix -f- 492 pp. Roy. 8vo, with detailed map.

Vol. II. The Influence of Light and Darkness upon Growth and Development,
by Dr. D. T. MacDougal, first assistant and director of the laboratories. An ac-
count of the author's extensive researches together with a general consideration of
the relation of light to plants. The principal morphological features are illustrated.
xvi -j- 320 pp. Roy. 8vo, with 176 figures.

Contributions from the New York Botanical Garden. A series of tech-
nical papers written by students or members of the staff, and reprinted from journals
other than above. Price, 25 cents each.

Vol. I. Inclusive of Nos. 1-25, vi -f- 400 pp. 35 figures in the text and 34
plates. ;S!S.oo.

RECENT NUMBERS 25 CENTS EACH.

No. 43. Some correlations of leaves, by Dr. Daniel Trembly MacDougal.
No. 44. Soil-temperatures and vegetation, by Dr. Daniel Trembly MacDougal.
No. 45. Studies in plant-hybrids, by Dr. William Austin Cannon.
No. 46. Some aspects of desert vegetation, by Dr. Daniel Trembly MacDougal.
No. 47. Anatomy and physiology of Baccharis genistelloides, by Miss Elsie M.
Kupfer.

No. 48. Mutations in plants, by Dr. D. T. MacDougal.

All subscriptions and remittances should be sent to

New York Botanical Garden

Bronx Park, New York City

CONTRIBUTIONS FROM THE NEW YORK BOTANICAL

GARDEN— No. 50

THE SPINES OF FOUQUIERIA

By WINIFRED J. ROBINSON

NEW YORK

1904

[Keprinted from Bulletin of the Torrey Botanical Club, 31 : 45-50. January, 1904]

[From the Bdllktin op thb Torkby B<itanical Club, 31 : 45-30. January, 1904]

The spines of Fouquieria *

Winifred J. RoniNsoN

The six species now comprised in the genus Fouquieria occupy
collectively a region extending from the vicinity of the City of
Mexico northward into California, Arizona, New Mexico and
Texas. Nearly all of the species find their habitat in arid districts,
and the localities from which they have been reported may be
comprised within the southern extensions of the Chihuahuan and
Sonoran deserts. Among other adaptations, the various species
exhibit a capacity for casting off the leaves during seasons of
drought, more or less irrespective of the time of the year, while
some interesting features of the formation and casting away of the
bark have been noticed.

All of the species bear strong spines and seem to be included
under the single colloquial name "ocotillo" by the Mexicans,
who use the living plants for hedges and sometimes erect barriers
by thrusting the ends of heavy branches into the ground in rows.
Messrs. Coville and MacDougal report having seen shrubs of
Fouquieria splendens planted within a few inches of young shade
trees in the streets of Alamogordo, New Mexico. The spreading
spiny branches were held in a cylindrical clump around the trunk
of the tree by means of wire and formed a most effective pro-
tection against damage by animals.

During the course of some recent work upon this genus, Mr,
G. V. Nash called my attention to the unusual manner in which the
spines were formed, which he described as " developed within the
petioles of the leaves on the new growth, becoming apparent when
these fall." t The general anatomical facts presented seemed of
sufficient interest to warrant a detailed examination, and the results
of my observations upon Fouquieria splendeus and F. Macdougalii,
specimens of which are growing in the conservatories of the New
York Botanical Garden, are presented below.

* The results described in the following paper were obtained by the aid of a Re-
search Scholarship in the New York Botanical Ciarden, in 1903.

fNash, G. V. A revision of the family Fouquieriaceae. Bull. Torrey Club, 30 :
449. 1903.

45

46

Robinson : Spines of Fououieria

On the lower side of the petiole of the primary leaves, a woody
thickening is developed (Figs, ii and 12, zc) which increases in
firmness as the leaf matures. This arises from the cortex, which

even in the bud is proliferated to form
a thickened area on the outer side of the
base of each leaf (Fig. i, c). As the
leaf emerges from the bud a conical
portion of the cortex on the lower side
of the petiole begins to lose its paren-
chymatous character. The cells lengthen,
their ends become pointed (Fig. 2, s) and
their walls thicken ; these changes are
,, . , , , J accompanied by a diminution in their

Fig. I. Section through bud. '^ •'

a, young leaf ; ^, cortex, thick- lumina (FiG. 3, s). With phloroglucin
ened to form spine ; v, vascular and hydrochloric acid they take the
^^^^^^- violet color characteristic of sclerenchy-

matous cells. The conversion of the parenchyma into scleren-
chyma is incomplete in F. splendcns, and a slender cone shaped
mass of thin-walled cells extends
about two thirds of the distance

from the base toward the tip

of the newly formed spine. This

is illustrated by the diagrams of s

longitudinal and transverse sec- j.^^ ^ p^^^;^^ ^^ longitudinal section

tions shown in Figs. 4, 6 and 7, through petiole, b, cortex ; t, separating
/. About the time when the layer ; 5, sclerenchyma.

leaf-blade is cast off, the parenchyma-core disintegrates and a hol-
low spine remains. In Foiiqiiieria
Macdoiigalii, on the other hand,
almost the entire mass of the cor-
tex on the outer (lower) side of
the petiole is transformed into
sclerenchyma, making a solid
heavy spine (Figs. 5 and 8, s).
The growth of the leaf and
Fig. 3. Portion of transverse section the development of the spine is
roug pe 10 e. accompanied by the differentiation

of a separatory layer of thin-walled, elongated cells (Figs. 2 and

ROUINSON : SlMNKS OI^ FoUQriF.RIA

47

3. t) in the cortex, between the sclerenchymatous tract and the
fibrovascular tissue of the petiole. The sclerenchymatous tissue
sustains only mechanical relations to the leaf, and the lamina may
be held for extended periods after the completion of the separatory
layer. When the vegetative season reaches its end by reason of
drought or low temperature, the leaf-blade is ca.st off; and as its
fall occurs simultaneously with its drying out, this desiccation may
be taken as the direct cause of the splitting of the separatory
layer above mentioned. The separation is identical with that
which occurs in the fall of the leaves of many deciduous trees
in the autumn in general procedure, but the author is not acquainted
with any other instance in which the petiole is cut in a longitudinal
plane making necessary a separatory layer two or three centimeters

Fig. 4. Diagram of longitudinal section through petiole of Fouquieria splendens.
p, parenchyma ; s, sclerenchyma ; v, vascular tissue.

Fk;. 5. Diagram of longitudinal section through petiole oi Fouquieria Mac-

dougalii.

Fig. 6. Diagram of transverse section through petiole of Fouquieria splendens near

the apex of spine.

Fig. 7. Diagram of transverse section through petiole of Fouquieria splendens,

near base.

Fig. 8. Diagram of transverse section through petiole of Fouquieria Macdougalii

near base.

Fig. 9. Diagram of transverse section through petiole of Fouquieria Macdout:^alii
near the apex of spine.

in extent (Fig. 13, d). Again, this unique method of excision re-
sults in leaving an elongated pointed portion of the petiole 3 or 4 cm.
in length attached to a mass of thickened epidermis and cork on
the stem where it is retained rigidly for an indefinite period. The
spine and the contiguous portion of the outer tissues of the stem
may easily be pulled from the underlying tissues in Fouquieria
Macdougalii (Fig. 10, /') when they are young, but in F. splendens
they are less yielding.

48

Robinson : Spines of Fouquieria

In Idria, another genus of the same family, a similar spine,
originating in the petiole of the leaf, has been described by Poisson
(Bull. Mus. Hist. Nat. i : 278. 1895) and in some species of the
nearly related Cantua, of the Polcnioniaceae, there is a woody ridge
at the base of each node. The latter, however, is more like the
structure which protects the axillary bud in Syrhiga and Philadel-
phus than the spines of the Fouquieriaceac.

The majority of the buds in the axils of the primary leaves do
not elongate to form branches, but develop small clusters of spat-
ulate leaves which are almost sessile and do not form spines
(Figs. 1 1 and 12, <r).

Engelmann (Bot. Gaz. 8: 338. 1883) was the first to call
attention to this peculiar mode of dehiscence of the leaves of Fou-

FlG. 10. Fouquieyia Maaiongalii, shQvimg a, manner of dehiscence of leaf ; k,
manner in which spine and bark may be peeled from stem.

Fig. 1 1 . Young shoot of Fouquieria Macdougalii. c, cluster of secondary leaves ;
7V, thickening which will form spine.

Fig. 12. Young shoot of Fouquieria splendens.

Fig. 13. Spine of Fouquieria Macdougalii. d, surface from which upper portion
of petiole has separated.

quieria. Plank (Garden & Forest, 9: 73. 1896) says: "The
bases of the spines appear to arise in the cellular part, leaving the
wood full of holes in their decay," which statement was doubtless
made because of the peculiar growth of the cork between the
spines and the portions of the outer tissues, which adhere to
them, making the spines, with their adhering pieces of bark, dry,
papery structures, which still have close enough contact with the
adjacent cells to leave a scar when they drop off. Van Tieghem

Robinson : Spines of Fouquieria 49

fjour. de Hot. 13: 295. 1899) has described the leaves of the
long branches as having sclerenchymatous cells prolonged from
the sclerenchymatous layer beneath the epidermis of the stem, on
the under side where the epidermis is in contact with the cortex.
He considers this subepidermis as the outer part arising from the
division of the cortex into two well-differentiated layers, and crit-
icises Baillon (Hist. PI. 9 : 242. 1888) for referring to the spines
as leaves reduced to a midrib. Such a section as that shown in
diagram in Fig. 9, where there are several layers of parenchyma-
tous cells between the sclerenchyma (s) and the epidermis, shows
that the sclerenchymatous cells arise not from a distinct outer layer
of the cortex, but within the cortex. Like the prickles of the
blackberry and rose, the spines of Foiiquieria have no vascular tis-
sue, but unlike them, the sclerenchymatous tissue arises from the
cortex, not from the cuticle and the abscission layer is nearly at
right angles with the axis of the stem instead of being parallel with
it and in the surface of the stem.

Various questions naturally arise in the study of such a struc-
ture as the spines of the Fmiquia-iaccae. How did these spines
orieinate ? Was the stimulus external or internal which caused
the original variation ? Is Fouquieria a genus that was separated
from its parent form at a remote period, so that the intermediate
forms between it and the other Poleinoniales have become extinct,
or is it a relatively new genus ? Is it a stable form, or do the
different plants vary in wide range ?

Spinose processes are so characteristic of desert flora that
the inference was long ago drawn that there is a close association
between their structure and environment. The usual corollaries
of spine development, reduction in surface and elaborate adapta-
tions for water-holding, as well as the protective character of the
spines themselves in warding off the attacks of animals, are such
important factors in the maintenance of a genus in arid regions,
that it is easy to reason that they are the outcome of influences in
tho.se localities external to the plant.

Kerner, Henslow, and others of the older botanists have been
inclined to emphasize the principle that " specific forms, on the
whole, fit the places they have to live in," rather than their inherent
tendency to variation. No experiments have been performed upon

50 Robinson : Spines of Fouquieria

Fouqnieria to test the effect of changed climatic conditions upon it,
but it has been observed that in greenhouses the leaves remain
attached to the plant a number of months, while in nature they are
usually cast off at the end of a few weeks ; and in some years, in
the wild state, no leaves are produced. Lothelier's experiments
upon the barberry (Rev. Gen. Bot. 2 : 276. 1890) showed that
when grown in moist air it loses its spines. Henslow (Jour. Linn.
Soc. Bot. 30: 223. 1895) has made a similar observation for Ononis
spinosa, one variety of which living upon sandy shores is covered
with spines, but becomes less and less spinose in favorable condi-
tions or under cultivation, and he cites the cultivated apple and
pear as similar examples. Miss Dale (Ann. Bot. 15 : 59> 497-
1 901) has noted that when tubers of Dioscorca sent out shoots in
light and without moisture, the leaves were scarcely developed at
all, and the same thing may now be observed in the Museum of
the New York Botanical Garden, where tubers of Dioscorea in the
exhibition cases have sent out branches, the leaves of which are
greatly reduced.

There is a wide gap between Fouqnieria and Cantiia, its near-
est relative among the Polemoniales, which fact together with the
small number of species in the genus, its confinement to a limited
area, and its stability may be taken to indicate that it is an old
form, though there is no geological record so far as is known of
any similar spine-bearing form.

JPUBLIO^^TIOIVS

OF

The New York Botanical Garden

Journal of the Ne^w York Botanical Garden, monthly, illustrated, con-
laining notes, news and non-technical articles of general interest. Free to all mem-
bers of the Garden. Toothers, lo cents a copy; ^l.oo a year. [Not offered in
exchange.] Vol. I, 1900, viii -}- 213 pp. Vol. II, 1901, viii -|- 204 pp. Vol. Ill,
1902, viii -|- 244 pp. Vol. IV, 1903, viii -|- 238 pp.

Bulletin of the Ne^w York Botanical Garden, containing the annual reports
of the Director-in-Chief and other official documents, and technical articles embodying
the results of investigations carried out in the Garden. Free to all members of the
Garden; toothers, ^3.00 per volume. Vol. I, Nos. 1-5, 449 pp., 3 maps, and 12
plates, 1896-1900. Vol. II, Nos. 6-8, 518 pp., 30 plates, 1901-1903. Vol. Ill,
No. 9, 174 pp., 15 plates, 1903.

Memoirs of the New York Botanical Garden. Price to members of the
Garden, ^l.oo per volume. To others, $2.00. [Not offered in exchange.]

Vol. I. An Annotated Catalogue of the Flora of Montana and the Yellowstone
Park, by Dr. Per Axel Rydberg, assistant curator of the museums. An arrangement
and critical discussion of the Pteridophytes and Phanerogams of the region with
notes from the author's field book and including descriptions of 163 new species.
ix -{- 492 pp. Roy. 8vo, with detailed map.

Vol. II. The Influence of Light and Darkness upon Growth and Development,
by Dr. D. T. MacDougal, first assistant and director of the laboratories. An ac-
count of the author's extensive researches together with a general consideration of
the relation of light to plants. The principal morphological features are illustrated.
xvi -(- 320 pp. Roy. 8vo, with 176 figures.

Contributions from the New York Botanical Garden. A series of tech-
nical papers written by students or members of the staff, and reprinted firom journals
other than above. Price, 25 cents each.

Vol. I. Inclusive of Nos. 1-25, vi + 400 pp. 35 figures in the text and 34
plates. ;5S5.oo.

RECENT NUMBERS 25 CENTS EACH.

No. 43. Some correlations of leaves, by Dr. Daniel Trembly MacDougal.

No. 44. Soil-temperatures and vegetation, by Dr. Daniel Trembly MacDougal.

No. 45. Studies in plant-hybrids, by Dr. William Austin Cannon.

No. 46. Some aspects of desert vegetation, by Dr. Daniel Trembly MacDougal.

No. 47. Anatomy and physiology of Baccharis genistelloides, by Miss Elsie M.
Kupfer.

No. 48. Mutations in plants, by Dr. D. T. MacDdugal.

All subscriptions and remittances should be sent to

New York botanical garden

Bronx Park. New York City

x<.

Now York Botanical Garden Library

3 5185 00293 3289

■-■"'-'■*-«■■.*.,

.. J r

^

ti

■•fc:.;> ■