CONTRIBUTIONS FROM THE GRAY HERBARIUM
HARVARD UNIVERSITY

Studies in the Boraginaceae
Pah pecan tir la coop... ES
A New Specs of Joncutom Clombia.

Lee ST) wr
RAP AS A Te

~- VIL. » I. M. Jounston. :

_ I. STUDIES IN THE BORAGINACEAE,—VII.
By Ivan M. Jounston
1. Tae Sourn AMERICAN Species oF HELIOTROPIUM.

In the present paper a provisional treatment is given of the
American species of Heliotropium known to occur south of Panama.
In addition to keys I have provided complete synonymy and —
citation of specimens. Attention has been given also to

to reserve the discussion of this matter for the present, planning
to give it detailed consideration after certain other studies in Helio-
tropium and Tournefortia. For convenience, therefore, the genus
has been accepted in its broadest sense and Heliotropium europaeum
L. has been taken as its genotype.

The last critical account of the South American species was in the

general revision of the family published by DeCandolle Pro-
dromus, ix. 531-559, in 1845. The few subsequent works have been

to sg eaincaige ii. 193-202 ory in the ppsibes) year. The Chilean

zenfamilien, iv. Abt. 3a, $2-97 (1899), nothing further, tga
general nature has t sul

_ the paper are the results of field- and A ir ‘made in
= a At thet time I examined the material in

4 JOHNSTON

the herbarium of R. A. Philippi in the National Museum and the

barium of the Pedagogical Institute. From Prof. Lorenzo R.
Parodi, Dr. Alberto Castillanos, Mr. F. C. Hoehne, and Prof. Miguel
Lillo I have received valued material for identification. The sendings
of Dr. Lillo were very extensive and have been of great assistance in
interpreting the species of Argentina. From Dr. Carl Christensen
of the Botanical Museum, Copenhagen, I received photographs of

written. Appreciating all the help I have received I would here
express my hearty thanks to all those who have assisted me.

Every specimen that has been studied during the preparation of
this paper has been cited. To the best of my abilities, with the
facilities available, the material has been cited under the country and
major subdivision of the country from which it came. Following each
collection, in parenthesis, are the initials indicating the herbarium
in which the collection examined may be found These initials and the
herbarium they indicate are as follows—“BD” for Botanical Museum

in Berlin-Dahlem, “BM” for British Museum of Natural History,
“RM” for Field Museum of Natural History, “G” for Gray Herbari-
um, “IP” for Pedagogical Institute of Santiago, ‘““K’’ for Royal Botanic

Garden at Kew, “MS” for National Museum of Santiago, “NY”
for New York Botanical Garden, and and “US” for United States
Herbarium.

National

caper a tell
Sor and -tube villous within... ..- VIL. § HELIOPHYTUM, 21,

and
Carpels very strongly ribbed; coarse weedy annuals with

C: if t all ribbed; t
SS a 2g agg

STUDIES IN THE BORAGINACEAE 5

a Bianroun Witte oc VIII. § Cocuranga, 25.
rolla vill WH. oe ee VII. § Hetiorxrom, 21.
Carpels — breaking in two along a definite dorso-ventral
medio-longitudinal commisure, at maturi gg each falling
away to form two uniovulate cane (ef. H. indicum L.
and H. a enero Soe

ciliate on the back Anais wag ee ee oe IX. § Heriornamnvs, 38.
t an ly 0.5 m
nutlets smooth, pubescent or glabrous; anthers glabrous
or pubescent or dular at tip only.
thi e commonly on a

mo ,
perennial, herbaceous or shrubby, usually erect.
§ OrtHostacuys, 46,

Stigmatie disk very broad and relatively thin, saucer-
shaped, closely oe ee plant prostrate or very loosel;

son a gee bs or rarely slightly fruticulose.
Fruit merely 4-suleate, the nutlets with plane contiguous
sides; plant succulent, glabrous, ——

Vig HaLwrnornra, 14,
Fruit deeply 4lobed, the nutlets saat and =
ventrally united; plant ee pubescent
wees ee V. § PLAGIOMERIS, 16.
_I. Section Coeloma (DC.) Johnston. bute a as § Coeloma
DC. Prodr. ix. 556 (1845); type-species, ingle m monostachyum
DC. Valentina Speg. [FI. Patag. ii. sso a —y Argentina
lii. 78 (1902); type-species, V. : Valentiniella
Speg. Anal. Mus. Nac. Buenos ok nag 3 9 Te. type-species
V. patagonica Speg
To this section Melons five American species and a few from Africa.
It is the most primitive group of American H. Its most
striking feature is the presence of malpighiaceous hairs. "Ses pcnaghe
the occurrence of this type of trichome has not been previously noted
as occurring in the Boraginaceae. These hairs are known to be devel-
Se a ee

included in the section, since whens de | malpighi :
(egg elhganae egal asap pil op paycgmcan, ay
section. Also lacking malpi

: g
which show affinities in the section Coeloma. LT should not be sur-

6 JOHNSTON

eventually find their way into the section. These species show many
characters of Tournefortia.

Most of the American species of Heliotropium can be derived from
the section Coeloma. One line from Coeloma has given rise to sections

stachys i is the only section whose relation to Coeloma is uncertain

Key To Species.

Plant fruticose, 5-20 dm. tall; leaves 7-25 em. long; ——,
in oo ae malpighiaceous Clee ee ow Coe Sue ey 1 tiaridioides.
Plant herbaceous, widely spreading or prostrate, less than 1 dm.
tall; hasan 1-5 em. long.
Plant t perfectly glabrous, somewhat glaucous; leaves a's
St WON ee. H. patagonicum.
Plant more or less obviously pubescent; leaves an oe al
ous, veins evident.
Plant with malpighiaceous hairs at least on the stems or in
the inflorescence; corolla-lobes linear or subulate-
hiat tekiig ‘sesipighineoss ‘hairs; corllaiobes oe
t us coro ovate-
Obie 6. GOS mm. long: : 2... 2.2.2: ee ee: 4. H. Schreiteri.

1. Heliotropium tiaridioides Cham. Linnaea iv. 453, t. 5, fig. 3
(1829); Chodat, Bull. Herb. Boiss. ser. 2, v. 483 Lense Molfino,
Physis vii. 179 (1924). Heliophytum tiaridioides DC. Prodr. ix.
(1845); Fresen. in Mart. Fl. Bras. viii. pt. 1, 47 (1857). Helio-

um monostachyum, var. tiaridioides Chodat, |. c. ser. 2, ii. 816
(1902). Tiaridium heliotropoides Cham. |. ¢. 453; Don. Gen. Syst.
i .

a . Te0 (G NY BD): Orin © Ris Piedras, 300 aca oe
; ieareiasiies ose m.,
); Tartagal, 500 m., Schreiter 3372 ” eT . Jusvr: El Gato, Lillo 9877

Puerto Famaren, Alto Parané, 1383, Niederlein (BD); Yocy. Rio Alt Alto eed
1, Alto Fi i; “a

Hieronymus & Lorentz 619 (BD).

Monte de Buenavista, 450 m., Steinbach 2045 (B a etghe
ae BD); Buenevate, 40m, Sianbock 619 Bia, ag ae sAtsiniach 279 i
BD); Buenavista, 450 m., Steinbach ee: x oe. » :

| ie Renate Tees cr K. =, BD); Valenzuela, ta

STUDIES IN THE BORAGINACEAE fa

ia (BM); Sapueay, Hassler 11754 (G, US, BM); Caballero, Morong 409

BRAZIL Rio E DO Sux: Neu Wiirttemberg, Bornmuiller 505
“este Porto — Tweedie 135 (K); indefinite, Reineck & Czermak 2 Ky.
A CATHARI Blumenau, Ule 853 (US, B BD); Desterro, Schenck 442
‘SD St. Catharina, Muller 1 149 (K). InpEFtnire: no no locality given, Sellow
(K, BD ; IsoTypPEes of H.tiaridioides).

iv. Abt. 3a, 96 yon Kia Rev. Gen. iii. be ra 205 (1898) ;
Briq. & Hochr. Ann. Conserv. et Jard. Bot. Genéve iii. 162 (1899);
Fries, Ark. Bot. vi. no. 11, 23 (1907); Molfino, Pikegpon vii. 179 (1924).
Heliophytum monostachyum DC. Prodr. ix. 556 (1845); Fresen. in
Mart. Fl. Bras. viii. pt. 1, 47 (1857); Warm. Kjoeb. Vidensk. Meddel.
1867: 19 (1868). Ti iaridium monostachyum Don, Gen. Syst. iv. 365

tropium monostachyum, var. ovatum Chodat & Tes. Bull. Herb.
Boiss. ser. 2, v. 483 (1905).

Northeastern oh csapneae apa ce and eastern Brazil; also in
— and Cen America

vii Sach ae Fi K. 83 the J drgensen (G,

bigs! soto _ Miso ken a vanes ae eG Sota

Posadlas, att ianchi 1 : Cor-

pend é Grande, 1883, Sy jahpelaop Nieder-

_ Janerro: Imbuhy, ass 7 (NY, | BM); Rs 5 i
_ Wilkes Exped. (G, NY, US), Gay (ND Leachenault (NY), Gato 1
BD), Burchell 1262 and 1763 (K), Gardner 5561 (BM).
: Caldas, Henschen 200 (US); indefinite, Claussen (NY, Kk) "Bamia: nea near

Bahia, Blanchet 864 (NY); Bahia, Guillos (NY).

if i Pohl 1582 (K, BD) ‘and 1591 (K), Sellow (K, BM, BD, ISOTYPES ft

monostachyum).
ae Indefinite: Eggers 3 15260 Aides ea

8 JOHNSTON

pads have both been distributed under the same number by Hassler
and Morong, which suggests that they grow together. The name,
H. tiaridioides, clearly belongs to the plant with non-sulcate carpels,

for Chamisso illustrated this form and a copious suite of specimens at
Berlin, collected by Sellow and determined as H. tiaridioides by
Chamisso, also represent the form with non-sulcate carpels. The
application of the name, H. monostachyum, is less certain. Chamisso
did not describe the fruit which appears to have been unknown to
him. In the herbarium at Berlin there are four sheets of a collection

= on the pours I have seen no authentic material of this species,
but am placing it in the var. schizocarpum since it comes from Rio
Janeiro whence I have seen only material with sulcate carpels.
Similar reasons cause me to place under that variety

: - geographical
H. pes. and H. monostachyum, var. ovatum, both of which

specimens.
2. HL an (Speg.), comb. nov. Valentina patagonica

ae Er Poe: 90) Act ae Argentina liii. 78 (1902);

Dusén, Ark. Bot. vii. no. 2, 32 (1907); E. & P. Nat. Pflanzenf.; Er-

ngsh. ii. 306 (1908). diggs patagonica Speg. Anal.
Mus. Nac. Buenos Aires, ser. 3, ii. 9 (1903); Macloskie & Dusén

[Rev. Fl. Patag.] Rep. Princeton ph Exped. Patagonia viii. sec. 3,
214 sible

ARGENTINA. Paraconta: Santa Cruz: Pan de Azucar near Rio
Chico, Dee. 5 1 1897, Ameghino (G). Guusvr: Puerto Madryn, 1907, Dusén

A remarkable species which, because of its s

— H. curassaricum, though in fact only very distantly related

ee ee H. “seo It suggests ‘that species in its habit, frequently .

leaves, saline habitats, and of course flowering and fruiting
s. Certinly H. patagonicum is not to be generically separ-
these indubitable relatives.
H. veri sicifolium Griseb. [Pl. Lorentz. 184] Abh. K. Ges.
x6ttin So ieee €. Xxiv. mn

STUDIES IN THE BORAGINACEAE 9

(1879); Hieron. [Pl. Diaph. 181] Bol. Acad. Nac. Cien. Cérdoba iv.
375 (1882); Giirke in E. & P. Nat. Pflanzenf. iv. Abt. 3a, 96 (1893).
H. salsum Griseb. [Pl. Lorentz. 185] |. : xix. 233 (1874) and [Symb.
Argent.] |. c. xxiv. 271 (1879); Giirke, 1. ec. H. repens Griseb. [PI.
Lorentz. 185] 1. c. xix. 233 (1874) and [Symb. decd, c. xxiv. 271
(1879); Giirke, |. c.; Kuntze, Rev. Gen. iii, pt. 2, 205 (1898); Fries,
Ark. Bot. vi. no. il, 22 (1906). H. repens, var. medium Kuntze,
l.c. H. repens, var. stenophyllum Kuntze, 1. c.

Northern Argentina and adjacent Paraguay and Bolivia, usually
in saline —

ARGEN CérDo: near Quinta Santa Ana, Kurtz 8863

(NY); Cirdoba, ho ay, (BM oO Cordoba, ‘71, Lorentz 91 and 609 (BD).
SANTIAGO - 1892, Kuntze (NY, 2 collections one the Typ:

“ ucuman,
Lillo 1465 (G); Tucumén, 500 = ae? 1278 (G); Tucumén, 1837, Twveedic
(K);Tu . cuman, Hieronym ed Loum tTe ts). ;Santiago t to Tucuman, Tweedie
PARAGUAY — nant pd eae eta 1439 (BD); oe
cepcion, Oct. 1892, ge hts: choc var. stenophyllum; G.
LIVIA. Tara ?: Villa Montes, Nov. 20. 1924, 460 m., oe
(US, BD),
A well defined species which is rather variable in leaf-outline.
Grisebach’s three species appear to represent merely ecological forms
of a single species. The luxuriant phase with mavee 2.54 mm. long

oes a iar variable in smn It is practically only this
character that serves to Sor gar the species from its close —

10 JOHNSTON

4. H. Schreiteri, sp. nov., humile herbaceum sparsissime breviter-
que villosulum; ramis 1-3 dm. longis parse ramosis inconspicu
seque villosulis decumbentibus apice ascendentibus; foliis firmiter

ebracteatis inconspicue villosulis; calycibus sessilibus, lobis ovato- -
lanceolatis 1-2 mm. longis acutis; corolla alba parva 1.5-2 mm. longa
subeylindrica, lobis ovato-oblongis ca. 0.8 mm. longis ca. 0.5 mm. ie
apice rotundis, tubo intus glabro; antheris ca. 0.8 mm. longis late
lanceolatis subsessilibus medio tubo affixis; stigmate sessili lato;
fructu glabro, carpellis biovulatis ca. 2 mm. longis leviter dorso-
suleatis indivisis apice bidentatis

sage ky aM og Piguirendo, Ordn, 500 m. alt., Feb. 9, 1925,

Evidently related to H. veronicifolium but differing in its complete
lack of malpighiaceous hairs, small corollas with short broad lobes, and

at maturity do not ocak, - into uniovulate nutlets.

II. Section Schobera (Seop.) Johnston. Schobera Scop. Introd.
158 (1777); type-species, * angiosperma Murr. — Raf.

5 caches Wesadaca sch ack % Maik iacly welch Wt
ection Hypsogenia. It differs, however, in having the corolla glab-
rous within and in its very different t habit of growth. The fruit in

I beliewe that Schobers is « dexivative from CC
5. angiospermum Murray, Prodr. Stirp. Géttingen 217
(1770). Schobera angiosperma Murray“ex Scopoli, Intr. 158 (1777);
Britt. & Wilson, Bot. Porto Rico vi. 134 (1925). Heliotropium parvi-
florum L. Mant. ii. 201 Aig Cham. Linnaea iv. 455 (1829). st
iflorum DC.

RES tae 89 (1838). cesexscaas "eo Wi x

STUDIES IN THE BORAGINACEAE i}

Lehm. Nov. Act. Acad. Caes. Leop. Nat. Cur. ix. 127 (1818) and
Asperif. i. 33 (1818); Cham. Linnaea iv. 455 (1829). Heliotropium
pion Nov. Gen. et Sp. iii. 89 (1818); Cham. I. c.; not
Willd. (1818). ? Heliotropium patibileense HBK. |. c. 87. Helio-
phytum foetidum DC. Prodr. ix. 553 (1845). Heliotropium foetidum
Salam. ex DC. |. c. Heliotropium rugosum Mart. & Gal. Bull.
Acad. Brux. xi. pt. 2, 336 (1844). Heliotropium oblongifolium Mart.
& Gal. 1. c. 336. Heliophytum portoricense Bello, Anal. Soc. Espafiol
Hist. Nat. x. 297 (1881).
Northernmost Chile (Tacna) and eastern Bolivia northward
through Cen America and the West Indies to sou
United States (Texas and Florida).

CHILE. Arica: Arica, Meyen (BD).
az: below Chuguiaguillo, east of La Paz, ca. 3750 m.,
Hauthal 174. (BD); Beni River, — 1434 (NY).
PERU. Arequipa: M Bay, Macrae (K); Arequipa, Meyen (BD).
Ica: Pisco, ca. or wait ceo i 5394 (BD). Lama: ea. 400 m.

P. near Viscas ca. 1900 m. Pennell
14458 (G, Ted Lar, 12 m., Puan i2217 (G, EM); south of Santa
Clara, Rose 18626 ( US); Rio Rimac, 1800 m 7, Safford (NY);
Valley of Rimac, 263 (BM); Chosica, 1877, pa (K); Callao
to Lima, 1876, Savatier ( ); Lima, lade Poherstme (8) Ball
), Seemann 692 (G, K), Cuming 1032 (BM), André K1395 (K),
Nation (K), Mathews 396 (K, BM), Meyen (BD), D’Urville & Lesson (BD),
Gaudichaud (BD). Huancave.ica: between an
1500 m., Weberbauer 5394 (FM). Lipertap : Barraneo in valley between
Pacasmayo and railhead, 2100 m., 1912, Forbes ~~ Prora: Talara
LaBrea, Heneht 38 (G). Inp a without data, Dombey (FM, BD);
; a sungystachspum

R. ' Guayas Cerro Magior Alto, 360 m., 19 1923 Anthony & Tate 5
(US); Guayaquil, 1926, Mille 4 (G) uayaguil, Sinclair (K); Chanduy, 1862,
Spruce 1886 US) and 6489 (BM); isla 836, Barclay 407 and 2447
(BM). Mawnasi: ues Bay, 1923, Geto é Tale 121 (US); Caraques
prverina 2 758 . Ganapacos IsLanps: Island,
Snodgrass & Heller 812 (G); Albemarle Island, Snodgrass & eller 80, 160
yang a than 180 & H w eller 463 (G), ma 3218 (G), Lee (G),
pg 21 (G), Agassiz (G. _FM, kK), Andersson (G), ‘eum

<& Heller 510 (G), ceed 3216 (G) and 3217 (K); ( e Island, Wh.

& Beebe 12 (NY); Gardner (G), Stewart
$219 (G); Hood Island, Baur $ Baur os 2G), Snodgrass ¢ tH Heller 7. 4 (cy. Stewart
(G); (G); Narborough Teak, Gentes & tolke s6a (GN: Tome ais oem

: ; Tower I
& Heller 792 ( )

COLOMBIA. Cunprnamarca: on ge ghee Rusby & Pennell
(G, NY, US); Villeta, Hartweg age A pgp , BD); La Mesa, 1400 m., 1856,
oe Triana 3752 (BD); ¢ Casas Viejas, 310 m., André 1847 (K); os

André ye (K). Sanranper vet Norte: vieinity of Ciieuta, ea. 215 ;

Killip <a ee ATuantico: Barranquilla, Piitier 1565 ws)

7 ca. 5 m., Pennell 12037 (G, NY). Botrvar: Tur ca.
S vm, “Pennell sist G, ‘NY, adie dapat Cartagena, ca. 110 m., .

12 ‘ JOHNSTON

Killip & Smith 14058 (G); Cartagena, bigger agh 10 (US); Cartagena
Schott 27 (NY). Maapauena: Santa Marta, 30 , Smith 873 (G, NY. US,
ag K, BM, BD) and 872 (NY); Santa hl! Schlim 950 (K). INDEFT
no locality given, Karsten (BD); Guarum, Magdalena, André 491 (i.
: Peninsula de Paraguana,
1167 (US). Canapo: ARABOBO: Porto Cabello, 1917, eae & Haman 1167 (Us):
Kareits 29 (BD). ARacu UA! Colonia oA Fendler 910
: fara Maracay, 440 m., Pi
Dist. Feperav: Blanco, Pittier 10193 (G, US): fo Guaira, Curran &
Haman 859 (G, 3, US) a and tre G NY, US), André 129 (NY), Birschel (K);
Lower Cotiza ca. 1000 , Pittier 7958 (US); Car
(G). Sucre: gato Rimland 268 A om IsoTYPE of H. scorpicides);

Humboldt 268 (Herb. Willd., type of H. latifotium); Benge A of
Cristobal Colon, Broadway 12, 610 and Bsr (G, NY, US). th-
out locality, Puncke 41 (BM), Funcke & & Schlim &

BRAZIL. Prrnameuco: Pernambuco, Gardner 1079 (NY, US, BM),
Ridley, Lea & Ramage — CearA: Villa do Ico, Gardner 1784 (K, BM).
Atac6as: Rio San Francisco near Propia, Gardner 1361 (G, NY, Us.

: Bahia, Blanchet 595, 973 and 1027 (BM), Herb. Martius (BD}, Salz-
mann (K, isoryee of H. foetida). INDEFINITE: via Miata:
Herb. Martius (K, BM).

A loosely branched weedy species ang oe in tropical

America. It is very readily oo a <raley

America. This Brazilian form has been described as Heliophytum
foetidum. At most, however, it is worthy only of varietal recogniti

III. Section Hypsogenia Johnston; type-species, Heliotropium
wcncueiees E&P.

A group consisting of merely two very closely sina eas of the
mountains and plateaus of western South America. It is immedi-
ately related to the section Schobera, but differs sufficiently in habit,
flowers and _ to merit recognition. I believe it to have evolved

_ from the section section Coeloma and that its closest relatives are in section
_ Schobera.
Wer 6. H. microstachyum.
Stems and leaves shaggy villous........................ 7. H. abbreviatum.
Loe 5. L microstachyum R. & P. FL Peruv. ii. 3, t. 110a (1799);
DC. Prodr. ix. 550 (1845). Heliophytum brachystachyum DC. Le.

554. Heliotropium brachystackyum Griseb. [Symb. Argent.] Abh.
Sertum §

KE. Ges Wiss. Gottingen xxiv. 271 (1879); Hieron [

Bol. Acad. Nac. Cien. Cérdoba iv. 64 (1882); Giirke in E. & P. Nat. a
age iv. Abt. 3a, 96 — Nova Acta Reg. Soc. Se.

STUDIES IN THE BORAGINACEAE 13

Upsala, ser. 4, i. 111 (1905); Rusby, Bull. Torr. Bot. Cl. xxvi. 149
(1899). Coldenia phaenocarpa Ph. Cat. Pl. Itin. ~~ 56 (1891).
Heliotropium phaenocarpum Reiche, Anal. Univ. Chile cxxi. 23
(1907) and Fl. Chile v. 195 (1910). Heiolveptun Bangii Rusby,
Bull. N. Y. Bot. Gard. iv. 414 (1907).

Northwestern Argentina northward along the puna and cordilleras
to Ecuador.

ARGENTIN : Los Berros, Sierra Fama Hieronymus &
Niederlein 591 BD). py Race agp ng eur: de Pedra, Stra Vela, Hierony-
mus & Niederlein 14 (BD). San JuaN Teonsi

Lomas de la Banda, Castillon 3180 (G); ‘Taft to. HG os) 2075 (G);
Valk 2000 m., ay: 3106 vate oa Guiaea, Feb. 1916, "Haxnen (G);
\ 25.

dios); Pe 7 * 3000 ms Asplund 3155 1G) Ris
m; photo.); Poreo, Uyuni, 8 1
chisla, 3425 m., Bender 29 (BD). La Paz: Tiahuanaco, Hill 327 (K); Sur

chacac m.,
NDEFINITE: withou ity, Bang 1971 (G, NY, US, FM,
ca BD}. 2505 Jka NY, i. K) and 2847 (N NY, TYPE of H. Bangii; G,

FM, K, BD, 1
sang 5 Hoes! Weberbauer 7484 Mapas Puno:
Doge S100 Shepard 120 120 (@); Ch ca.
13358 (G, FM): Laguna de f hirer wast woe (BD); ‘Aloe de Toledo, "4650 m.
Meyen (BD). Arequipa: Cotahuasi, ca. 2650 m., Weberbauer 6871 (BD).
Cusco: Ollantaitambo, ca. 3000 m., Pennell 13660 (G, FM). Hvancavetica:
Huaytard, ca. 2650 m., Weberbauer 5 5427 (G, FM). Ica: mountains
Pisco, ca. 2650 m Miele fe nad (BD). Junin: Tarma, ca. 3150 m.,
Weberbauer 2370 “BD). Lima: Mantucana, 2400 m. , Machride & Feather-
stone 202 (G). Huanuco: Ambo, 2100 m., Macbride 3168 (G); Huanuco,
Kanehira 258 (G). InpeEFIntre: without gare as (FM, Ae
Pavon (herb. Willd.), Ruiz (BD, tsoryPe of
ECUADOR. Cxrmmorazo: Ri —__ (G, 7s (Gk BM)

oo 2800 m., Rimbach 155 (BD). wl ase ae

The well known name, H. brachystachyum, is here replaced by the
unquestionably synonymous and very much older H. microstachyum.
ecsmgraenag bisa - a H. eo _—* no doubt

a oe ane H. "permachode. he sees, H. ——* :
1 send ieevamnaers he species. —

14 JOHNSTON

7. H. abbreviatum Rusby, Mem. Torr. Bot. Cl. iv. 224 (1895);
Kuntze, Rev. Gen. iii. pt. 2, 205 (1898).

Known only from Bolivia in the western ay of Cochabamba.

BOLIVIA. ory Rio Tapacari , Mare — 1892, Kuntze
(NY); ba, 2600 m., M Sch toe ae (G, NY, US. , BD); Cocha-
bamba, Bang 924 (NY, Tyee; G, US, FM, K, BD, tsoryres). INpEFinrTe:
no locality, Bridges (Ks, BM).

Very obviously related to H. microstachyum and perhaps no more
than a pronounced geographical variety of it. It agrees with that
species in habit and reproductive characters and differs primarily
in its very shaggy-villous rather than short appressed pubescence.

IV. Section Halmyrophila Johnston; type-species H. curassav-
icum L.

This section consists of a single species which occurs in saline soils
in temperate and tropical America and in scattered localities in the
Old World. The plant is a completely glabrous succulent, with the
4-suleate fruit crowned with a very broad stigma in which | Son cohmun
The inamediate affinities of the section are uncertain. I am inclined
to believe, however, that it is related to sections Hypsogenia and
Schobera and that it is an American :

8. H. curassavicum L. Sp. Pl. i. 130 (1753). H. glawewm Salisb.
Prodr. 113 (1796). H. glaucophyllum Moench, Suppl. Meth. 147
2) henopodioides

= : ebehumgeeragitns: at Tel, Jour. Linn. Soe. xsi. 227 (1884).

le sed Cie aed en Bolivia along the coast to northern
ee South America, thence northward oo Central America, and the
West Indies to the United St Z

- Rio Nexo: General Rocas, Fischer 230 (G, NY, US,

a a Batueo, 1913, Baeza (IP); Cerro de Renea, Philippi
2 alsin se Lag gunn Alo, alo, 818, ete, MS, 1sorrees of ee

Page (RS A ce ee ey ey Rl ee a eR ee crc eee) pete ee Ge ee me ne ee i

STUDIES IN THE BORAGINACEAE 15

Santiago, Philippi (MS); Punta, Claude-Joseph 2142 (US); Cartagena,
Claude-J oseph 977 aL Montaria, Claude-Joseph 1326 (US). VALPARAISO:
— pan del a —- uchtien (US); Liai Llai, Bridges 318 wr ACcON-
267 ( uimBo: Algarroba, 1915, Ochoa
ap): Fy reese sing * Hastings 584 (US). Atacama: Puerto r+ Hot 1
Volckman (MS); Co , Johnston 5008 (G). Tarapacd: Tamentica, 1809,
Johow (IP); near quique, 1881, ———~ _ Tambillo chica, 1
Rahmer MS). T Tacna: Arica, Buch goth rsd 4378 and 4379 (US); Arica,
1882, Bail C rie ke Bat) peg i 1880, Ortiga (G); Azapa, ca. 150 m., Werder-
mann 711 BM); Tacna, 1880, Ortiga (G); Taena, 1890, W oitschach

BOLIVIA. aera near snow-line, Mt. Tunari, Bang 1039 (G,
PA: "Mollendo, Weberbauer 388 (BD); Lomas -

AREQUIPA: near , Weber

bauer 5732 (BD). Ica: Paracas Bay near Pisco, 1912, Forbes (BM). Lima:

Shovacen tion 5881 ( ag Rio Rimac near Lima, 1882, Ball (US); ny
reille &

oECUADOR. ee as: Chanduy, Spruce 6491 (K). Manasi: Bahia i
ae Eonrores BT 753 (NY). ESMERALDAS:
tp gar, peared, iad Island, Stewart 3196 (G); ihenacls
Stewart 3197 (G, US), Belt ( ); &

a Chatham Island, Agassiz (G, US), SO). Charles Iden Keke
427 (K); Gardner Island, Stewart $201 (G); Hood Island, Snodgrass & Halles
395 785 (G); Indefatigable Island, Snodgrass & Heller 660 (G), Stewart $202 and
3203 (G, US), ao Wheeler, Rose & Beebe 15 (G, NY, US); Seymour

eller 560 (G).
i oe (NY, K);

LENA: Santa arta, ca. = =

544 (G, NY, US, FM, K, BM, BD); Santa Marta, Schitm 926 (K). Goasma:
at =a 538 (K). Inpermire: no locality, Linden 1526 (K) and
1

BM
OS RILISH GUI SG Cee Hitchcock 16572 (G, NY, US); indefinite,
Jenman 2165 on 4466 (NY) and 5471 (US, BM).

Var. argentinum, var. nov., a varietate genuina differt corolla

ARGENTINA. Dcmnon Acne Arres: Carhiie, Lorentz 131 (G, BD); ———
Parodi 7698 (G); Buenos am hage Hauthal (NY); Buenos Aires, "weedie
5 (BM); nos Ai Rodrigue: 2 208 (G). Menpoza: Rafael,

and Jachal,
): Hineonada, 1876,

re kx keen Vinching, Hieronymus ce Nideran 255 (BD);

Las ranquitas, es ee + ee

16 JOHNSTON

MARCA: Le Schickendantz 238 (BD). San Luts: Bajo de Velis, Kurtz
8485 (NY ); Queb. Salado, 1882, Galander (BD); Santa Barbara, 1882, Ga
(BD). Cérposa: Posta de San Agustin, 1879, Galander (NY, foam
Grandes between Balde de Nabor and the frontier toed La Rioja.
mus & Niederlein rd (US, BD). Santiago: Pinto, Oct. 1892, 2, Runize (NY (NYS;
indefinite, Lorentz 61 ( BD). TucuM4n: Amaic 2000 m.,
(G); Tapia, 600 alg Schretter 191a (G); Cadillal, 500 m., . Schriter 191 (G);
Dulce, Schreiter 45. 49 (G); Cuesta Grande, Trancas reiter
344 (G); Rio Salf, pedo m. , Venturi Tarami, Rio Salt, Lillo 2489 (G).
Jusvy: Voledn, 2000 m. , Schreiter 2602 (G); indefinite, Castillon "10152 (G).
- sik s Palmas, Feb. 1 oe Ser assure 43 (G, rrPe). F osa: Fo
eb. 1 Orgensen 2246 . Lyperintre: Laguna de : Banedaltind,
a Lorentz (BD); Las Raluras, Aug. 30, 1892, no collector given, no. 335

URUGUAY: Piriapolis, Smith 114 (US); Punta t. Maldonado,
oe 17676 (BD); Malvin, = 76710 (FM); m Moet 1839, Tsabelle

Aeon a Bing woes, Morong 790 (G, NY, US, FM); Laguna de las
o River, Morong 1058 1G Ny, US); Rio Palado, Hassler

3257 (BM, BD): ; athens NY, US, FM); lower reaches
of Rio Pileomayo, Rojas 188 (B. a ge a near Rio Apa, Malme 1016

(BD).
GRANDE DO Sut: indefinite, St. Hilaire C2249 soins and
Malme 314 BD). Sao Pavuto: St. Paul, Glaziou 19681 (K,BD). Inp
no locality, Sellow (G, BM, BD).

A widely distributed and characteristic species of tropical America,

ne

forms of HT. cormemnioun Shock aocar ta the leaiad & States, i. e. var.
obovatum DC. (= H. xerophilum Cockerell), being less succulent,
searcely Ad aged and having a less firm much more decidedly funnel-

_

pe pela wand Oran pepe a

o@ Section Plagiomeris Johnston; type-species, H. paronychioides ©
DC.

y calyx,

STUDIES IN THE BORAGINACEAE iz

Key To Species
Herbage (especially the younger parts) stipitate-glandular;

corolla 3-5 mm. broad; leaf blake ORONOUIAE. 206) 9. H. Geissei.
Herbage not at all glandular; co rolla 1-3 mm. broad; oe
munnneniite te Gntee. 3S. eS i el paronychioides.

9. H. Geissei F. Phil. Anal. Univ. Chile xc. 352 (1895); Reiche,
Anal. Univ. Chile exxi. 237 (1907) and Fl. Chile v. 195 (1910).

Central Chile; rare and local.

LE. Santiago: El Alba near 1891, F. re. eg

Phe mon ol lake, Batueo, 1913, Bac oo Aconcagua: San Lo
1889-90, Geisse (MS, Trpe; G, photo

A very distinct species aoe of limited distribution. It is
known only from central Chile where it appears to be rare and local.

iche, |. c., suggests that H. Geissei might be one of the European
species a ie into Chile. I ae however, been unable to
match it among European collectio As it agrees in very many
details with H. paronychioides I am gaia to believe that it is an
indigenous species and a close relative of that species of Chile and
a

H. paronychioides A. DC. Prodr. ix. 565 (1845); Clos in Gay,

FL a iv. 459 (1849); Giirke i in E. & P. Nat. Pflanzenf. iv. Abt. 3a,
96 (1893); Speg. [Fl. Patag. ii. 37] Anal. Soc. Cien. Argent. liii. 79
(1902); Reiche, Anal. Univ. meee 8 236 (1907) and Fl. Chile v.
194 (1910); Macloskie, [Rev. FI. tag.] Rep. Princeton Univ.
Exped. Patag. viii. sec. 3, 212 shld Tekes: Physis ii. 113 eRe
Hauman, [Veg. Cord. Mendoza 149] Anal. Soc. Cien. Argent. Ixxxvi
303 (1918).

Along the Andes of Argentina and Chile between lat. 28° 30’
and 41° S.

ARGENTINA. Rio Necro General Rocas, ca. 300 m., Fischer 234
a K, BM); Saw Oselne ak de Bariloche, Lake Nah 2.
m., Buchtien 1338 (G) and 117 (US). EUQqUEN: Cal , Nazarre 10

trahue,
(G); Sonata Parcds 2156 (G); Zapala, a 135 (K); Las Lajitas, 1000 m.,
Comber 231 (Hh). MENDoza: Rio Salado, Gillies (K). San JUAN: Arroyo

Chan, 155, Germain : (MS, K, ei ef de, Chillan, wad ta

ges 1877, Philit i (MS); La Cueve,; 1887, Gets TM) Cnt ac de Chillan,
1800 m., Bi ta (BD). ‘Tatea: Andes de Talea, 1922, Silva (IP).
Gontc6: Cord bash Sag > | 2250 m. ee ers 220 OR) a 1

18 JOHNSTON

Cord. de Santiago, ype ig fd (BD). Coguimso: Cord. de los Patos,
1844, Alamos (MS); An f Coquimbo, 1861-2, Volckmann (MS).
A small prostrate annual herb with small white corollas that com-
= turn bluish in drying. Although somewhat suggesting H.
acrostachyum in gross habit, it is very distinct and is only distantly
mete to that species

VI. Section Tiaridium (Le hm.) Griseb. Fl. W. Ind. 485 (1861).
Tiaridium Lehm. Asperif. i. 13 (1818); type-species, H. indicum L.
Tiariditum § Hieranthemum Endl. Gen. 646 (1838); type species,
H. elongatum Lehm. AHieranthemum (Endl.) Spach, Hist. Nat.
Veg. ix. 31 (1840). — § Scorpiurus Kuntze in Post &
Kuntze, Lexicon 271 (1904

This section consists of ce species, one a widely distributed weed
in the Tropics of the Old and New World and the other a rare plant
known only in eastern South America. The section is a very well de-
fined one because of its coarse weedy habit, coarse pubescence, salver-
form corollas and especially by its conspicuously ribbed multicellular
glabrous . The affinities of the section are uncertain. It
appears, however, to be most related to the section Heliophytum.

Key To Species

— di
OTS on a. Ok oe ee ee apices nl .
Carpels becoming avert, mmonly 2.5-3 :
long, with strongly bi in cues Pee . oe es ie. H. indicum.

11. H. elongatum Hoffm. ex R. & S. Syst. iv. a Giirke
in E. & P. Nat. Pflanzenf. iv. Abt. 3a, 96, fig. 39c (1893).
: Lehm. Asperif. i. 16 (1818) and eis 1, t. ‘ ors
R. & S. Syst. 1. ¢.; Cham. Linnaea iv. 452, t. 5, fig. 1 (1829). Helio-
phytum elongatum DC. Prodr ix. 555 (1845); Fresen. in Mart. FI.
_ Bras. viii. pt. 1, 47, t. 10, fig. 8 (1857); Warm. Kjoeb. Vidensk. Meddel.
1867: 20 (1868).
oa. we genuinum.—Corolla small, with a tube 3-4.5 mm. long and
a limb 2-3 mm. broad - ymy given above. |
Brazil

ee ‘GU. =e - Cerro Largo, —_ 18458 gp =
- “sat BD); sae of Rio sag ee 132 G. Commer!

nian, Morong 56 (NY, BB; Goalies de -

—Synon:
en ee ——

STUDIES IN THE BORAGINACEAE 19

BOLIVIA hago Paz: Beni River, Rusby 1 NY). Santa Cruz:
Buenavista , Steinbach 5100 (NY, FM) po Pia (K, BD); Bafiado
i y, 4 ; ‘Steinbach 7130 (FM).

1 Pird: hey

BRAZIL. Minas Gerass: Pauso big 1927, Hoehne 19208 (G). Banta:

acd — 7913 (G). ‘Guani Fortaleza, Rolland (K). Frernanpo

locality, Moseley (K, “SM Ridley Lea & Ramage 104 A & B

(BM) and 104 (G). Inpertnrre: Bras. trop., Sellow (BD, Trpz?; K, tso-

‘ar. Burchellii, v ., @ varietate genuina differt corolla con-
spicua, tubo 5-7 mm. oe limbo 7-10 mm. lato.

BRAZIL. Goraz n Conceicao and Natividade, upper Tocantins
River, Burchell 3191 ie’. paar K, 1soTyPe).

A peculiar plant very closely related to, and much resembling
H. indicum in all details of habit. It differs from that species only
in the structure of its fruit. The fruit consists of two strongly rib
brown glabrous << whiels each tardily | break up into two angulate
nutlets. The carpe only v
apically, 3-5 mm. long, and hows entire or phacavehy. hidentinulate
apices. In H. indicum the carpels are usually 2.5-3 mm. long,
become very strongly ie ce and have conspicuous bidentate

apices.

12. H. indicum L. Sp. Pl. 130 (1753). Ttaridium ailicute Lehm.
Asperif. i. 14 (1818). Heliotropium horminifolium Mill. Gard. Dict.
ed. 8, no. 3 ree) Heliotropium cordifolium Moench, Meth. 415
(1794). Heliotropium foetidum Salisb. Prodr. 112 (1796). Eliopia
serrata Raf. ea Tellur. 90 (1838). E. riparia Raf. 1. c. Helio-
phytum indicum DC. Prodr. ix. 556 (1845).

From northern Argentina extending northward through the tropies
to southern United States; also in the tropics of the Old World. Ap-
parently rare on the Pacific Coast of South America.

ARGENTINA. Tucum4n: Las Herreras, Lillo 14147 (G). Sauta: near
eee Hieronymus 350 (NY, BD); Oran, Lillo 18098 (G);

ee ox , Schreiter 3607 (G@ Oran to Rio de los
Pie (G). Cusco: yer bapa iat ra sy 7 oso a 2244 (G, US). Formosa:

oie Rs israye jer (HBR Wb saad Parsee
PEM) and Hard, oe 1 si Eth

ada, Sladen 620 (K, BM); Care, Hoch aoe BM). Ric
: artius ms
comayo betw. S. Cruz and Diamantina, Moore 633 (NY, ‘pM. BD Sio

anelro. ie
sag) 14138, tudichaud (BD), (2). Bed. (NY), Andean (U8), lain
: ‘tpbi; ‘Bahia, Chase 7880 (G), Lee nes ee

i

20 JOHNSTON

FM). Pernampuco: Pernambuco, 1838, Gardner (K, eg Fernando
Naronha, Darwin (K). MaraNnHAao i Asst, Snethlage 377 (BD). Pari:
— at co a Spruce 476 (K ): M te Alegre, 1 1873, Traill (K). Ama-

o Jurud, Ule 5222 eB D). INDEFINITE: Cac hambi,
1300, A itera ira (BD); Gedelaite, Sellow (K, BD), Burchell 1009 (K), Pohl 1592

FRENCH GUIANA: Iles de Salus, Sagot 445( BM); Cayenne, 1845, Rothery
fe dahon ee Broadway 8 (NY); Maroni , Wachenhei m 287
290

DUTCH GUIANA: Paramaribo, Samuels 160 (G).
BRITISH GUIANA: Kamakusa, upper Mazaruni, La Cruz “ee (G, NY);
iver, Jenman 5476 (US, BM); Lamaha Canal, J n 4568
and 5044 (NY); Waramuri Mission, Moruka River, samlebcan ict,
La Cruz 2600 (G. NY, US); Comaca, Moruka River, La Cruz 1058 (NY, US);
: 16684

Succes obal Colon, 20 (G,
Moore ee mG Moritz ee “aM. BD). Gallmer (BD). Tek: Colonia
ovar, DLOMBIA. 3 ac K). Carasoso: Puerto Cabello, Karsten 30 (BD).

a AGDALENA: Santa Marta, 90 m., Smi ( NY,
US, FM, BM, BD). Artantico: Barrang se Paul 16 (US), Ariste-Joseph
709 (U8); Puerto Colombia, ca. 5 m., Posse 2 2035 (G) and 12036 (G, US);

G
delta of Rio Magdalena ni ear Barranquilla, Kailbach 19 (BD). Botrvar:
Pennell i Heriberto 6 (US), : Behott 5 (NY); Tierra a Rio Lear ca. 65 m.,

3 (NY); Cafiabetal, ca. 95 m., Pennell 3871 (NY); Bodega Central,
Rensch 1 (BD); Turbaco, ea. 250 m., Killip & Smith 14219 ‘Gh CunpDI-
NAMARCA: ta Mesa, — ist? (SY, BK, vost Huma: La Plata, ca.
1200 Lehmann BT 6 , NY, FM, :
wy, Triana 3752 CBM). Cauca, 1600 m., Pics (BD ——
de ‘Caraman Rio Cura, ca. Pen: nell 10832 ACS i, US); Boca

ra Masa sf: Recren es kee: 590 at 15009 (PM : ng a
Chanduy, Spruce 6492 ;’ Guayaquil, Hitchcock 1 NY, US ),
Mille 2 (G), aga (K). Gatapaaos Isias Charles I. Snodgrass &

NDS:

ms Age gs Lee (G), Stewart $208 (G); Chatham Island, Stewart 3206
PERU. meee Lower, Itaya, Baja Amazonas, 110 m., Tess-

BOLIVIA, (BD). Prora: Piura ura, Townsend 1393 (US).

La Pas r, Rusby 1433 ( at rg dae fe
savin. Woleone 100 (NY, US, Tess Sant. uz: Bafiado
de Piray, 450 m., Steinbach 7130 (BM, B D).

_- sty recognized, coarse, Ww weedy annual, which exhibits remark-

- I suspect that the species is indi t America, although -
se the doe of Amen its weedy Se

STUDIES IN THE BORAGINACEAE 21

it to extend its original range there. The rarity of the plant west of
the Andean Crest is a striking fact. Because of this peculiarity of
range and the circumstance that its only close relative, H. elongatum
(a species more primitive in its fruiting structures), is endemic to
eastern South America, I am inclined to believe that H. indicum is
Brazilian in origin. Certainly none of the Heliotropiums of the

~

Old World seem at all closely related to it.

VII. Section Heliophytum (Cham.) Griseb. Fl. W. Ind. 485 (1861).
Heliotropium *** Heliophytum Cham. Linnaea iv. 458 (1829) ; type-
species, H. anchusaefolium Poir. Heliophytum DC. Prodr. ix. 551
(1845).

This section is confined to southern South America east of the
Andes. It is most closely related to the section Cochkranea, from
which its villous corolla-throat and -tube, its commonly multilocular
carpels and its different habit seem to warrant its recognition as a
distinct section, at least provisionally. The recognized are
not clean-cut and must be subjected to extensive field-study before
they can be given a satisfactory and detailed treatment.

Key To Spectres.
Cauline leaves sessile or subsessile.
Stem — lea aeongent tawny-villous, somewhat

ie
Peele Ga ee H. amplericaule.
Stem and er buh white villous, sparingly if at all ela

WOR rca ee ee H. phylicoides.
Cauline leaves with a a petiole, 1-4 em.

long.
Leaf-blade a cuneate base...... 15. H. nicotianaefolium.
— selec sins ti

13. H. amplexicaule V: Symb. Bot. iii. 21 (1794); Lehm.
Asperif. i. 25 (1818). Tournefortt sessilifolia Poir. Encyc. v. 360
(1804). Heliotropium anchusaefolium Poir. Encye. Suppl. iii. 23
(1813); Cham. Linnaea iv. 458, t. 5, fig. 4 (1829); Arechav. FI.
Uruguay iv. 173 (1911); Herz. Medel. Rijks Herb. xlvi. 10 (1922);
Prain, Bot. Mag. cxxxix. t. 8480 (1913). Heliophytum anchusae-

anc ia Giirke i
. iv. Abt. 3a, 97, - 39n (1893). Heliophytum lithospermi-
‘DC. L. c. 554 554. ropium anchusifolium, var. lithospermi-

poe Griseb. [PL Re 184] Abh. K. Ges. Wiss. Gottingen xix.

| = (1874) and [Symb. Argent] I. e. xxiv. 271 (1879). Heliotropium
ok ifolium Speg. Contr. Fl. Sierra Ventana 44 (1896). Helio- —
phytum anchusaefolium, var. -latifoliwm De. L e. 554. - eames

22 JOHNSTON

anchusifolium, var. latifolium Kuntze, Rev. Gen. iii. pt. 2, 205 (1898).
Cochranea angustifolia var. latifolia Hicken, Apuntes Hist. Nat.,
Buenos Aires ii. [Chloris Plat. Argent.] 194 (1910). Hidlisslgius
anchusaefolium, var. angustifolium DC. |. ec. 554. Heliotr opium
anchusifolium, var. eS Griseb. [Symb. Argent.] l. c. xxiv.
271 (1879); Kuntze, 1. 205. age pe anchusifolium, var.
latifolium, f. simaidlocaen. Kuntec, l. c. 205. Heliotropium anchusi-
folium, var. latifolium, f. parviflorum Kuntze, |. ¢. 205. Heliotropium
anchusifolium, var. angustifolium, f. grandiflorum Kuntze, |. c. 205.
Heliotropium anchusifolium, var. angustifolium, f. medium Kuntze,
l. c. 205. Heliotropium bolivianum Rusby, Mem. Torr. Bot. Cl. iv.
225 (1895). Heliotropium montevidensis Arechav. Anal. Mus. Nace.
Montevideo, ser. 2, i. 66 (1911) and I. ec. ser. 1, vii. [Fl. Urug. iv.]
176 (1911). Heliotropium semiamplexicaule Larrafiaga, Escritos
[Inst. Hist. y Geogr. Urug.] ii. 73 (1923).
i“<¢ f) Tirncnaw i +k. Rp ‘7

,horthern and central

ARGENTINA. Burnos Arres: Naporta Chica, Bahia —— Claraz
119 in (K); Bahia Blanca, 1884, where (BM: Sierra Ventana, Lor
299 and 301 (BD); Sierra Ventana, Hauth al (NY); J. A. de Desa. Pi Parodi 7710
(G); A ag v 1892; Kuntze (NY, Trpzs of H. anch, formae mediu

: Pavon, Miers (BM); “Buenos Aires et Monte Video,”
Commerson (NY, ? E . an et um) 5 Aires, Hauthal
643 (NY), Tweedie 383 (G), 23 (NY), Gillies 1317 (K), Bettreund é

Floyer
Koester 339 (BD), Parker (G), Arnott (NY). Menpoza: To
(K, BM). Cérposa: Los Caldenes, Gillies (K); Las Pefias, Lorentz 92 BD);
[eas AG" BBD) Coa Merona teh FAL” BD!
Kurtz 6949 ), Lorene 88, £5, 68 and B08 (BD), Looser 5 and 12 (G):
318 BD). ag Seve Sonagasta, Sierra F; Hieron & Niederlein

) , Hosseus gee (BD); —. 1600 m., Hosseus 538
(BD). Chr ATAMARCA: 1075 (G, US); Rio del Valle,
Castillon 1096 BD) Valle del astillon 1093 (G); Capillitas, Schicken-
dantz 187 (BD); Siguil, Schunk 127 (BD); Yaeu tula, J 176, 239
and 245 ( if oe ond D).

| Poser: Juluy, Sept 1802,

= of H. anch
> illo i en pate Taft del Valle oe age 197 (G);
, Castillon 372a (G); Raco, 1500

STUDIES IN THE BORAGINACEAE 2

BOLIVIA. Tanrua: SS. Fiebrig 2052 (BD). Cocua-
— Cochabamba, B 20 (NY, nc G US, K, BM, BD, tsorypgs).
NDEFINITE: Bolivia, Fag oe

URUGUAY: Ramirez, Hert 3668 (G); Atahualpa, Herter 79759 (NY,
FM); Arequita, Herter 81269 “G); Pentarol, Herter 68106 (BD). fe Paz,
Canelones, Filippone 4581 (K); Marincho, Osten 3227, 3228 (BD); Mercedes

egro, 3080 (BD); Montevideo, Gilbert 20 (K), Re a —
(KBD), pr posi ae are 3050 (G, photo of authentic material o:
BRAZIL. Rio GranvE po Sut: Nova Wiirttemberg, Bornmiller 606
(BD); Rio Grande, 1837, "Tuccdie (K). Parand; 858, Gilbert 20
(K). ae Pavto: Sao Paulo, Glaziou 19679 in pt. ride: BD) and 19680 Brasil”
BD). em: Botafogo, Schenck 1370 (BD). InpEFINITE
Sellow iB D).

An extremely variable species as to width of leaves, character of the

indumentum on the herbage, and in the size and color of the corolla.

4. H. phylicoides Ch. am. feibes | iv. 460 (1829); Griseb. [Symb.
Argent.] Abh. K. Ges. Wiss. Géttingen xxiv. 271 (1879); Giirke in
E. & P. Nat. Pflanzenf. iv. Abt. 3a, 96 (1893); Arechay. FI. Urug.
iv. 174 (1911). Heliophytum phylicoides DC. Prodr. ix. 554 (1845);
Fresen. in Mart. Fi. Bres. viii vill. pt. 1, 46 (1857); Glaziou, Bull. Soe.
Bot. France lvii. Mem. 3, 479 (1910). Heliotropium phylicoides f.
luciliaefolia Arechav. |. c. 175.

ARGENTIN Concepcion del Uruguay, 1877, Lorentz
(G, US, BD) a 1112 ii? (Dy ‘Gobcareion del Uruguay, Lorentz 58 and 37. 374
B 3 Concepeis 1 Uruguay 84 (G); Parada Chaviy, 1920, Denis
(G). Mistonzs: between San gas Arroyo F , Niederlein 1748

BD). Buenos Ares: “Buenos Ayres,”
: Phvdt Ne in, 1926, Herter cee “BD); a Oe Lucia, Gilbert
BRAZIL: S&o Pauto: Serra do Cubatao, ee 19681a_ ig In-
muvrnrre: Brasil. merid. Sellow UX) BD, isotrres of H. phylicoides
A ch laine of Hampi and pai ay ae
Ais Sa ena

15. H. nicotianaefolium Poir. Enc thong a iii. 23 (1813);
Kuntar Rev. Gem it pt 2, 205 (808); Hick puntes Hist. Nat. _
Buenos Aires ii - [Chlor. Plat. Argent] 194 0810; Molfino, Physis —

24 JOHNSTON

vii. 179 (1924). Heliophytum nicotianaefolium DC. Prodr. ix. 554
(1845). Heliotropium sidaefolium Cham. Linnaea iv. 460 (1829);
Hicken, |. c. 194; Archev. Fl. Urug. iv. 174 (1911); Molfino, |. e. 179.
Heliophytum sidaefolium DC. |. c. 553; Fresen. in Mart. Fl. Bras.
viii. pt. 1, 46 (1857). Tournefortia heliotropioides Hook. Bot. Mag.
Iviii. t. 3096 (1831); DC. 1. ec. 515.

Northern RE southern Paraguay and eastern Bolivia; also
reported from Uru

ARGENTINA. Buenos Arres: ‘Buenos Aires,’’ Commerson (BD, 1s0-
type of H. nicotianaefoli um)! ‘Buenos Aires,” herb. Miers (BM). C6rpopa:
Cruz de Ele, Isler 71 (G); Bo rero Arroyo, Cahuayo, Sierra Chica, Kuriz 8760
(NY); between La Calera fot La Puesto de Arriba, Sierra Chica, 1881,
Galander (BD): Cérdoba, 1880, Galander (BD); (?) Dique near Cérdoba,
Dee. 1891, Kuntze (NY gal Bd Fe: Ocampo, Venturi 207 (G). Tucum4n:

1 4h (
Rio Campo Blanco, Oran, 500 m., Schreiter 3606 (G); Salta, Nov. 1892,
Kunize (NY, BD). ” Jurur: Jujuy, 1200 m., Lille 5255 (G).

BOLIVIA. Santa Croz te de Santa Cruz, 450 m., Steinbach 1224
(BD); Chacras de Buena vista, ore m., Steinbach 1270 (BD).

PARAGUAY: Sud "Paraguay, Sept. 1892, Kuntze (NY).

BRAZIL: “‘Brazil,’’ Sellow (BD, 1soryprs of H. sidaefolium).

This is an obvious relative of H. amplexicaule from which - Gibers
in its looser habit and definitely petiolate leaves. From its other
close relative, H. leiocarpum, it differs in ae a more sou y
and westerly range and in having the leaf-blades eae rather
than abruptly contracted into the petiole. It is quite possible that
future study of more copious material will reveal reasons for treating
the present species as only a variety of H. amplexi
- 16. H. leiocarpum Morong, Ann. N. Y. Acad. Sci. vii i 168 (1892);
Chodat, Bull. Herb. Boiss. ser. 1, vii. suppl. 1, 79 (1899) and l. ec.
ser. 2, ii. 816 (1902) and 1. c. ser. 2, v. 483 (1905); Molfino, Physis vii.
179 (1924). H. leiocarpum, f. minor Chodat, |. c. ser. 2, ii. 816
(1902). H. leiocarpum, f. albiflora Chodat, |. c. ser. 2, v. 483 (1905).

Paraguay and adjacent northeastern Argentina.

Posadas, 1900, — (G); Santa Ana,
eter

ane GQ); = inet Puerto Monteaguda and San Pedro, Ni
Tue BD); 1 (G); between Santa Ana, 1884, no collecor given 2096
PARAGUAY: Villa Encarnacién, Beitfreund 21 6 (BD); Cord. Altos,
__ Fiebrig 72 (G, K, BD) Hassler 3087 (G, K, BM, BD); Cord

6188 (BM, ; Lacus Ypacaray. Hassler 11704 704 (G, US, BM, BD); Loma

4 ached Fror , Balansa ,
Alta Parand eased ts, ca BM, BD); ~~ Sept. srg

Rojas 2619 (G A, : : Sas Bernardino, — 382

STUDIES IN THE BORAGINACEAE 25

is readily recognized by its leaves, which are long-
sollte and i a blade abruptly contracted to form a rounded, or

more commonly, a cordate base. It is very clearly related to and
probably intergrades with H. nicotanacfolium, but appears to range
to the north of that species. It the common member of the

section Heliophytum in Paraguay. The forma albiflora is based upon
Hassler 6148. I have seen no material of this collection.

VIII. Section Cochranea (Miers) Reiche, Anal. Univ. Chile exxi
234 (1907). Cochranea Miers, Trav. Chile. ii. 529 (1826), nomen;
and Ann. & Mag. Nat. Hist. ser. 4, ii. 124 (1868); type-species, C.
conferta Miers. Meladendron Molina, Saggio, ed. 2, 143 (1810);
type-species, M. chilense Molina.

This section is known only from the coastal hills of northern and
central Chile and = age Peru. It is closely related to the section
Heliophytum, which occurs east of the Andes, differing from it pri-
marily i in having the corolla glabrous inside. The fruit of Cochranea
is superficially very similar to that of Heliophytum, but does not have
the infertile cavities which are frequently present in the latter section.
In habit Cochranea is definitely a shrub. Heliophytum is at most an
undershrub. The section Cochranea is of particular interest because
of the extreme localization of its species. As would be expected the
plants are in need of further field-study.

Key To Species
Leaves sinuate, more or less rugose above with conspicuous
veins.
—- of lage simple; corolla 7-11 mm. broad; tube
PSO Set er Man pctnine | 17. H. sinuatum.

ae mm. long, densely :
st tomentose; corolla 4-6 broad with tube
SS a H. taltalense.
Oe ealyces 2-3 mm. long, sparsely ; corolla

lo hirsute;
mm. broad with tube 2.5-3 mm. long...... 19. H. Krauseanum.
Leaves pat aagea not rugose a

26 JOHNSTON

does ge us-tomentose; stigmas 1.5-2 mm. Be
very agg ao eloped Se Tae Ae . sclerocarpum.

Plant evidently strigose or villous, not glandular; corolla

to ed tne WO cv 23. H. mysotifolium.
Plant - emgggenad or puberulent, or mages and obscurely

strigose, ewhat glandular; corolla 7-9 mm

Stigma practically sessile; leaves a 4-10 mm.
long, spatulate-linear, apex rounded........ 24. H. filifolium.
Stigma on a presi strle; leaves lanceolate to lance-
ear, 2-6 cm. 2 ages
note evidently ‘an abundantly scree
; leaves lanceolate, 3-8 mm
25. H. glutinosum.
—— sage sos _ ot if at all glandular;
ce-linear, 1-3(-4) mm.
oe ee ee eee oe 26. H. stenophyllum.
Stigma usually shorter than the style, rarely porgry
longer, varying between EAs aoe 25 beg of le (cf.

SS a eae he ces Gee strigose,
epidermis usually co ith resinous secretions and

-margins lia 6-9 mm. broad.
Leaves obtuse, 1. 5-3 cm. long, finely short-strigose;
ps exserted.. 27. H. huascoense.
Leaves acute, 3-6 em. rio iehoes ; corolla 7-9
mm. ‘broad; cateaetie melded. ...... 28. H. longistylum.
pustulate and ciliate; corolla 9-12 mm.
BU SEG I Pa I OE ig ae a, 29. H. megalanthum.
Plant coarsely and obviously’ strigose, not resinous nor
Leaves spathulate, somewhat fleshy, with very y
revolute margins and hence q a oe

or semiscandent, 1-2 m. tall; corolla 5-8

Z 31. H. Philippianum.
Shrub decumbent, 3-6 dm. tall; corolla 8-10 mm.

‘Corolla oraupe-colored. -. 6s. linariae
Corolla white (or ? purplish). . : - ts ign einen
17. H. sinuatum (Miers), comb. nov. Cochranea sinuata Miers,
Ann. & Mag. Nat. Hist. ser. 4, ii. 127 (1868) and Contr. Bot. ii. 196
(1869); Ph. Anal. Univ. Chile xe. 342 (1895). Heliophytum floridum,
var. Bridgesii DC. Prodr. ix. 553 (1845). Heliotropium floridum,
var. Bridgesii Clos in Gay, Fl. Chile iv. 457 (1849). C. conferta, var.
‘Miers, Ann. & Mag. Nat. Hist. ser. 4, ii. 126 (1868) and
‘Comte: Bot. ii. 195 (1869); Ph. corre pes Chile xe. 339 (1895).
Heliotropium rosmarinifolium Ph. niv. Chile xliii. 514 (1873);

Tae cal cay aes 2 a a een
Shrub erect
Sted ok

—

a

STUDIES IN THE BORAGINACEAE 27

not Heliotropium rosmarinifolium Bertero, nomen (1845). C. ros-
- Ph. Anal. Univ. Chile xc. 349 (1895). Heliotropium
zagae Ph. Anal. Univ. Chile xc. 355 (1895).

serch ete Chile; southern and western parts of the province of
Atacama in the departments of Copiapé, Vallenar and Freirina, lat.
27° 45’-28° 40’ S.; picts probably erroneously, from Coquimbo.

CHILE. Atacama: urrias, Geisse (MS); ge of ATacaMA, Geisse
173 (NY); Carrizal baj an 1885, Philippi (MS, B M); Chafiarcito near

arrizal, Sept. 1885, Philippi (MS, Type of H. I Izagae; G, photo; BD, 1so-
TYPE); os sa de Huasco, 883, lippi
(MS, Tyre of H. rosmarinifolium; G, photo; BD, IsOTYPE); Mina A bo,
Vallenar, 1000 m., W. nn 145 (G, FM, K, ee dry hills and valleys
between Huasco and Copiapé, a 1342 (BM, hoto.) INDEFINITE:
without locality, Bridges (BM, TrPE of Me sinuata; G, yeaa Bg corer
BMD (K); “ N. N. Chile.” Lobb 442 in t. (K, TYPE Pe oes auriculata

A dark green, erectly branched shrub 1-2 m. tall, growing on dryish
rocky hillsides. The corolla is white. The two collections upon
which H. sinwatum is founded were collected by Bridges and given
as from Coquimbo. This geographical data is obviously incorrect.
A study of Bridges’s plant-list shows that his number 1342 came from
between Copiapé and Huasco, and not from Coquimbo or, as one
finds in Hooker’s Herbarium, from Concepcion. Bridges unnumbered
collection which is the type of the species I believe to have been
material of his number 1342 that became igs from its data.
Lobb’s number 442, the type of C. conferta auriculata, given
from “‘N. Chile,” at Pearce’ 8 collection Be “Coquimbo, sg probabl ly

The type collection of H. sinuatum is a form with large leaves

(ad 3.5 em. long) and very pubescent calyces and under leaf-surfaces.

se respects it suggests the common phases of H. taltalense. It

has, however, the simple leaf-veins and large flowers characteristic of
the Vallenar-plant.

18. H. taltalense (Ph.), comb. nov. Cochranea ce aaraa Ph.
Anal. Univ. Chile xc. 349 (1895). H. rugosum Ph. nae 38
ing and Viage Des. Atac. 20, 24, 25 and 212 (1860); Reiche, Anal.

. Chile exxi. 239 (1907) and Fl. Chile v. 197 Gaen sink i.
. Mart. & Gal. (1844). ©. rugosa Ph. Anal. Univ. Chile xe.
351 (1895).
- Northern Chile on the coastal hills of the Dept. of Taltal, lat.

30’ S.

24° 30’-25°

28 JOHNSTON

CHILE. Anroracasta: Taltal, Oct. 1889, Darapsky 30 (MS, Type of
Cc. taltalensis; G, photo.); Taltal , Dee. 1888, Borchers (BM); Taltal, 200 m.
837 ( K i of Tal

G, photo.; BD, tsoryre); arid shrubby crests, E] Rincon bind: poso, Johns-
ton ee (G); ridge above Aguada de Panulcito, Johnston BANG PG: rocky
da Cardon, 2 eetethe §293 (G); dry rocky crests, Migue
Y Jekiaion 5414 (GQ).

ibcsatl evidently a close relative of H. sinuatum, this species is
unquestionably distinct, differing as it does in its smaller corolla,
longer corolla-tube, forked leaf-veins and detached very natural
range. It is an erect dark green bush 1-2 m. tall growing on the
dryer, usually u upper slopes of the coastal hills. The corolla is ae
with a little yellow in the throat and occasionally with a rose-tinge
outside.

19. H. Krauseanum Fedde in Just, Bot. Jahresb. xxxiv. pt. 3, 72
(1908). H. sazatile Krause, Bot. Jahrb. xxxvii. 633 (1906); not
Brandg. (1905).

Coastal hills of extreme southern Peru and adjacent Chile.

PERU. Arequipa: og Johnston 3533 (G); Mollendo, Weberbauer
1552 (BD, TYPE; G, photo.); between Atiquipa and the port of Chala, 250 m.,

88 ( BD).
vOHe shes NDEFINITE: “Cobija, Iquiqui et Arica,’’ 1831, Cuming 955
(K, BM; G, photo.).

A very interesting species finding its relatives agg the shrubby
Chilean species

it is separable by its more slender, more remotely flowered spikes,
slightly smaller corollas, smaller less densely pubescent calyces, and
smaller fruits. Possibly future study of more numerous collections

and inaccurate labeling. Cuming’s printed label and his plant-lists
give the data I have cited above. ;

20. H. inconspicuum Reiche, Anal. Uniy. Chile exxi. —
and Fl. Chile v. 203 (1910). Cochranea parviflora Ph. Anal.
: ~ xe. 350 A omuiide? not — —_—- L. (1771).

coast in the southwestern
the province of Antofagasta and — — lat. 25° 20

STUDIES IN THE BORAGINACEAE 29

AntoraGasTa: Breas, Dept. Taltal, Larrafiaga (MS, Type; G,

A loose a bush 6-10 a tall, growing in dry rocky BO or
on sunny slopes of quebradas. The species is quite distinct from the
members of the H. chenopodiaceum-group which are its closest
relatives. The style in relation to the stigma and column is much
longer than in the group of H. ——— and the linear or
rarely spathulate-linear leaves are quite characteristic because of
their definite petioles and evidently orien tees undersurfaces.
The small corolla is white.

21. H. chenopodiaceum (DC.) Clos in Gay, FI. Chile iv. 458
(1849); Reiche, Anal. Univ. Chile exxi. 243 (1907) and Fl. Chile v
201 (1910). Heliophytum chenopodiaceum A. DC. Prodr. ix. 553
(1845). Cochranea chenopodiacea Miers, Ann. & Mag. Nat. Hist. ser.
4, ii. 132 (1868) and Contr. Bot. ii. 201 (1869); Ph. Anal. Univ. Chile
xe. 348 (1895).

Var. genuinum.—Plant glabrate, at — usually puberulent only
in the inflorescence.—Synonymy given ak

Northern Chile back from the coast, lat. ‘oF 30’-30° S.

ogy en vu Fe ve Rivadavia, Johnston 6260 (G); ass
eiche (MS); Tres + mal 1871, ” Phitzppe (MS); j just above Guanta,
yi 6232 Gy ‘Saruno Serena ete., ’ Oct. 1836, Say “Ms): =
» mene Philippi (Us); indefinite, Gay (G) and Gay 294 ( ko.
Transito, Johnston 5884 (G); Yerba Buena, Sept. 1885, philips otis):
Yerba ge Haske Oct. 1871, King MIS, K); Yerba Buena, Reed (G); Queb.
Conchas above Cerrillos, Johnston 4979 (G).

Var. ericoideum oo Reiche.—Plant strigose throughout with
short stout usually curved hairs.—Anal. Univ. Chile exxi. 244 ae )
and Fl. Chile v. 202 (1910). Cackionss ericoidea Miers, Ana!
Mag. Nat. -~ ser. 4, ii. 130 gees and Contr. Bot. ii. 199 pete
Ph. Anal. Univ. Chile xe. 344 (1895). Heliotropium Pearcei Ph.
Anal. Univ. Chile xe. 352 (1895). Eritrichium glabratum Ph. Cat.
Itin. Tarapacdé 56 (1891).

Northern Chile back from the coast, lat. 24° 50’-27° 50’ S.; re-
ported from Coquimbo, but no doubt incorrectly so.

Coqumpo: ‘“Coquimbo,”” Pearce (MS, tree of H. Pearcei).

Aracama: Chaheruillo, Sept. 1885, Philipp: (MS, BM): Queb. Serna, 1885,

MS); Queb. Puquios, 1 S); . Puquios, 1885,
i vy Co 7 1887, Gigoux (G); Pabellan, 1885,
(MS); “mountains c. : of Co) ,”’ Bridges

ohnsion "5003 (G): Tierra Amarilla, Werdermann 399 (G, FM,
: ot 1m, Gros BS) Desert of Atacams, Morong

30 JOHNSTON

1183 Bigs baits = and 1188 (NY). AnToracasTa: high dryish hie =
Johnston 5545 (G); ridge above Aguada Pan
Jane 5477 ? « ).

A slender-stemmed, usually loosely branched bush 6-10 dm. tall.
The epidermis is more or less resinous. The plant grows in arid
situations on rocky points, cliffs or dry gravelly stream ways. The
~~ corolla is white.

species is very closely related to H. sclerocarpum and H. myoso-
Tess and perhaps should be extended to include them. These
ose relatives, however, seem to have larger flowers and a character-
istic habit even though the variation of characters is such that I have
been unable to discover any invariably stable and hence universally
diagnostic sa a by which the three species might be decisively
separated. It is perhaps significant to note that the other members
of the section Cochranea are all sharply, readily and obviously de-
imited. It would seem that the H. chenopodiaceum-group is an
exception or that I have inconsistently persisted in segregating it
more than the actual state of speciation justifies. The true condition
can only be determined by field studies.
var. ericoideum is merely the coarsely strigose form of the
species which appears to be more northerly in range than the typical
glabrate forms. The type of C. ericoidea has an indument of coarse
appressed falcate hairs but differs from most of the material referred
to the var. ericoideum in being a more slender plant with more herba-
ceous slender leaves that are not so broadly attached as in the plant
classified with it. Plants = similar to the type of C. ericoidea are,
Morong 1342 and Johnston 500,

It is to be noted that AS aE established his H. chenopodia-
ceum on a collection of Gay and upon Bridges’s number 1342. Miers,
however, restricted the name to the species represented by Gay’s
collection. Bridges’s number 1342 is the original collection of the
es epee H. filifolium.

H. sclerocarpum Ph. Anal. Univ. Chile xliii. 515 (1873).
Pc selerocarpa Ph. |. c. xe. 351 (1895). H. chenopodiaceum
_ var. sclerocarpum Reiche, Anal. Univ. Chile exxi. 244 (1907) and Fl.
. Chile v. 202 (1910).
Northern Chile near the coast, lat. 28° 30’ S.
CHILE. — F Oct. 1 Philt MS,
e a . Dept. Freirina, 366, ppt (
‘This species ie known only from the type-collection. It is a loosely
branched shrubby plant 1.5-2 dm. tall and is. villous-tomentose

ae _ throughout. a —— os

_

[ep] o= Jon

STUDIES IN THE BORAGINACEAE 31

| are 6-15 mm. long and 1-2 mm. broad, have rather strongly revolute

margins and are sessile by a broad base. The corolla is white and
about 4 mm. broad. The stigma-column, which is borne on the very

| well developed disk, is 4-6 times as long as the style. The stigma-

column is ca. | mm. long.

23. H. myosotifolium (A. DC.) Reiche, Anal. Univ. Chile exxi.
243 (1907) and FI. Chile v. 201 (1910). Heliophytum stenophyllum,
var. mysotifolium A. DC. Prodr. ix. 552 (1845). Heliotropium steno-
phyllum, var. mysotifolium Clos in Gay, Fl. Chile iv. 456 (1849).

| Cochranea mysotifolia Miers, Ann. & Mag. Nat. Hist. ser. 4, 0. 128

(1868) and Contr. Bot. ii. 197 (1869); Ph. Anal. Univ. Chile xc. 343
(1895). C. hebecula — c. 130 and |. c. 199. C. hispidula
Miers, 1]. c. 132 and I. ec. 201; Ph. |. c. xe. 347 (1895). Heliotropium
hispidulum Reiche, |. ec. _ 243 nage c. 201. C. sentis Ph. |. ec. 351.
Heliotropium canum Ph. 1. ¢. 356. Heliotropium hispidulum Ph.

ic é ‘
Northern Chile back from the coast, lat. 27° 10’-29° 40’ S.
Atacama: Bandurrias, 1885, Geisse (MS); Desert of Atacama,

ic ee tall. It is branched from the base
and usually (not always) has simple strict branches 1-4 dm. long.
The plant is strigose or villous, frequently very densely so. The
corolla is white with paboauls Widaak and is about 7 in di-
ameter. The stigma-column is 1.5-2.2 mm. long above the disk. The
style is 0.5-0.2 mm. long. The leaves are 1-2 cm. long, 2-3(-4) mm.
broad, and ane Bikes or oblomp knekr ix outone. The lower leaves
tend to be oblanceolate and to be more or less definitely petiolate.

The species is very closely related to H. chenopodiaceum, differing

in its usually larger, more herbaceous, whiter corollas, usually larger

leaves (the lower ones being somewhat oblanceolate and petiolate),

: None of these characters is invariably crucial although the species
A careful

does seem to be, for she mont part, readily recognizable.

field-study is needed before this species can be well malicatond

Se JOHNSTON

Heliotropium mysotifolium is taken as based upon —— 1338
which has been incorrectly stated to come from “‘prov. Coquimbo.”
DeCandolle who first described the plant considered it no more than
a variety of H. stenophyllum distinguished by its “sericeo-incanis”’
indument. Miers who also cited Bridges 1338, and apparently
took his specific name from DeCandolle, considered the plant to be a
distinct species. Miers, however, established his C. hebecula also
upon a collection by Bridges. The types of C. mysotifolia and C.
hebecula are so similar in details that they might well be parts of one
collection, as I believe them to be. The type of C. hebecula appears
to be merely a specimen of Bridges 1334 which has become dissociated
from its num

There is a rather conspicuous range in the abundance of pubescence
among the specimens cited above. One extreme is densely villous,
almost tomentose. Thetypeof H.canum issuchaform. The other
extreme, represented by the types of C. sentis Ph., H. hispidulum
Ph. and C. hispidula Miers, is rather sparsely strigose or appressed
villous and not densely villous. These two extremes I had intended
to treat varietally, but upon discovering that the types of C. myosoti-
folia and C. hebecula were quite intermediate as to the character of
the pubescence, I have thought it hest to refrain from publishing the
varieties under consideration.

24. H. filifolium (Miers), comb. nov. Cochranea filifolia Miers,
Ann. & Mag. Nat. Hist. ser. 4, ii. 131 (1868) and Contr. Bot. ii. 200
(1869); Ph. Anal. Univ. Chile xe. 346 (1895). H. chenopodiaceum,
var. filifolium Reiche, Anal. Univ. Chile exxi. 244 (1907) and FI.
Chile v. 202 (1910). C. Kingi Ph. |. ¢. 350. H. Kingi Reiche,

CHILE. Aracama: Carrizal?,” Sept. 1885. kes (MS, Tree of
C. Kingi; G, photo BD, vs mer ae Mra a between Huasco
and Copiapd, Bridges 1 1345 (BM, rrre of C. fljolia: G, photo eG EM
ISOTYPES) ; Vallensr allenar, 1900, Reiche (BM

This eoRGg eR, It is appar-
sally (dae, tek adc kee coi , finely villous-scurfy
with stoutish curved hairs. Though not glandular it is somewhat —
_‘Tesinous. The ne flowers, according to Bridges, are white. As Reiche

d otable in it: tion because of its broad
sessile or subsessile stigmas. In the material studied the stigma-—
__ colurnn is 0.4-0.7 mm. high and short-cylindrical or frustoid, being
- oe irene i we: 68 :

remit iie

STUDIES IN THE BORAGINACEAE 33

developed disk which in turn is apparently seated directly upon the
ovary.

25. H. glutinosum Ph. Fl. Atac. 38 (1860) and Viage Des. £
212 (1860); Reiche, Anal. Univ. Chile exxi. 242 (1907) and FI. ret
v. 200 (1910). Cochranea glutinosa Ph. Anal. Univ. Chile xc. 349
(1895).

Northern Chile ia the mountains of the Dept. Chafiaral, lat. 26°-
26° 30’ S.

CHILE. Aracama: Queb. de Salado, San R G, photo.);
ger 1854, Philippi {gad TYPE; G, Fens eyo a Potrerillos,
Johnston 3698 (G); near Rio Sal S n, Queb. de Salado, Johnston ida

(G); rig . 9 ‘las Chiea, — 1886, Gigoux (G); ‘ Atacama, ’
Philippi (BM, BD).

A rare and remarkable species known only from the department of
Chafaral where it occurs in the quebradas in the mountains east of
the port of Chafiaral. The type and most of the subsequent col-
lections have been made in the general vicinity of Agua Dulce near
Potrerillos, at an altitude of about 2000 m. Outside of this area it is
known only from near Dojia Inez Chica about 40 km. to. the north-
ward. The plant forms a small light-green bush 3-10 dm. tall pro-

| ducing many widely spreading or decumbent branches. It grows in
The

gravel of stream-ways or on the talus at their margin.
is puberulent and is conspicuously provided with extremely numerous
glands. The corolla is white with a touch of yellow in the throat.
26. H. ste . & A. Bot. Beechey Voy. 38 Pesce,
Clos in Gay, Fl. Chile iv. 456 (1849); Reiche, Anal. Univ
8 (1907) and Fl. Chile v. 199 (1910). Heliophytum ke
. Prodr. ix. 552 (1845). Cochranea Se sateoang Miers, Ann. &
be Nat. Hist. ser. 4, ii. 128 (1868) and Contr. Bot. 197 (1869).
Meladendron chilense Molina, Saggio, ed. 2, 143 (1810); Ph. Anal.
Univ. Chile xxii. 714 (1863); not Heliotropium chilense Bertero es
Hehotropium rosmarinifolium Bertero ex Steud. Nom. ed. 2, i. 7
(1840), nomen. Heliophytum stenophyllum, var. rsmarinolio
DC. Prodr. ix. 552 (1845). Heliophytum rosmarinifolium
ex DC. Prodr. ix. 552 (1845). Heliotropium stenophyllum, var.
rosmarinifolium Clos in Gay, |. c.; Reiche, |. c. 242 and lL. ec. 200.

_ C. conferta erta Miers, Trav. Chile ii. 529 (1826), nomen; Miers, Anal. &

Mag. Nat. Hist. ser. 4, ii. 125 (1868) and Contr. Bot. ii. 194, t. 53
(1869). C. congesta Miers, Anal. & Mag. Nat. Hist. ser. 4, ii. 126
(1868) and Contr. Bot. ii. 195 (1869), lapsus ?

Geel Coe mtn meraaie of, Yoleeoren, Aconcagua and
foqninho, lat. 30°-33° S.

34 JOHNSTON

CHILE. Vaupraratso: Vifia del Mar, 1863, no collector given (US); Calera
Quillota, Bertero 1042 (G, NY, FM, BM, 1sotrpss of H. rosma: marinifolium m);

esta ota M
, Cuming 377 (G, K, BM); rig —o. forse a ACONCAGUA:

a de Ocampo, Sept. , Sept. 1
Reiche (MS); Illapel, na 1862, Phileppi rary BM, BD): Tilapel, Rose
19346 (NY, us); “Cumbre, Andium Cl m, 1825,’ acrae (K). Co-
ea : :

Hare (G K), Ball (G:, NY, K), Gaudichaud 64 (FM, BD), Werdermann
te BM). Coppinger (K), Macrae (K), Bridges (K).

A very distinct species well known since it grows near the ports of
Valparaiso and Coquimbo. In the collections at Kew there is a col-
lection by Caldcleugh given as from “Conception” and one from
Reed given as from “Yerba Buena.” These are, no doubt, incorrectly
labeled as is also Macrae’s specimen which is said to come from
“Cumbre.”

27. H. huascoense, sp. nov., fruticosum resinosum sparsissime [
glanduliferum breviter strigosum ; baie paullo congestis anguste spath-
ulatis 1.5-3 mm. longis 1-2 mm. latis, margine revolutis, apice rotundis,
basem versus gradatim attenuatis; paniculis terminalibus corymbosis,
spicis 4-5 ca. 3 em. longis; calycibus subsessilibus ca. 1.5 mm. longis
fere ad basem in lobos lineares ascendentes incisis; corolla alba 6-7
mm. diametro, tubo 2-2.5 mm. longo calycem superanti, lobis obtusis,
antheris paullo extrusis; stigmate anguste conico ca. 1 mm. longo;
stylo 1-1.3 mm. longo.

Northern Chile near the coast, lat. 28° 30’ S

CHILE. ama: Huasco, 1920, Lopez (G, TPE); Punta del
1889, Vidal MS). i INDEFINITE: ““Coquimbo, ” Pearce in pt. i: ey.
Chile,” Lobb 442 in pt. (K).

This species is clearly related to H. stenophyllum, but differs in its
obtuse narrowly spathulate leaves, smaller corolla, shorter style, as
well as more northern range. It is with certainty known only from
about the port of Huasco. The collection made i bdo was deter-
mined by Philippi as H. hispidulum. The specimens 1 by
Pearse and by Lobb almost certainly came from Huasco or near there,

for both are a mixture of H. huascoense and H. sinuatum.
eee ee

.

Anal. Univ. Chile xlii. 515 (1873);

m Ph.

: ert parE Chile exxi. 240 (1907) and FI. Chile v. 198 (1910).
oe a <c. —. H. vernicosum Ph. I. ec.
cs — a

STUDIES IN THE BORAGINACEAE 35

Northern Chile near the coast, about lat. 28° 5’ S.

CHILE. Ara a a Bajo, Sept. 1885, Philippi (MS, Type of
vernicosum; G, Sees ISOTYPE); Carriza i Bajo, hae 1871, King
(MS, TYPE of H. rewards , photo.).

Apparently a low decumbent shrub 3-4 dm. tall, which is glabrate
or sparsely strigose in the inflorescence. It is past aed glandu-
lar. The leaves are numerous, narrowly oblanceolate and acute,
being 3-6 cm. long and 2-6 mm. broad. They are very ay in texture
with concolorous faces and narrowly revolute margins. The corollas

29. H. megalanthum, nom. nov. Cochranea corymbosa Miers,
Ann. & Mag. Nat. Hist. ser. 4, ii. 126 (1868) and Contr. Bot. ii. 195
(1869) ; Ph. Anal. Univ. Chile xc. 340 (1895). H. corymbosum Reiche,
Anal. Univ. Chile exxi. 242 (1907) and Fi. Chile v. 200 (1910), not

. corymbosum R. & P. (1799). H. crassifolium Ph. Anal. Univ.
Chile xliii. 515 (1873); Reiche, 1. c. 240 and |. c. 198, not H. crassi-
folium Boiss. & Noé (1856). C. crassifolia Ph. 1. ec. xc. 349

Northern Chile in southern part of the province of Atacama, ca.
lat. 28° 30’ S., incorrectly reported from Coquimbo.

CHILE. AtacaMa: Huasco, Sept. 1900, Reiche (MS); Huasco, Oct. 1866,
Philippi (MS, tree of H. crassifolium; G, photo.; BD, sac eapsbe Chafiarcito,
Sept. 1885, Philippi (G, US, i, BM, BD, MS MS); dry valleys and hills between
Huasco and Copiapé, Bridges 1 (BM, of C. corymbosa; G, photo.;
G, K, Shae Rea IND: ear C wi ay ” Pearce 40 (K).

A striking plant, with certainty known only from the department
of Freirina, Chile. It is apparently a weak spreading or decumbent
bush 1.5-3 dm. tall which is glabrous or sparsely strigose on the stems.
The leaves are broad and coriaceous, being oblanceolate to ovate-
spathulate, 2-3.5 cm. long and eit Reid mm. broad. They have an
obtuse or rounded apex and a contracted base. The surface of the
leaves is quite ikea but the ae are ciliate and pustulate.
The corolla, according to Bridges, is white
30. Heliotro pycnophyllu Ph. FL. Atac. 38 (1860) and
Viage Des. Atac. 15, 17, 18 and 212 (1860); Reiche, Anal. Univ. Chile
_ exxi. 238 (1907) and FI. Chile v. 196 (1910). eon ege pyenor
_ Ph. ex. Reiche, 1. c. H. breanum Ph. Anal. . Chile xe.
(1895). _H. = Ph. Anal. Univ. Chile xe. e357 gene
Northern Chile near the coast, lat. 24° 30/-26° 30

to near Chafiaral, Johnston 4807 (G); Cachinal
de | ie Con Costa, sag 1353, Philip ippi (MS, Tree of H. pycnophyllum; G, photo.);
; Bi ). Anroragasta: Taltal, 100 m.

Johnston be ga
Werdermann 849 ( FM, K, BM); Quebrada de San ni
Teltar Johnston 5155 S&S Hueso Parado, Oct. ——— Borchers (MS, TYPE of

:
’
‘
é

36 JOHNSTON

H. brevifolium; G, photo.); B wn ned aR (MS, Typs of H. breanum;
G, p ag a); Punto Colorado, Sociaien 5242 (G); Des ney Pohang Philippi 703
(BD :

A very distinct and readily recognizable species of the dryish
plains and hills of the western parts of the departments of Chajiaral
and Taltal. The corolla is of a somewhat sordid white with a touch
of yellow in the throat. At extreme maturity before drying up a
corolla becomes more or less tinged with purplish or violet.
corolla is most certainly not orange-colored as originally given by
Philippi and subsequently accepted by Reiche. No doubt Philippi,

Jiage Des. Atac. 14 (1860), confused his notes regarding H. pyenophyl-
lum and H. linariaefolium, for as I have checked by numerous field
observations, some even at the type station (Aguada Grande, i. e.
Cachinal de la Costa of Philippi), the latter species has conspicuously
orange corollas and the former has white or purplish ones as I have
described.

Heliotropium pycnophyllum is a pronounced xerophyte and usually
forms very dense globose bushes 5-10 dm. tall. The leaves are
thick, almost terete when fresh, a trait due partly to their somewhat
fleshy texture and partly to their strongly revolute margins. It is a
very difficult plant to dry, in this — much suggesting the various
Nolanaceae with which it associates.

31. H. Philippianum, sp. nov., fruticosum erectum vel sub-
scandens 1-2 m. altum ascendenter ramosum; foliis firmis planis vel
margine paullo revolutis oblanceolatis vel spathulatis 1-3 em. longis
1-6 mm. latis strigosis apice rotundis vel obtusis basem versus in
petiolum gracilem gradatim attenuatis; paniculis terminalibus
dichotome ascendenterque ramosis a ebracteatis; calycibus
eylindricis 4-5 mm. longis 1.5—2 mm. crassis sessilibus, lobis linearibus

strigosis; corolla fragranti 58 mm. diametro alba, faucibus
abrupte dilatatis evidente aurantiacis, tubo ca. 4 mm. longo sepalis
subaequilongo; antheris inclusis vel non rariter apice paullo extrusis;
stylo gracili 2-2.5 mm. longo; stigmate 1.5-2 mm. longo bifido colum-
nari basi in annulum stigmaticum angustum incrassato; nuculis laevi-

bus.
Northern Chile on coastal hills, lat. 24° 30’-25° S
CHILE. AnrToracasTa: Paposo, Philippi (MS; G, photo) 0 back of
Punto Grande near Paposo, Johnston 5233 (G, TYP: E); —-
; demnelen, BE 78 ( (G);-Aguada Cardon, Sok ohnston 5. 294 (G); Aguada
Miguel Diaz, Johnston 5415 (G); Miguel Diaz, Dee. 1853, Philippi (MS; G,
photo.); Des. Atacama, Philippi i 708 in pt. (BD).
A shrub frequent in the fertile, fog-bathed middle-slopes of the
coastal hills of the Department of Taltal, from Paposo to Miguel

STUDIES IN THE BORAGINACEAE St

Diaz, in which area it seems practically to replace H. linariaefolium.
It is usually a large bush and, often leaning on or growing through
other shrubs, reaches a height of one to two meters. The corollas
are white with a conspicuous orange center and the flowers are fragrant
with the odor of the common garden heliotrope. The species was
first collected by Philippi, who obtained material from near Paposo
and Miguel Diaz, but who failed to distinguish it from the decumbent
plant with larger orange-colored flowers which he obtained at Cachinal
de la Costa and to which I have restricted the name H. linariaefolium.

32. H. linariaefolium Ph. Fl. Atac. 38 (1860) and Viage Des
Atae. 12, 15, 18 and 212 (1860); Ph. Anal. Univ. Chile xe. 354 (1895);
Reiche, Anal. Univ. Chile exxi. 239 (1907) and Fl. Chile v. 197 (1910).
H. linearifolium F. se Cat. Pl. Vase. Chile. 212 (1881). H. longi-
oe Ph. Anal. Univ. Chile xe. 354 (1895)

Yorthern Chile near the coast, lat. Pe S.

CHILE. AracaMa: Caldera, Sept. 1900, Reiche (MS); Barquito near
Chatiaral Johnston 4760 (G); Las Animas, Dec. 1853, Philipps (MS); Cachi-
nal Costa, Dec. i H. linariaefolium; G

reas, 1888, Larra m; oe
1887, Borchers (MS); Taltal, 200 m. = Werdermann 767 (G, FM, K, BM);
eb Guani t Posado, Dee. 1925, Johnston 5600 (G); Des. Atacama,
Philippi 708 in pt. (BD).

A plant with loosely decumbent stems, forming a low light-green
shrub about 3-6 dm. tall and 6-9 dm. broad. It is readily distin-
guished by its large orange-colored corollas. It is a very handsome
species and merits a place among the cultivated garden _ The
plant usually grows in the coastal hills, where it occurs o
rocky slopes or more commonly along dry gravelly sincera in
the quebradas. Occasionally, with H. pycnophyllum, it is — in
the very arid country back of the hills facing the ocean. It is kno
only from the departments of Copiapé6, Chafiaral and Taltal, extending
- from about Caldera @ Paposo.

ao. A. DC.) Clos in Gay, FI. Chile iv. 457 Laphee
Ph. FL Atac. 38 wad Viage Des. Atac. 10 and 212 (1860);
Anal. Univ. Chile cxxi. 240 (1907) and Fil. Chile v. 198 (1910).
Heliophytum floridum A. DC. Prodr. ix. 553 ( 1345). Cochranea

ida Miers, Ann. & Mag. Nat. Hist. ser. 4, ii. 129 (1868) and
Contr. Bot. i. 198 (1869). Heliotropium floridum, var. “aya
ee Anal. Univ. Chile xliii. 516 (1873) and I. c. xe. 343 (1895

—— Chile in the department of Copiap6, ca. lat. - 30’ to
S.; reported, no doubt incorrectly, from Coquimbo.

38 JOHNSTON

CHILE. Aracama: Calde 1922, Gigoux (G); Caldera, Sept.
Philippi TBM, BD, MS); Calders, pry 1876, no collector given (Ms) ); Galders
1853, Philippi (MS): of Atac ama [? Caldera Region|, M orot

f ng
1236 (G, NY, US, FM); Co reste BF ic Sept. 1885, ay Roman (MS);
Atacama D : : : ak

: Geisse 23b (NY); Bandur 1885, Geisse (MS); Carriz
Bajo, De Dec. 1871, King (MS, 3 collections one the Typx of var. latifolium);
: n Copiapé and Bridges 1340 (G, K, BM).
ESSRTEREES:; ig ’ Gay 1182 (G, FM, tsoryrss A H. floridum);

oquimbo, FAs ap 858 (G, K, BM); indefinite, Bridges

A loosely decumbent shrub with large white corollas. It varies
much in the size of the leaves, the var. lati ifolium being a rare extreme
in which the leaves attain about a centimeter in width. The type is
given as coming from Coquimbo, but came in fact, I feel confident,
from the Caldera region. It might be noted that Brides 1340 was
incorrectly cited as H. stenophyllum by Miers.

HELIOTROPIUM sp.—CHILE: open rocky canyon, Antofagasta
100-300 m. alt., April 3, 1925, Pennell 13022 (G, FM); =
Dec. 1913, Salinas (IP; G, frag.). These two collections from
arid north-coast of Chile represent plants related to H. floridum, TL
linariaefolium and H. Philippianum, but apparently distinct from
all three. The habit seems to be that of H. floridum though the
corolla is only 6 mm. broad and the calyx noticeably smaller. The
flower-color is uncertain. The small leaves have a sparse but loose
pubescence. More material is needed before the exact status and
relations of this apparently unnamed plant can be properly ascer-
tained.

IX. Section Heliothamnus Johnston; type-species, H. arborescens

With the exception of one species which extends northward to
southern Mexico the members of this section are restricted to north-
western South America. Iti is an extremely well defined group, prob-

ti Cochranea, an erized
iy evi ing ellipsoid, lacunose or - sculptured nutlets, which are sitachiea
medially and ventrally by a definite elliptical areola. The anthers

: eral species of the section are usually readily distinguished and

STUDIES IN THE BORAGINACEAE 39

recognized by gross habit alone. Strangely, however, they are very
difficult subjects to key out. This is particularly the case with the
Ecuadorean species. Careful field studies are distressingly needed
for the clarifying and the determination of interrelationship of such
pecies as H. submolle Kl., H. Urbanianum Krause and H. rufipilum
(Benth.) Johnston.
Key To SPEcIEs
Style elongate, surpassing the stigma in es (ef. H.-submolle).
Leaves plane or a weakly rugose, the veining only weakly
sence ste surface ahalwate 6 or finely pubescent

imp
or somewhat tomen
Leaves large, 20-25 m Genk 4-9 em. long........ 34. H. arborescens.
Leaves small, 7-13 soit broad, 2.5-5 cm. long... a8. ‘H. lanceolatum.
Leaves coarsely rugose with deeply impressed veins, the
lower surface — = whi mentose.

seabrid with rather sparse
ascending bristles, nes an app pu nce;
coro CORR. ee 36. H. incanum.
Upper ol pao me and abundantly strigose or even
—— weakly if at all seabrid; corolla 2-3.5
mm
Leaves lanceolate, 4-5 cm. lon: or eo 37. H. erianthum.

— obovate or obovate-oblong, = no long, apex
38. H. lippioides.

‘ ay ape be Ree ae a
Plant deeded pallid, copia appressed white-villous,

not at all glan dular; leaves ovate................5 'H. argenteum.
t green or at most so cinereous or fulvescent, not
white-villous, —— somewhat glandular; leaves
oblong to lanceo

Calyx small, 1-1.5 mm. becoming 1.5-2 mm. long, not

conspicuously acerescent in fruit.
Ovary and nutlets glandular; ; leaves 2-5 apn long. 40. H. adenogynum.
Ovary and nutlets not glandular; leaves 7-14 em. ae
. H. Mandonit.
Calyx large, 1.5-2 becoming 3-5 mm. long
Leaves —~ — smooth above with the veins weakly
if e mpressed; we loosely branched; spikes
pr aio "3-20 On We a 42. H. rufipilum.
Leaves firm, w our uatios with impressed veins and hence
re or leas rugose; plant more erec t and loosely
branched; spikes 2-6 em. lo
Leaves 3-6 cm. long, tending to be broadest at or
above the middle; style ackicia: more than

16-4¢ as long as the stigma.

34. H. — ens L. Syst. ed. 10, 913 (1759). H. perurianum
L. Sp. Pl. ed. 2, 187 (1762); Curtis, Bot. Mag. iv. t. 141 (1790-95);
Baill. Hist. 3 x. 353 figs. 269-272 (1891). H. edoratum Moench,
Meth. the (1794). H. odorum Salisb. Prodr. 112 (1796). H.

. & P. Fl. Perwv. ii. 2, t. 107a (1799); Sims, Bot. Mag.

40 JOHNSTON

xxxix. t. 1609 (1814); Bonpl. Pl. Rar. Malm. et Nav. 83, t. 32 (1813).
H. grandiflorum Donn, Hort. Cantab. ed. 6, 42 (1811), nomen;
Schrank, Pl. Rar. Hort. Acad. Monoc. i, 2, t. 2 (1819).
Var. genuinum. Herbage with an appressed, not very dense
pubescence; leaves not conspicuously if at all canescent beneath.—
dante my given above
Known only from ee Department of Lima, Peru.

PERU. Lima: ag rel Rose 18639 (NY, ag oe 2400 mk,
Machride 2938 (G); Matue 1877, Savatier Moa
Pennell 14352 (G, FM,); Cenk. ta, ’ Mathews (K); Atoe con. nell
14759 (G, FM); San Damian, ca. 2700 m., Hrdlicka ( (U8); hehe Obrailo,
Wilkes i gga (US); Purroe huca ca, Mathews 499 (KK); Chrachucos, r
2284 (BM); Sauga, Barclay 2353 (BM); Amancaes, ca. 500 m., We oe ee
1637 (BD); Rimac Valley, 1882, Ball (US, ou Lima, André K1397 (K);
near Lima, Nation ifi in pt. (K); Lima, Dom x herb. Ruiz (BM). In-
DEFINITE: no locality, McLean (kK), Cuming ad " (K) and 1077 (K, BM),
Forbes (BM).

Var. grisellum, var. nov., a varietate genuina differt pilis longi-
oribus laxe ascendentibus et magis griseis, foliis subtus saepe evidenter
canescentibus.

Southern Peru and adjacent Chile.

CHILE. Tacwa: without data, vss ge (G).

U. Moagusewva: Carumas, 27 ping > gg sec wl beef é FM, gh
G, — Arequipa: Cotahuasi 2 (FM, BD);
Poseo, ca. 550 m., Cook & Gilbert 0 ( (ua) Mellende. i Badont 29394 (Us);
Mollendo, ca. 400 m., Weberbauer 1466 (BD). Ayacucno: below Coracora,
2900 m., We berbauer 5799 ( BD).

The var. genuinum from the region sake Lima eae ~ be von
form f: from eliotrope hac heen denvue
Miller’s plant from the Chelsea Gardens, _ type of which is in ma
British Museum, is ¢ clearly this form. _ The form of the species found
in southern P. pious, more spreading pubescense .
and because of these differences, coupled with its distinct range, the
plant has been given varietal recognition. Possibly referable a the
variety is a collection made by Weberbauer, no. 2710 (BD), in the
Department of Ancash below Oercs at about 2650 m. alt. This is
north of the range of the var. genuinam but quite separated from the
other stations known for the variety. Another very peculiar form
comes from river banks near Loja in southern Ecuador where it was
collected by Hartweg, no. 810 (K, BM). I doubtfully refer it to the
var. genuinum since in its small flowers, slender habit and scanty
pubescence it is not hay
= lanceolatum R . & P. Fi. Perwy. ii. 4, t. Ila (1799).

STUDIES IN THE BORAGINACEAE 41

PE Lima: Lurin, 1881, Ortiga (G); — Wilkes Exped. (G, US);
near Lima, 1862, Nation 111 lin ‘pt. (i): ‘ Lima-Oroya R. R. betw: ween Narquinia
Matucana, ca. Weberbauer 96 (BD). Inpevuwrre: ‘in Peruvia

- ad Huanuco Lima ete," "Ruiz (BD, 1sotyre of H. lanceolatum).

5

The original material is given by Ruiz & Pavon as coming from
Huanuco and the adjacent village of Pillao. The authentic material
of the species from the herbarium of Ruiz now presery ed at Berlin
gives the source as from “Huanuco Lima etc.” All the material
with definite data comes from the region about Lima. It agrees
remarkably with both the original description and_ plate. The
species is ‘characterized by its small scabrous lanceolate leaves and
loose corymb of long-tubed flowers., It is probably most related to
i. nn but certainly appears to be th ages agar

36. H. incanum R. & P. FI. Peruv. ii. 2, t. 108a (1799

Central and sah Peru, iiesiaaune: pearics east of sa western
cordilleras.

PERU. Cusco: Urubamba Valley Herrera. Si (G, —
Vileanota Valley, 3000 m., Herrera 1080 xc, Us); Ollantai
Herrera 231 (BD); Ollantaitam tambo, ca. 3050 m ennell 13640 (G, ;, FM).
URIMAC: Ri acienda Co sacle ; 2950 m., Weber
5855 (BD). Huancavetica: Man Jalley near La jorada, ca. pin
m., 7 , FM). Junin: Cabe ve Hertas, :
Macbride & Featherstone 1327 (FM); vicinity of a, 76 (NY,

Macbri Oroya, Kalenborn
ra Hvanuco: Huanuco, 2100 m., Macbride 3237 (G, FM); Huanuco
Ruiz (BD, 1soryPe). INDEFINITE: “ Colombia, ” Lobb 97 (K); indefinite,
Dombey (FM).

A very characteristic species with ovate, usually conspicuously
bicolorous leaves that are very rugose very scabrous or rough-
pubescent above. The type came from Huanuco. Its nearest
relative is H. arb 1 f. proach that species closely.

37. H. erianthum, sp. nov., fruticosum canescens 2 m. altum;
ramis teretibus canescentibus tomentosis et sparse villosis laxe
ramosis longiusculis; foliis lanceolatis alternis 4-5 cm. longis 10-17
mm. latis apice acutis basi in petiolum 5-8 mm. longum # attenuatis
supra obscure viridibus tomentulcsis tessellato-rugosis cum nervis
impressis subtus prominenter venosis albescenter tomentosis margine
obscure crenulatis paullo revolutis; spicis 2-5 em. longis ebracteatis
tomentosis et villosis in corymbum terminalem vel axillarem dicho-
ome ramosum dispositis congestifloris; calyeibus sessilibus dense
duce villosis aurantiacis quinquifidis in anthesi 2-3 mm.
longis sed demum 4-5 mm. longis, Nobis anguste lanceolatis vel subu-
latis; corolla alba siccata brunnea hypocraterimorpha 4-5 mm. lon:
extus dense adpresseque villosa intus glabra, tubo cylindrico 25-3
mm. longis calyce paullo longiore, limbo ca. 3 mm. diametro, lobis

42 JOHNSTON

late ovatis rotundis patentibus; antheris lineari-oblongis ca. 1 mm.
longis subsessilibus in medio tubo affixis dorso apicem versus pilis
obesis dense obsitis, apice separatis et extrusis; ovario glabro; stig-
mate ca. 0.8 mm. longo conico; stylo 1-1.3 mm. longo; nuculis ignotis.

PERU. Prora: eg of cultivated land north of Huancabam m road
to Jicate, ca. 2050 m., May 12, 1912, Weberbauer 6332 (G, sea ‘FM, BD,
SOTYPES).

Clearly a relative of H. incanum, from which it differs in its pubes-
cence and smaller flowers. It is also suggestive of H. argenteum of
southern Ecuador but differs in its longer style and narrow strongly
rugose bicolorous leaves. Prof. Weberbauer, who collected the type,
notes that the corollas are white with a yellowish throat and that they
were fragrant with an odor similar to that of “ H. peruvianum.”

38. H. lippioides Krause, Bot. Jahrb. xxxvii. 632 (1906).

PERU. Casamarca: Santa Cruz, ca. 1650 m., 1904, Weberbauer 4122
(BD, trre; G, photo.).

A very distinct species known to me only from the type-s
It is characterized by its small obovate silky-strigose leaves and small
white corollas. The relationships of the plant are with H. —
and H. incanum, as its rugose leaves with deeply impressed v

39. H. mteum Lehm. Nov. Act. Acad. Caes. Leop. Nat. Cur.
ix. 139 (1818) and Asperif. i. 73 (1818). H. lanatum HBK. Nov.
et Sp. iii. 89 (1818) and I. ¢. 451 (1820).
Seabees Ecuador.
ECUADOR. Azvar: Paute, ea. 2400 m., Lehmann 6547 (G, US, K, BD);
igs ea m., Jameson (K); Nabén, Cuenca, ca. 2600 m., Lehmann 884

K, BD); ceo, Jameson 155 (K). Losa: “ Loxa
: BD, tsoryrr of .. lanatui m); * Am Merid. _ Humboldt Hert Willd.
vtTyPe of H. argenteum). Inperintre: “‘Colombia,’”’ Lobb 44 (K).

The type material came from Ecuador in the province of Loja,
near Loja and Gonzanama. The latter is a town situated ca. 35 km.
southwestward of Loja, about equidistant from that city and the
Peruvian boundary. A very distinct species readily recognized by
its ovate white-villous leaves.

40. H. adenogynum, sp. nov., frutescens; ramis teretibus laxe

ramosis longiusculis breviter villosis rariter subvelutinis plus minusve
ge sg folis alternis sparsis herbaceis venosis integris ovatis
tis saa em. longis 6-28 mm. latis sparse villosis
cee nd , ap tis, basi acutis vel obtusis in petiolum
oF $8 mm. 1 longum atienuatis, pera paullo pallidioribus, margine

STUDIES IN THE BORAGINACEAE 43

integris vix revolutis; spicis gracilibus 4-10 cm. longis ebracteatis
multifloris in corymbum laxum terminalem dispositis; calycibus
sessilibus globosis 1-1.5 mm. longis demum ad 2 mm. longis villosulis
saepe glanduliferis ad medium vel ultra in lobis_triangularibus
vel elliptico-lanceolatis erectis vel conniventibus divisis maturitate
plus minusve deciduis; corolla alba hypocraterimorpha ad 4 mm.
longa extus sparse villosula et non rariter glandulifera intus glabra,
tubo quam calyce 2- ad 3-plo longiori, limbo ad 4 mm. diametro,
lobis orbicularibus patentibus; antheris oblongo-limearibus 1-1.3 mm
' longis subsessilibus infra medium tubi affixis, apicem versus cum
pilis crassis dense obsitis separatis inclusis; ovario glandulifero;
stigmate conico ca. 0.8 mm. longo; stylo subnullo; nuculis 4 ovatis
rugosis 1-1.5 mm. longis stipitato-glanduliferis.
PERU. Lima: Chosica, 900 m., — & Featherstone 4 FM, Tr

G, photo.; US, meer E); Chosica 1050 m., Weberbauer 5312 > EM, BD)

hosic ca, Rose 18 543 (US); near Peal Wilkes’ Exped. (G, NY, US);
Gaudichaud (BD).

A remarkably distinct species characterized by its small flowers,
glandular fruit and ovary, and small scarcely accrescent more or less
deciduous calyces. The fruit is covered with long-stiped glands.
In age the glands tend to become broken and the fruit appears to be
sparsely villous. The species is most closely related to H. Mandonu
but also has an evident relative in H. rufipilum. According to
Professor Weberbauer, in sched., the plant is a small shrub a half-
meter high and has flowers which are at first pale violet but later
become white. He also adds that the odor of the corolla is entirely
different from that of “ H. peruvianum.”

1. Heliotropium Mandonii, sp. nov., fruticosum 1-2 m.
ramis teretibus laxe ramosis longiuscule dense retrorse fulv sie
demum sparse villesis vel glabrescentibus; foliis alternis sparsis 7-14
em. longis 2.5-6 em. latis late lanceolatis vel oblongo-ellipticis, apice
acutis, basi in petiolum ca. 1 em. longum fulvo-villosum attenuatis,
margine obscure erosis vix revolutis, supra scabridis et sparse villoso-
hispidis saepe cum nerviis impressis plus minusve tessellato-rugosis,
subtus pallidioribus dense velutinis fulvo-canescentibus vel vere
canescentibus; spicis 5-12 em. longis gracilibus ebracteatis breviter
villosis multifloris in corymbum terminalem dichotome ramosum
dispesitis; calycibus sessilibus sparse villosis non rariter glanduliferis
globesis ca. 1.5 mm. longis demum 2-2.5 mm. longis, lobis lanceolatis
vel ovato-lanceolatis acuminatis maturitate ovatis sinubus rariter
paullo plicatis; corolla alba 3-4.5 cm. longa extus pubescenti et non
rariter sparse ginndalters intus glabra, tubo eylindrico ca. 2 mm.

=—_—S—“‘i‘“C—C;~;~C~C~™~™~™~™

44 JOHNSTON

longo calyce 1.5-2-plo longioribus, limba 3-5 mm. diametro, lobis
ovatis obtusis patentibus; antheris lineari-oblongis ca. 1 mm. ongis,
apice extrusis libris, dorso apicem versus cum pilis congestis crassis
ascendentibus obsitis; filamentis perbrevibus paullo infra medium tubi
affixis; stigmate sessili ca. 0.5 mm. longo; ovario —— nuculis 4
glabris lacunosis ca. 1.5 mm. longis compressis late affixis
BOLIVIA. La Paz: Cerro Iminapi, vicinity of Sora
387 (G, TypE; NY, K, - ISOTYPE); re 1920, Holwayy ~~ 580 (G. WY US),
PERU. Cuse usco: Hacienda Potrero, Valle de § : Prov. n-
vencion, ca. 1100 m. Farce 899 (BD). Hvanuco: Yanan sing 1800 m., at
Machride 3660 (US, FM); Rio Huallaga below Rio Santa Domingo, 1200 m
1923, Macbride 4206 (US, FM; G, photo.).

A coarse plant somewhat suggesting a Tournefortia in gross habit.
It is most related to H. adenogynum and H. rufipilum, but quite
distinct from both. _ It is characterized by its small scarcely accrescent
calyces, and its large more or less velutinous leaves; a very well marked
and distinct species.

42. H. rufipilum (Benth.), comb. nov. Saag ahaa he By
Benth. Bot. Sulph. 140 (1844). H. physocalycium Donn
az. xlix. 457 (1910). H. jaliscense Macbr. Proc. Ais peng i
542 2 (1916) and Contr. Gray Herb. xlix. 17 (1917).
ar.genuinum. Nutlets beiehig: aueerces long-stiped glands.—
Synonymy given above.

Southern Mexico (Jalisco and Oaxaca) to Salvador; Ecuador where

not common.
ECUADOR. Guaras: Milagro, 50 m., Hitchcock 20200 (G, NY, US);
Chanduy, Spruce 5983 (G, K, BM); Guayaquil, Orton (G); Baleo, Eggers
(us). Manasi: Salango, 1342, Sinclair (K, tree of T. rufipila; G;
. snag Picuincua: silv. suband. M. ‘Carazon etc., Sediro 112/20 (BD);
subtrop. Mindo, Gualea, ete — 112/16 ( ae veep:
Suquibl, west ‘alee of contin 500 , Rimbach 30

Var. anadenum, var. nov., a varietate genuina differt nuculis
maturitate glabris nullo modo ee sparsissime glanduliferis.

Ecuador and adjacent Peru with an = station in Bolivia.

La Paz: near Yungas, 1 an sg ate Bad ho
ae ogee: Talara, Haughi 70 ‘Ge Parifins Valley
ADOR. Azuay: Cuenea, Rose, Pachano & on
Curmporazo veinity of Hier Rose Ive 22/31 ( ‘GNY. NY, a); Baie,
| Holway 818 US); Spruce 5534 (G, K, BM); Pallatanga

valley, Soir 113/18 (BD). AR: in silv. trop. a sapamba,
: Sodie i12/ 12/ seta 5 PIcHINCHA: Gaite, 2100 m., Jameson 428 (K, BM).

Goaras: ¢ |, Mille ; :
(K), Sena 406 odio
ene

a oe Yen men ISOTYPE) ;
BM), Hortwen 6 681 c.4 , BD). ee 243 ‘BR. cMaxani: El

>

Spey ream apen noun

STUDIES IN THE BORAGINACEAE 45

roo 10, 1876, André 4037 (NY, K); village of Lasarauga, 1874, Seemann
90 (K); Central Andes, Barcley MD indefinite, Fraser mai Jameson (US).
The typical form of the species is known from San Salv ador, Guate-
mala and southern Mexico where it is practically confined to the
Pacific slope. It reappears in Ecuador but seems to be much less
common there than the endemic South American variety, anadenum.

In habit and gross aspect the material of H. rufipilum from Central

and South America seems quite the same, although the var. genwinum
has the fruit provided with coarse gland-tipped villous hairs and the
var. anadenum has the fruit glabrous and usually quite devoid of
glands. The form with glandless fruit is apparently absent north
of Panama, although it appears to be overwhelmingly the prevailing
one to the south. The species is very well marked by its lax habit

of branching, mages pubescence, and broad inflated mature
calyces with almost cordate lobes. It seems to be rather variable in
the character of caecaue and in size of flower.

43. H. Urbanianum Krause, Bot. Jahrb. xxxvii. 633 (1906).
napa Ecuador and adjacent northern Peru.

AMAZONAS: Chachapoyas, Mathews (K, NY). INDEFINITE: no

osCUA ADO Cumporazo: Alausi, 1864, Jameson Petes Prov. de Homer,
Jameson 154 ( K). TUNGURAHUA: betw Pelileo selene Quero, ca. 2550 m.,
Lehmann 5779 (BD, TrpPe; G, K, isoTyPEs s). cHA: Quito, Jameson 183
(K); Pomasqui, ca. 2600 m Jameson By). Pomascui , Jameson 318 (K,
BM); Pomasqui, 2400 " uae 570 (NY): San Anto! ‘0, 2700 m ameson
84 (K). Inwpapura: Rio ‘Chota, Ibarra, ca. 1650 m. "Lehmann 6238 (K, BD);
Betw. La i : K); Arenales
de Sali . 1650 ‘BD). “gee oe Ambato a
Calicali (Quito), ea. Lehmann 379a (B

= 112/17 (BD): sedeheree Fraser (BM) and pay’ 1316 16 (NY. K, arg
Mest closely related to H. rufipilum but although distinguished
from it at a glance, characters by which it may be said to differ
constantly are singularly lacking. The present species usually differs
in having much less elongate spikes, its flowers have less elongate
tubes, the calyx does not become globose and the lobes are not con-
spicuously broad. The leaves of H. Urbanianum are smaller than

in its relative, usually rugose with impressed veins — tend to be

broadest at or above the middle. The pubescence on the foliage i is
usually more copious and the hairs feeaientty have pilbots bases.
44. H submolle Klotzsch, Gartenzeitung xx. xx. 89 (1852);

Allgem.
Regel, Gartenfl. ii. 259, t. 63 — H. incanum, var. glabra
imate Garden. Chron. ser. 2, xxii, 809, fig. 140 (1884). Tour-

nefortia stenosepala Krause, Bot. Jahrb. xxxvii. 631 (1906).

46 JOHNSTON

Colombia to Peru; not common.

PERU. Avyacucna: above Yanamonte, ca. 2550 m., Weberbauer 5600 G,
FM); road between Tambo and Rio Apurimac, ca. "2550 m., We
5600 (BD). Huvanuco: Huacachi near Muifia, ca. 2000 m., Macbride 1187
(G, US, FM). Lisertap: Hacienda Liaguen, Prov. Otuzco, 2600 m., Weber-
bauer 6990 (BD). Casamarca: below San Pablo, ca. 2300 m., Weberbauer

ECUADOR. a HIMBORAZO: Pallatango = Sn ca. 1750 m., Lehmann
5780 (BD), TrPE of T. stenosepala; G, K,

COLOMBIA. / Celer Canaan, Mt. rena ‘a "2050 m., Killip 6665
(G, NY, US).

The above cited plants seem to be conspecific and to represent the
plant figured by Regel and by Masters. I could not find any authen-
tic material of the var. glabrum at Kew, nor any material so labeled.
Lobb’s collection, no. 41, mentioned by Masters was not to be found.
At Berlin I saw a specimen of apparently camera as . submolle.
This was from the Botanical Gardens and was grown from seeds
provided by Warszewicz. This appeared to be the same as the
plants I have cited.

I doubtfully place under H. submolle four other collections at
Berlin. These are as follows Tied ” Pavon (Herb. W illd) ; Alausi,
Humboldt (Herb. Kunth); Prov. Loxa, Warszewicz, and “‘in suband.
m. Pich. prp. Chiquilpe,”’ Sodivo 112/18b. These specimens have
firm hairy foliage 6-8 cm. long and 2.5-4.5 em. broad. In varying
degrees these specimens suggest H. rufipilum, H. Urbanianum and
H. sulbmolle.

Weberbauer’s two collections cited above can only be referred here.
They are noteworthy, however, because of their very short styles and
conspicuously eege corolla-tubes.

The relationship of H. submolle seems to be strongest with H. ar-
borescens. In fact the species seems to be one of the intermediate
forms connecting the group of H. arborescens with that of H. rufipilum.
In the latter group it oo H. Urbanianum most closely.

_X. Section Orthostachys R. Br. Prodr. 493 (1810); DC. Prodr.
ix. 539 (1845); Benth. & Hook. Gen. Pl. ii. 844 (1876); type-species,
H. foliatum R. Br. Orthostachys Spach, Hist. Veg. ix. 32 (1840).
Preslaea Mart. Nov. Gen. et Sp. ii. 75 (1826); type-species, P. para-
doxa Mart.; not Preslia Opiz (1824). Euploca Nutt. Trans. Am.
Philos. Soc. ser. 2, v. 189 (1837); type-species, E. convolvulacea Nutt.
Schleidenia Endl. Gen. Pl. 646 (1838); type-species, P. paradoxa
Mart. Pioctonon Raf. Sylva Tellur. 88 (1838); type-species, P. antil-
lanum Raf. Sareanthus Anderss. Vet. Akad. Handl. Stockh. 1853:

pes ve bsbcisietsicane S. asperrimus Anderss.; not Lindl. (1821).

1

STUDIES IN THE BORAGINACEAE 47

The section Orthostachys is the largest section in the genus and
probably the most difficult. It contains about fifty American species,
approximately half of which occur in South America. In the Old
Norld the section is also represented by species in southern Asia,
Australia and Africa. As a group it is recognized by its ovate or linear
anthers which are glandular or hairy at the narrowed, frequently
coherent apices. In most of the species one of the inner faces of each
nutlet is marked by a circular or elliptical pit. When present, this
latter development is very characteristic of the section. The section
breaks up into at least three subsections.

The subsection Azillaria is characterized by having the flowers
borne along the leafy stem and not aggregated into a definite spicate
or racemose inflorescence. The fruit is not lobed, but is merely sulcate
and is usually drawn up into a beak. The faces of the nutlets are
not pitted. The five species of the section all occur in South America,
although some of them range northward to the Greater Antilles and
Mexico. .

The subsection Ebracteata consists of over a dozen species, practi-
cally all of them American and the majority of them from South
America. It is a very natural group characterized by having its
flowers in completely bractless spikes or racemes. All the species
have pitted nutlet-faces.

The subsection Bracteata has bracted spikes or racemes. It is
cosmopolitan and contains the most of the species of the section.
Among American Heliotropia at least, it presents to the systematist
most puzzling and difficult problems. The species are very variable
and illdefined and seem to be singularly lacking im statable differences.
Although extremely unsatisfactory the classification and keys to the
subsection presented here are the best I have been able to work out
with the time and materials available. Rather than omit the sub-
section completely the provisional treatment of it is, therefore, pub-
lished for I suspect that despite its inadequacy it may be found useful.

Most species of the subsection Bracteata have equally lobed fruit
and nutlets with pitted inner faces. In the North American H. con-
volvulaceum Gray and in the South American H. chrysanthum Ph.

-~ and H. catamarcense Johnston the fruit is strongly compressed and
the nutlets not pitted. The fruit of these three species is peculiar in
the section but I do not believe that it should be unduly stressed.
To do so would result in the separation of such closely and i
takably related species as H. chrysanthum Ph. and H. catamarcense
Johnston from H. mendocinum Ph. and H. Greggti Torr.

48 JOHNSTON

Key To SussEcTIons

Flowers borne = a Hale ee eta See 1, AXILLaRIA.
Flowers borne i ikes o reo
florescence ‘cea sharp acy... ee EBRACTEATA.
pay msg racted, the bracts usually evident but some
i filiform, eaducous and inconspicuous........... . BRaAcTEaTa.

= TO iam OF SUBSECT. AXILLARIA

Plaits at corolla-sin most _e poomnages into inconspicuous
tooth-like Inbal; cheats 2-4.5 mm. lon
Plant — * overed by an inconspicuous appressed
* pubescence
Cora lia. 23. mm. long; fruit with a conical apex; plant
MN MON Se 45. H. lagoense.
Bisson S451 5 mm. long; fruit usually beaked; plant —,
Caw a eee OU oe Tete H. antillanum
Plant 1 viltcwn, the hairs spreading and ¢ conspicuous. . ‘7 H. humistratum,
Plaits at corolla-sinuses eigen eer into linear falea
lobules; t usually
a rele aa 4—§ mm eg Hasna APRNs NE RS oe ah reiflorum,
rolla large, 8-11 mm. ies pe ee ee oe oy it sea eb

Key To Spectres or Sussecr. EBRACTEATA

Anthers ovate or ovate-oblong, glabrous, acute or acumina
terminated by a conical ai subulate p rolongation = the
connective; a widely distributed semaine Goi
ENON ee ee ee a ve "50. H. procumbens.
Sethe | linear, obtuse, more or less pubescent or glandular
at apex.
Stigma-column evidently penicillate with long white hairs;
P — eke ar ean aa ere re sea H. barbatum.
tigma-co umn not penicillate, glabrous or sparingly
ously strigose.

inconspicu
Corolla-lobes
Corolla large, 4-5 mm
t can t; — de te-strigose;
stigma ca. | long; corolla sparsely long-
slclbies inside above the middle; northern Peru. .52. H. Lobbit.
Plant green; foliage inco icuously ‘ees ose; stigma
ea. 0.5 mm. long; pane 3 glabrous maidie; southern
POG ee a 53. H. toratense.
Corolla ~—_, 1-3 mm. broad.
Plant s lowsely branched shrub 5 dm. tall; corolla
glabrous within, yellow............_. polyanthellum.
Plant an annual herb or a spreading wri
—- corolla ao hairy in throat and 2
above = a We . H. pilosum.
EE cee acute or
Corolla-tube not sanpunanae the calyx; insular. . H. Anderssonii.

Corolla-tube at least twice surpassing the calyx; pated

' STUDIES IN THE BORAGINACEAE 49
Key To Species or Supsecr. BRACTEATA
Fruit not compressed; nutlets very obscurely didymous,
densely Wecabv licen, 1-2 mm. long, commisural faces .
Decca ee Cy pecs poe ee ce 59. H. mendocinum.

oe ae once

Kadhens sah halk toheaeh: tena sehel ae Oe
63. H. filiforme.
Anthers coherent at — > bracts oblong or elliptical,
Reape es ag-etonwties 1c3) _ surface
inconspicuous, 1— mm. upper
of leaves glabrous except for appressed hairs

rege ow tam age yb ee
Bracts conspicuous, a mm. long; upper face
ves evidently strigose............... 65. H. ocellatum.
Minne tices an 3 See short style

66. H. polyphyllum.

Plant erect or rarely decumbent; Paraguay and deca
Leaves es lanceolate to to ovate, if = on both surfaces
j then elliptical or ovate; plant erect or decumbent.
| Lea’ and pots ett . eT
ves narrow elongate, strigose on 8
Pubescence very fine and obseure............ 68. H. pallescens.
Pubescence coarse and obvious.......... 69. A. distantiflorum.
Leaves with merely hairs ve;
leaf-margins more or less strongly’ revolute.
Plant herbaceous, annual, with foliaceous H. fruticosum.
Plant frutescent, ly i
pie aves dark-colored; Colombia to Argentina.
~co s
Plant drying ligh Agee
t ing
} ‘oliage more or less scabrous, ap a
} not densely no: — strigose; Carib- .
Foliage not scabrous, densely and sect piaiieetty as ae come
' not stri- a 4
i Brazil and adjacent Guiana.......... .T3. a. Fumana.

ee es a ee
— Schleidenia lagoénsis Warm. Kjoeb. Vidensk. _
Meddel 1867: 15 Seca i
(1912).

50 JOHNSTON ’

Eastern Bolivia northward through Brazil to Venezuela and
Trinidad.

BOLIVIA. Santa Cruz: Portochuelo, Prov. Sara, 400 m., 1924, Stein-
bach 6610 (BD); Laguna secas de Buenavista, Prov. Sara, 450 m., 1916,

Goyaz: betw. Cavalcante and Conceicao, Burchell —— A
K); betw. Natividade and Porto Real, Burchell 8323 (G 203
Flaziou 2 ec K N Proce lake-shore

arming (G, photo. and frag. of TyPE). Rio Sapo,
(RD) Traill : 571 ht INDEFINITE: without locality. Sellone 18 1 90 ( <BD) eas 3649

EZUELA. Cosepes: Culebra Lagoon near San Carlos, 1925, Pittier,
11703 (G, US, K).

A slender usually annual herb, which has been frequently sinha
with H. filiforme, which it somewhat resembles in gross habit and
from which it is readily separable by its axillary flowers and different
fruit. The species is mcst closely related to H. humistratum, from
which it differs in scarcely more than lack of conspicuous spreading
pubescence. It is also related closely to H. antillanum. Through
the kindness of Christensen I have been able to examine a fine
photograph as well as fragments of the type of Warming’s H. lagoense.
The plant i is pegecsmeonns conspecific with H. trinitense, from the
Island of , the type of which I studied at Berlin

46. Irban, Symb. Ant. iv. "528 (1910).
Colombia, Panama and the Greater Antilles; apparently also in |
Mexico.

COLOMBIA. MagpaLena: near Lake Sapatora, Chiriguana, 1924, |
Allen 162 (K).

Very closely related to H. lagoense and perhaps intergrading with
x Hh appears to differ, however, in its larger corolla, usually beaked
hirsute ial root and more northern range. It is best
known from Cuba, Inte of Pines and Porto Rico,

47. H. humis' m Cham. Linnaea iv. 462 (1829). Schleidenia
‘kennel He: Sa Mart Fl. Bras. viii. pt. 1, 34 (1857). S. -
— lulans Fresen. 1. ec. H. pullulans Giirke in E. & P. Nat. Pflanzenf.

iv. Abt. 3a, 97 (1893).

i, SS

Ss Alvez, Oct. 1839, G
PP cog Geen Faz. de Vindae 1818, Sellow ere ‘BD, ISOTYPE of H.
freumistratum); between Ouro Preto and Marianna, Glaziou 13822 (K, BM,
BD). Inverrrre: without locality, Sellow og Pohl, herb. Vind. 1756 in
pt. (BD), herb. Vind. 1757 in pt. ( (K), Ferreira 967 (BD).
ee a ee ne by Sellow, is conspicu- —
ee ously vllon he plant, becom-

a

|
e
5

A

|

STUDIES IN THE BORAGINACEAE 51

ing sparsely pubescent and glabrescent towards the upper extremities
of the stems. The cited specimens of Gardner and Ferreira are quite
typical. The other specimens cited are villous throughout and are
apparently referable to H. pullulans. The collection by Glaziou is
rather robust and except for unbeaked = quite like a collection by
Palmer, no. 466, from Acapulco, Mexi

48. H. parciflorum (Mart.) Giirke in in E. & P. Nat. Pflanzenf. iv.
Abt. 3a, 97 (1893). Preslaea parciflora Mart. Herb. Fl. Bras. 183
(1837). Sehleidenia parciflora DC. Prodr. ix. 557 (1845).

Known only from southcentral Brasil i in the states of Goyaz and
Matto Grosso.

tro Grosso: Cujabd, herb. Mart. 277 (NY, K, ISOTYPES);
near Cujaba, 1 6 Maser — spy” Goyaz: Nanoe Alvez, — 1
ner 3359 in pt. (K, BM); Natividade, , Gardner 8 (K);

jbetw. Canciones and Conceicao, Burchell 7958 a :

A species related to H. paradoxum from which it differs primarily
in its smaller corollas. It has been much misunderstood and has been —
even with such distinct species as H. lagoense and H. an-

49. H. paradoxum (Mart Mart.) Giirke in E. & P. Nat. Pflanzenf. iv.

76, t. 164 (1826). Scbleidenia paradoxa DC. Prodr. ix. 557 (1845).
_S. macrodon Fresen. in Mart. Fl. Bras. viii. pt 1, 35 (867). zz.
-macrodon Giirke, |. c. 8. Gardneri Fresen.|.c. H. Gardneri Giirke,
ie.
_ Eastern Brazil in the states of Piauhy, Bahia, Goyaz and Minas
Geraes.

BRAZIL. Bann: near Rio San Francisco, Joazeiro, 300 m. ,Chase 7912

(G); phages og Jacobina and St. Thomé, Blanchet 71 (BM); Serra
bina 2653 (NY, K, BD); Taboleiro near Remanso, 1906, Ule
42 (BD); indefinite, Martius (BD, ‘tsotree of P. paradoza?). Gora:

| Boa Feb. 1839, Gardner 2262 (NY, US, K, BM). Inperrvrre:
i watt ia Pohl, herb. Vind. 1756 in pt. (BD); herb. Vind. 1757 in pt.
' "The material cited above from the state of Bahia is ‘ovddeutly « con-

specific and exhibits only a little variation in the abundance of — :

: pubescence and in the development of the beak on the fruit. The
other collections differ from those from Bahia in having the corolla-

anes eteags el a ee |
— oe

52 : JOHNSTON

be indistinguishable. They are villous and canescent plants very
similar to typical H. paradorum, in fact appear to differ from it only
in the more elongate corolla-lobes. Although Gardner 2262 was one
of the three plants cited under the original description of S. dasycarpa
it is clearly a close relative of H. paradorum and is here referred to
that species. Except for the size of the corolla Gardner 2262 and 2263
and Burchell 8184 seem to differ little from the typical material of
H. paradoxum. They do, however, seem to have more slender wiry
stems than any of the other specimens I have referred to the species.

50. H. procumbens Mill. Dict. ed. 8, no. 10 (1768). H. inun-
datum Sw. Prodr. 40 (1788) and FI. i. 343 (1797); Cham. Linnaea iv.
456 (1829) ; DC. Prodr. ix. 539 (1845). Schleidenia inundata Fresen.

Leop. Nat. Cur. ix. 129 (1818); Asperif. i. 39 (1818); Kunth, Flora i}
603 (1818); HBK. Nov. Gen. et Sp. iii. fe (1820). H. canescens
Lehm. Nov. Act. Acad. Caes. Leop. N . Cur. ix. 128 (1818) and

Asperif. i. 38 (1818); R. & S. Syst. iv. 70 (1819). H. ietilind
HBK. |. c. 88 (1818). H. canescens HBK. 1. c. 88 (1818). H.
cinereum HBK. |. c. 89, t. 206 (1818). H. Humboldtianum R. & S.
Syst. iv. 737 (1819); HBK. 1. c. 452 re H. simplex Meyen,

Torr. Bot. CL iv. 224 (1895). H. riparium Mart. ex Chodat, Bull.

Herb. Boiss. ser. 2, ii. 817 (1902), nom. subnud. 4H. inundatum f.
elliptica Chodat, |. ec. v. 483 (1905). H. inundatum, var. chacoénse |

Fries, Ark. Bot. vi. no. 11, 22 (1906). H. Eggersii Urban, Symb. Ant.
-v. 481 (1908). H. inundatum, f. elliptica, subf. pusilla Hassler, Trab.
Mus. Farmacol. Buenos Aires xxi. 100 (1909).

N E:

orthern Argentina northward through tropical America to
: southem United States; rare and local on the Pacific Coast of South \
America.

sce oe Niederlein 184 (BD); islands

P tiiaignasininesy

‘Be Crees Bas “Misiones IONI : Posadas, Vattuone ala 4 S36 |

45 (G, BD). ee
otis La Banda, Lillo 6136 (. CaATAMARC.

To : Mefieeas, 500 m., Schreiter

450 m., ., Lillo 2172 (G); — Hieronymus &

1

STUDIES IN THE BORAGINACEAE 53

Loreniz 172 (G, BE) 7 Sg (BD), fag (G), 1108 (G), pe igen Villa
Haney ang 2415 (G); Rio Salf, 600 m., 2 (G);
far Pozo 300m, Venturi ad (G); Siambon, ening & yo thtead

_ mus 848 (B BD) Di la Aguadita, 700 ‘m., Lillo 7317 (G); Vipos, 780 m.,
Schreiter 1718 and roe (Gh Cadillal, 500 m., Schreiter 75 ‘4 (G); Tapia, 500 m.,
301 (G). : Ordn, Feb. 1916, Hauman (G); Metan, 850 m.,

rnelli (BD); ficy de Ina Piedras, “Rodrigues 45 (G); Sen José iosé, 1873, Lorentz &
Hieronymus (BD); Kilom. 1390, = GC. Embarcacién a Yaeuiba, 0 Oran, 500 m.,
Schreiter 3605 (G); Salto, 1864, Pearce (K). Jusvy: San Lorenzo, Lorentz &
Hieronymus $57 Bena Jujuy, 1302, 2, Kuntze (NY); Jujuy, Castillon 366b (G).
_ BOLIVIA. mayo near Ft. Crevaux, Fries 1614

: 450 m tei ;
BD}, 6 1 (FM , K, DB). Inperrrre: Villamontes, Agua Blanco, 500 m.,
Pflanz a0? (BD), 4047 (US); no locality given, Bridges 19

PARAGUAY: Lacus Y Hassler 3893 Sagditeg K, BM, BD), 11490 (G,

/US, BM, BD), 12374 (G, U ORM, BD); Salines de Lambari, Balansa 2033
\¢ ; Paraguari, Balansa Ae ; between 1 Rio Apa and Rio Aquidaban,
Picbrig ni Br (K, es Alto Paraguay, Chaco, , lat. 21°, Fiebrig 1380 (BD);

pha _ olonieberg Chaco,

and Siedelung, Fiebrig 532 (BD); Gran
Mor a (NY ¥); + Asuncion, Morong 77 (G, NY, US, BM); near a
Hassler 7341 (BM, BD; isotypes of H. inund. v. elliptica); lower course of
Pileomayo, pany i2 (BD), 13 (BM), 23 (BM, BD); indefinite, Hassler

URUGAY: pag ong cio getting 8 (KK).
Pisa oa Martro Grosso: ae Robert 787 (BM); Santa Crus,
oore 354 eee ge and 603 (BM, BD); lerere, Paraguay
BM). Falls of the Madeira, Rusby 1 an NY, US, K, BM, BD). Sio
Pauto: §. Jou da paces? Piracicaba, Puttemans 11276 (G). Rto
Janetro: Rio Janeiro, Glaziou 11292 (BD), 13032 plang Mrnas :
/ indefinite, Gastn th pol (BM, BD), St. Hilaire B 567 (K). Bawa: Serra
Jacobina, Blanchet 2666 (K, BM. na oe um); near Bahia,
FM, BM); Ri Sin Francisco near , Martius (KS),
Chase 7908 (G). Goyas: Barra de Lagoa Feia, Glaziou 21778 (NY, K, BD); :
ee rene eee ee K) and 8894 (K). Piavny: —
| indefinite, Gardner 1983 (BD). Pernampuco: Olinda near Pernambuco, —

Ule 5192 (K, BD). Inpertnrre: no locality given ” Sdive G. US, K, BM,
)), Herb. — Vind. 1593 (K, BD, coryrs of B longipetiolata)’ Hoffm
(BD, wre of H. canescens); Rio Belmonte, Martius (K, BM).
coe GUIANA: indefinite, Appua 1762 (K); Schomburgk 1024 (K,

BM, BD) ote 1026 3 (I).
crRE: Cumana, Humboldt 57 (BD, rvee H. decumbens);
Coimana, Hombni (BD, sore oH. oemaentl; Cus nt, Heelan Fi

54 JOHNSTON

set ISOTYPE of H. procumbens); Cumana, Humboldt 5. 8 (BD, tsoryre of
H. cinereum) and 5202 (BD); “Rio Apure . onpland 58 (FM).
cua: St. Mathio, Valle del as, Otto 812 (BD): Valle M

ARAGUA: ragua,
786 (BM, BD). Bottvar: Cuchivero, Playa Urané, Passarge & Selwyn 824

(BD); Isla Degrero near Ciudad Bolivar, 60 m., Bailey 1877 (Us). Gu Aric
La Rubiera near Calabozo, Pittier 27 (BD). Caraposo: Rio Chico ly
1256 (G, NY, US). Cosepres Carlos and San ael de «

: Raf;
Pittier 11723 (US). Porrucursa: Portuguesa River, Pittier 12035 (NY,
( A. Bourvar: Barro Blanco Pag” Sinu, Pennell
iss Me NY). Tierra — on Rio Sinu, ca. 65 m., ennell $871 ( (G.. NY, ont
Prati Quaymaro, ca.

338 (US). InpEriniTe:
Schultze 419 ge

ECU. R. Milagro, 50 m., Hitchcock 20170 (G, NY ag
Bodegas, Sodiro en ; (BD); Rio Daule near Guayaquil, Spruce 6490 (K

BM).

PERU. Prvra (K). Loreto: middle Rio Ucayali, ca.
160 m., Tonia one TBD) -
CHILE. Tacna: Arica, Meyen (BD, Type of H. simplex).

A widely distributed and somewhat variable weedy species which
is the American counterpart of H. ovalifolium Forsk., of tropical
Asia and Africa. The most important variation is that avobring the
size of the flower, the plants from the southern part of the range
appearing to have flowers almost twice the size of those from further
north. Some of the plants from the Bolivian Plateau appear to\
become perennial. Typically, however, the plant is annual.

51. H. barbatum DC. Prodr. ix. 541 (1845). Schleidenia barbata
Fresen. in Mart. Fl. Bras. viii. pt. 1, 43 (1857). S. leptostachya
“Fresen. |. c. H. leptostachyum Giirke in E. & P. Nat. Pflanzenf. iv.

Ale i 97 (1893).

5 tern Brazil, State of Bahia.

BRAZIL. Banta: Serra Serra Jacobina, Blanchet 2726 (FM, K, BM, BD, 1so-
Types of H. barbaun); near Jacobina, Blanchet 3610 (BM, coryre of S.
leptostachya) ; pr. Bahia, Martius (BD, tsotrre of S leptostachya).

A well marked species most related to, though quite distinct from
H. inundatum. Tt also appears to have affinities with the Peruvian

é a .
tain
a en

|
\ STUDIES IN THE BORAGINACEAE 5s

tico-oblanceolatis firmis conspicue costatis sed inconspicue nervatis
m versus in petiolum 1-2 mm. longum gradatim contractis apice

spicis geminatis multifloris ebracteatis 1-2 cm. longis; calycibus
scendentibus 2.5-3 mm. longis dense strigosis, lobis lineari-lanceo-
latis ca. 0.5 mm. latis paullo inaequalibus; pedicellis crassis 0.5-1
mm. longis; corolla 4-5 mm. longa infundibuliformi extus strigosa
intus in faucibus sparsissime longeque strigosa, tubo cylindrico ca.
mm. longo ca. 0.8 mm. crasso calyce subduplo longiori, limbo 2.5-3
m. diametro concavo lobis ovatis ascendentibus; staminibus medio
tubo affixis; antheris linearibus ad. 1 mm. longis sessilibus apice ob-
tusis glandulari-puberulentis leviter coherentibus; ovario glabro;
selon cylindrico ad 1 mm. longo, disco crasso et angustissimo;
stigmate 2-3-plo breviori; nuculis ignotis.
| PERU. Amazonas: Chachapoyas, W. Lobb (K, TYPE).
t+ very distinct species most related to H. pilosum, with which it
in habit but differs conspicuously in the size and form of the

|medium basem v tim attenuatis; spicis geminatis vel
| ternatis terminalibus scorpioideis 3-7 em. longis 4-6 cm. longe
1 rue 1 Pee =< ate oe > mm longis ornati is;

| p cum p :
| calycibus subsessilibus 3-4 (demum ca. 5) mm. longis sparse strigosis,
| Jobis li i inatis i ibus strigosis ascendentibus

' 0.6 mm. longo 4lobato subsessili disco crasso; ovario sparse villoso;
a: Torata, ca. 2250 m., March 17-18, 1925, Weber-

PERU. Mogureva:
baer 7407 (FM, TxP; G, 180 | = :
A very well marked species of uncertain relationship. It is known —

|
|

56 JOHNSTON

54. H. polyanthellum, sp. nov., fruticosum 5 dm. altum dense
strigosum siccatum paullo fulvescens; ramis numerosis gracilibus
ascendentibus; foliis alternis 8-20 mm. longis 1.5-4 mm. latis firmis
costatis sed enervatis anguste lanceolatis vel oblanceolatis acutis
basem versus in petiolum ad 1 mm. longum gradatim attenuati
margine paullo revolutis; — solitariis vel geminatis gracilibus
multifloris ebracteatis 4-6 cm. longis; calycibus subsessilibus ascen-
dentibus 1-2.5 (maturitate 1.5-3) mm. longis, lobis anguste lanceolatis

inaequalibus; corolla sordide flava crassiuscula
igus extus dense strigosa intus glaberrima, tubo cylindrico 24
mm. longo 0.7-0.8 mm. crasso calyce saepius ca. duplo longiori, limba
2-2.5 mm. diametro subplano, lobis ovatis; staminibus medio tubs
affixis; antheris linearibus ad 1 mm. longis sessilibus, apice obtt

glandulari-p

ovario sparse villosulo; stigmate cylindrico ca. 0.8 mm. longo nia
disco crasso et pace . fructus maturitate ut videtur sessili;
stylo brevissimo quam stigmate 3-4-plo breviori; nuculis strigosis ca.’
1 mm. longis dorso convexis agri angulatis.

PERU. Casamarca: between Jaen and the valley of the Rio st
750 m., inet 25, 1912, gankoe 6184 (FM, tyre; BD, tsoryp

relatives. It is known only from Satie Pere According to|
Weberbauer, who collected the type and only known specimens, the )
plant has sordid yellow corollas and is a loose light green bush a half
apes tall.
55. H. pilosum RB. J &P. Fi. Peruv. ii. 3, t. 110b (1799).

gra,

oY

pooaunee only the growing parts ‘ adisleucals canescent.—Syn- |

PERU. Laima: open sandy slopes, A ea. 375 m., Pennell 14786 |
(G, FM); “iB i the sanunkiion aint or Chosica, 1350 m., Weber- _
Se ); [Amancaes iz (B } kane aiieies a bls

ar Barranco, ‘m., pfrodtge od tg oa ge cog sea,

t

Sea a var. nov., a varietate genuina differt foliis et cauli-
tomento-strigosis. ;

bus dense albideque
_ Middle western Peru.

: Lima, 1 Nation (K); along the Rio Rimac |
ne as ee ce = sre HOR a, Weber: 4

STUDIES IN THE BORAGINACEAE bi

The lower plant on Ruiz & Pavon’s plate, that lettered “b” and
ot “a” as cited in the text, is obviously that described in their text
d represents the plant in Ruiz’s herbarium which is labeled as
. pilosum. The illustration of the flowering and fruiting parts are
quite incorrect in the original plate. The four nutlets in the authentic
aterial at Berlin are 1 mm. long, with a sparsely hirsute-stri
trongly convex back and a decidedly acute face. The corolla is
ubsalverform with the stout cylindrical tube not surpassing the calyx
lo The var. albatum is merely a form with an abundant appressed
white pubescence. In gross aspect it is reminiscent of coarse forms of
rocumbens e collectors of both varieties speak of the corolla

as white or sordid ae and of the plant as prostrate or as a spreading
ren
56. "E. ‘Andexmonii Robinson, Proc. Am. Acad. xxxviii. 192 (1902).
Sarcanthus asperrimus Anderss. Stockh. Akad. Handl. 1853: 209
854). H. asperrimum Anderss. Freg. Eugenies Resa, Bot. 86
(1861); not R. Br. (1810).

4

-_~
i

ECUADOR. Gatapacos Istanps: Indefatigable, 1852, Andersson (G,
FM, K, BD; tsorypss).
Endemic to the Galapagos Islands whence it is known only from a
single collection.
57. H. oxylobum, sp. nov., —— 3-6 dm. altum sparse
strigosum cum pilis arcte adpressis; ramis cilibus numerosis;
foliis alternis lanceolato-linearibus me oblanceolatis 1-2.5 cm. longis
1.54.5 mm. latis firmis costatis sed enervatis margine paullo revolutis
basem versus in petiolum brevem ad 1 mm. longum attenuatis apice
acutis subtus pallidis; spicis solitariis vel geminatis ebracteatis graci-
libus 2-6 em. longis; calycibus ascendentibus 1-2 mm. longis; pedi-
cellis ca. 0.5 mm. longis, lobis linearibus vel lineari-lanceolatis sub-
aiconboas: corolla flavescenti hypocraterimorpha extus dense strigosa
intus supra medium sparse opel villosa 4-5 mm. longa, tubo
cylindrico 3-4 mm. longo 2-2.5 mm. diametro, lobis ascendentibus
ovato-deltoideis conspicue acutis; aa medio tubo affixis
- inclusis; antheris linearibus ad 1 mm. longis sessilibus apice obtusis
glandulari-puberulentis leviter coherentibus; ovario glabro vel puberu-
lento; stigmato ca. 1 mm. longo cylindrico hispidulo cum disco crasso
et angustissimo ornato quam sty lo ad ta Oo nuculis strigosis
ca. 1 mm. longis dorse convexis intus angu
PERU. Apwurrmac: a of Apurimac, Jan. sie 2400-2700 m., Pearce

u Rio Man Sureubam

(K, Trp; BM, tsoTyPe ee ee taro below ba;
ea. 1250 m., 1913, asics

58 JOHNSTON

Very closely related to H. campestre, a species extending north te
La Paz in Bolivia, from which, however, it differs in its more slende
habit, smaller leav es, more appressed pubescence and pubescent inne:
surface of the corolla. Weberbauer gives the flower-color as sordi
yellow.

58. H. campestre Griseb, [PI]. Lorentz. 186] Abh. K. Ges. Wiss.
Géttingen xix. 234 (1874) and [Symb. Argent.] |. c. xxiv. 271 (1879);
Giirke in E. & P. Nat. Pflanzenf. iv. Abt. 3a, 97 (1893); Kuntze, Rev.
Gen. iti. pt. 2, 205 (1898); Fries, Ark. Bot. vi. no. a 23 (1906).
H. andinum ey Mem. Torr. Bot. Cl. iv. 224 (189.

Central Argentina to middle western Bolivia, and cE also
in northern Paraguay.

ARGENTINA. San Luts: Queb. Salado, 1882, —— (BD). Cérposa:
Capello del Monte, Kurtz 6625 (NY); Alto Gracia, Parodi 7443 (G); Sierra
de Achala, Parodi Riga ( vues oo — near Cirdcba. 1877, Hieronymus (K);
Cérdoba, Loreniz 150 an BD), Hieronymus (G, NY, Us, FM, BM.
BD), Fielding ( BM). are “NY. oS BD). Caramarca: Fuerta de Andal-
galé, Schickendantz 92 (G, K); Andalgal4, 1917, Jérgensen 1076 (G, US);
Andalgaldé, 1915, Jérgensen 1076 ( US); Andalgal4, White 94 (BM); Rio de.
Valle, Castillon mm 1097 (G). _Tucumn: San Pedro de Colalao, 1200 m., Venturi
3731 (G); ¢ © de Medina, 1400 m pee 15999 Bee Traneas, 736 m.,

6 m..

i
289%

hear '
eae 2800 m , Rodrigues 1278 (G). Juscy: Loma del Tambo, Voledn,
2200 m , Sehreter 2600 (G); indefinite, C i 9644 (G). INDEFINITE:
Kindion, 4 (NY).

LIVIA. ‘Tana: H uayavilla, 1800 m., Fiebrig 2506 (G, US, K, BM,
BD); along Rio Frou near Ft. Crevaux, Fries 1593 ( US). CocHas
Parotani, 2400 m., March 20, 1892, Kuntze (NY, US); Tunari, 1600 me

929 (N £ of H. andinu

(US). emir aee no locality given, Fiebrig $21 3218 K. B) D).

PARAG see nate Rio Apo and Rio Aquidaban, Fiebrig 4016 (US,
BD) and got (G, BD).

A very characteristic, though somewhat variable species most
nearly related to the Peruvian H. orylobum. It also approaches
Ging less closely H. angustifolium Torr. Bot. Mex. Bound. 137
(1859), of northeastern Mexico and adjacent United States.

59. H. mendocinum Ph. Anal. Uniy. Chile xxi. 400 (1862),
Linnaea xxxiii. 186 (1864) and Anal. Univ. Chile xxxvi. 190 (1870);
Griseb. [PI. Lorentz. 186] Abh. K. Ges. Wiss. Géttingen xix. 234
(1874) and Wik Argent.] |. ¢. xxiv. 271 (1879); Giirke in E. & P.
Nat. Pflanzenf. iv. Abt. 3a, 97 (1893); Kuntze, Rev. Gen. iii. pt. 2,
205 (1898); Hicken, Physis i. 30 (1912); Sanzin, Physis iv. 43 (1918).

Middle » western Argentina.

»

|
\ STUDIES IN THE BORAGINACEAE 59
|
| ARGENTINA. : Monte Gorman near Ramacaida, Rio Dia-
te, Kurtz 7217 NY Poa an lat. 34° 1500, m., Kunize (NY, US);
ravesia, Miers (BM); = rral del Tra ravesia, Miers (BM);
Mendoza, ca. mat Ge King fgets Mendoza, » Gillies 528 (K); Mendoza,
Stibel (BD); Maniaia Diaz (MS (MS, Tre; G, photo.); Mendoza, 1885, Goldsack

BD); F , 2000 , Schreiter 90 (G); Andalgal4, Jdrgensen
1074 (G, ts): nes near Fuerte de Andalgal4, Lorentz 363 (BD) and 364 (G, BD);
{Campo de Pilciao, Schickendantz 172 (NY, K, BD) and 292 (G, BD); Cam

po
ide Pilciao, White : 95 in pt. (BM); Catamarea, Castillon 1105 (G). TucumAn:

‘Amaicha, Valle Ca tehaquies, 2000 m., Schreiter 559¢ (G); Valle de Amai-
‘cha, 2090 m., Castillon 2450 (G); Arenal ui o las Arcas,
‘Dept. Trancas, 2000 m., Schreiter 2517 (G); Tucuman, 1837, Tweedie (K).
{Santraco: La Banda, "Till 6090 6090 (G); Santiago, Gillies 1. CérpoBa
ah Cérdoba and Las Salinas Grandes, 1876, Sitibel (BD). Saawecren:

cenzie, Gillies (K); no locality, Hosseus 1258 and 1263 (BD).

» This species is characterized by its white, or occasionally rosy or
| yellowish, funnelform corollas and 4 equidistant non-didymous very
‘hairy nutlets. It is very closely related to H. Greggii Torr., Bot.
| Mex. Bound. 137 (1859), of northeastern Mexico and shunias
United States
| 60. H. catamarcense, sp. nov., anyon ramis 1—1.5 dm. longis
_gracilibus ramosis laxe ascendentibus pallide strigosis; foliis lance-
olatis 1-3.5 em. longis 2-5(-6) mm. latis strigosis 1-3 mm. longe
petiolatis, margine paullo revolutis, apice acutis, basi abrupte con-
| ieeetin: infimis oppositis ceteris alternis; floribus racemosis vel rariter
ibus; racemis lateralibus ne em. longis cum foliis non suffultis
| ebracteatis vel pauci-bracteatis; bracteis caducis filiformibus 2-3
| mm. longis dum adsunt juxta vel nie infra pedicellos lateralibus;
pedicellis ascendentibus 2-6 mm. longis; lobis calycis anguste lanceo-
_ Jatis 3-5 mm. longis ca. | mm. latis fructum valde tibus
| Paree: sieigonts; corolla alba infundibuliformi 7-12 mm. longa, lobis
- oblo: ascendentibus, faucibus Plicatis, — tubo intus a
antheris inclusis medio tubo ti
senate! stylo 0.5-0.9 mm. longo s stigmate subaequilongo; fructu

|

700. Apel 2, 1917, Jorgen ines ta my Campo de
m., arpusiet Type; U “ISOTYPE ;
_ Pileiao, Jan. Bi shee 172a 1 (G, BD).

a

|

60 JOHNSTON |

This plant most suggests H. mendocinum in flowers and habit, but
has the fruit of H. crea naenyy It seems to be most closely relat
to the latter, as shown by = poner, fruit and non-thickened

corollas are quite piltceit from ene of that species. I donot believe
on it is a hybrid.

um Ph. Anal. Univ. Chile xxi. 401 (1862)
teas xxxiil. 187 (1864); Crack, [Pl. Lorentz. 187] Abh. oe
Wiss. Got ttingen xix. 235 (1874) and [Symb. Argent.] |. e.
(1879); Giirke in E. & P. Nat. Pflanzenf. iv. Abt. 3a, 97 > (1803
Hicken, Physis ii. 113 (1916). H. auratum Ph. Cat. Pl. Itin. Tarapaca |
56 (1891). Tournefortia ? aurata Ph. Anal. Univ. Chile xe. ed
(1895).

Western Argentina.
NY US OM. & — — 2 of seoeager ge — 101 G,.
gees i given |

:

'. Caramarca: C. del povek 1917, Jorgensen 1723 (G, US); San )

Schreiter’s plant nt the province sen Tucuman is given as | }
white corollas. The other collections listed evidently have golden -
i more

yellow flowers. In all the material ex: is
elongate than in H. mene has a firmer texture with the
color of the inner surface more vivid than the outer.

}

acutis 2-5 cm. longis” 26 mm. latis laxe eee ann
sparse oe hispidis basem versus in petiolum 1-25 mm. |
attenuatis margine | revolutis subtus set
lanceolatis vel
ceols to-linearibus 4—5 m Stas hisp me Bd ornatis; 1s ee
2 mars ascents lobis peter am herbaceis sparse hispidis
aye Tanceolatis. vel lanceolato-linearibus conniventibus 3-4 mm. longis,

STUDIES IN THE BORAGINACEAE 61

glaberrimo; stigmate 0.3-0.6 mm. longo subsessile, columna brevi a
pilis sparsis rigidis erectis longi-penicillata, disco ca. 0.6 mm. lato;
puculis ca. 1.5 mm. longis in rostrum breve attenuatis dorso valde
convexis sparse hispido-villosis in utroque facie fovea elliptica notatis.

» COLOMBIA. Maepaiena: Valle Santa Marta, July 1844,
Purdie i inh — ay gre wrdie (K); = 1849,

Purdie ; ta M iguana near e tora,
Allen 24 (K); — set (K). "cas DE SANTANDER: Ocafia, 1850,

fe gross habit suggesting H. ternatum Vahl, but differing in its
annual duration, large differently shaped corolla and subsessile hirsute
stigmas

_ 63. H. filiforme Lehm. Géttingische gelehrte Anzeigen 1817: 1515
(Sept. 1817); an Act. Acad. Caes. Leop. Nat. Cur. ix. 128 (1818)
and Asperif. i. 37 (1818); Kunth, Flora i. 603 (1818); HBK. Nov.

‘iv. Abt. 3a, 97 (1893); Chodat, Bull. Herb. Boiss. ser. 2, v. 483 (1908);
‘Pulle, Enum. Pl. Surinam 399 (1906); Herz. Medel. Rijks Herb. xlvi.
10 (1922). Schleidenia filiformis Fresen. in . Bras. viii.
pt. 1, 40 (1857). H. tenue R. & S. Sout, j iv. 737 (1819); Cham. I. C5
'Fresen. |. c. 41. Preslea stenostachya St. Hil. Voy. Dist. Diam. ii.

: zuela and the Guianas; dion in Beitid Hidavas anil ot Svbled.

BOLIVIA. Et Bent: Reyes, Amazon Basin, 300 Cardenos 1381 (G,

1016, Steinbach { ie teen i Rune oti 3, Sicinach 3181
? ee eB

a 8 Rio Palometil-

: TTO BD);
near Cujaba, M: 312 (BD) is Rio in uarha, Moore 428 (BM);
iP osrgpege Sieh Noy Ee K. BM, BD); indefinite,
Nov. 1908, Hoehne 822 (BD). ie Nati vidade, Gardner "3361 (NY,
8311 (G, US. K) 3163 (0, 2 DOR (Ge Ay jindein ome ¢ . ee a
iy +: ?) INAS .
indefinite, St. Hilaire B1492’ (NY), Claussen’ ans K) 302 (BM). Rio
(BD), 13031 es) 3 Pari: Monte Alegre, Traill 70 (K); Rio Tre
Sees E ri, Traill 568 (K); Santarém, Spruce 400 0 (hk); Liman Spruce :

62 JOHNSTON

400 a Awazowas: Lip so Traill ee AS cogs , Spruce 1115 Ag kK.
Inp 0 locality given, Claussen NY), Sellow ( G, US. k , BD),
Martius (B (BD), Herb. Mus. Vind. 1587 iK) pe beens (Herb. Willd, TYPE
tenue).

BRITISH GUIANA: Essequibo River, 1881, Jenman 1095 (K); Esse~

BM eed 321 (BD): indefinite, Schomburgk 228 (K) and 351 (US,
Bourvar: Tigrito, Passarge & beara 615 eo

Botillns ge & Selwyn S08 awe INDE : ones mbold
1202 (Herb, Wil Willd, type of H. filifo

umboldt (Herb. Willd).

A very distinct species which somehow has been frequently confused

ith H. procumbens and H. fruticosum. From both species it is
readily separated by its subsessile stigmas and subulate or filiform
bracts. It is a slender annual herb which commonly dries dark.
The pubescence is usually fine and closely appressed. The type
came from Venezuela along the Rio Apures between El Diamante
64. Heliotropium margaritense Hassler, in herb., gracile laxe,

SEIU PE’) ry

ramosum sparse strigosum ut videtur annuum; ramis 5-15 em. longis:
ca. | mm. crassis; foliis linearibus vel lineari-oblanceolatis 1-2 em. 7

ae ad. mam, long hak Malt: iemade ics ta: €4 ba Saany
disco ad 0.4 mm. diametro, columna ad 2 mm. longa; nucellis ca. _
a 05 mm, longs dors conveis spare tren ventre angulatis per- —
ce PARAGUAY: Caleareous region on the upper Rio Apa, Hasler 10998
(BD, tyre; K, ISOTYPE). — .
Oby ssl related to H. fliforme Uhm. bat differing in its coarser

sppe ear Tn addition the leaves of

STUDIES IN THE BORAGINACEAE 63

Hl. filiforme are gradually tapered to the petiole whereas those of
H. margaritense, while tapering toward the base are abruptly con-
tracted just above the petiole. The derivation of the specific name
is uncertain. It is to be noted, however, that Hassler, Fedde, Repert.
xi. 167 (1912) has used the name for a plant collected near “Cerro
Margarita.”

65. Heliotropium ocellatum Cham. Linnaea iv. 463, t. 5, fig. 5
(1829). Schleidenia ocellata Fresen. in Mart. Fl. Bras. viii. pt. 1,
41 (1857).

’ BRAZIL. Inpertnrre: “Bras. merid.,” Sellew (US, K, BD, 1sorypss).

ARGENTINA. Misiones: Posadas, Barranca del Rio Alto Parani, 1884,
Wiedrtein 100 (BD).

A rare plant, apparently most nearly related to H. fruticosum and
rhaps only a southern variety of that species, but differing in its
ile stigma, nearly glabrous corolla-throat and copiously bracted

pikes. :

66. Heliotropium polyphyllum Lehm. Neue Schrift. Naturf.
es. Halle iti. pt. 2, 9 (1817), Nov. Act. Acad. Caes. Leop. Nat. Cur.
ix. "136 (1818), Asperif. i. 63 (1818) and Icones 11, t. 8 (1821); Cham.
Linnaea iv. 462 (1829). Schleidenia polyphylla Fresen. in Mart. Fl.
Bras. viii. pt. 1, 36 (1857). H. foliosum R. & S. Syst. iv. 737 (1819).

Var. genuina. Leaves oblanceolate, usually broadest above the
middle; corolla-lobes somewhat ligulate. —Synonymy given above.
_ Venezu Guiana.
_ VENEZUELA. Botivar: ae, te to, Passarge & Selwyn 514 (BD).
_GuArico: ae ee ee , Grisol 17 and 27 (NY).

APURE: Apure, Humboldt 808 (Herb, Willd., TrPE of H. foliosum); Rio

Apure, Bonpland (BD); Orinoco, no collector given (BD).

Var. Blanchetii DC. Leaves lanceolate, usually broadest below
the middle; corolla-lobes deltoid-ovate.—Prodr. ix. 544 (1845).
Preslea linifolia St. Hil. Voy. Dist. Diam. ii. 162 and 433 (1833).
Schleidenia linifoha DC. l. ce. 558. 4H. linifolium Giirke in E. & P.
Nat. Pflanzenf. iv. ote 3a, 97 gence # k. Sanienee DC. L. ec. 544.

bahiensis in Mart. FI. .

3 eoprig fo Rao Js k 3812 (
Boss Soe eae ee
ee: . j }; DEtW. ; § ¥ 2 OL CetGw (Ei);

64 JOHNSTON

US, FM, K, BD; tsoryrs of var. Blanchetii); Bahia, Lockhart (BD), Saltzmen
(FM, K; ? sotyre of H. Bahi lense), Blanchet 92 (NY). AvacG6as: near

6073 (K, Ohi ae Sedkiog I 162 (B D). gg oiacnmiae no locality given, Sellow
(US), Gardner 1365 (K), Blanchet 977 (BM).

All the South American material referrable to H. hag
appears to be prostrate-spreading and to frequent sandy places. The
Brazilian var. Blanchetii differs from the typical plant of the ark
Valley only in the form of its leaves and corolla-lobes and seems to be
most common on dunes along the coast. Its type is said to have
come from Serra Jacobina in Bahia, some 300 km. inland. One
wonders, however, if the plant actually came from that place since
all the other material from Brazil has been collected on or very neat
the coast. Perhaps they may have been some confusion of data i
this particular collection of Blanchet’s. The color of the corolla i
H. ‘polyphullum seems to be either white or yellow. corolla.
lobes of the Brazilian variety are broader than in the typical form bu
veutes to those plants of Cuba and Florida, cf. Addisonia iv. t. 133

H.
— and that they will eventually be treated as varieties of
—
7. Heliotropium ‘i ldenens Chodat, Bull. Herb. Boiss.|
ser. af i, 817 (1902).
_ . PARAGUAY: near Igatimi, Sierra de Maracayt, Hassler 4761 (G, K,
BM, BD; tsoryres).
Clearly related to H. Hasslerianum tee tag ts i ayer
referred to it are three puzzling collecti so from P: y.
collection from Sierra de Amambay, Hasler 1 isto (G, kK, BM, BD), ,

| The name 1. celia Hass, Ad Bi | oay zk "18 (1917), nomen
- nudum, has been applied to it. Another collection, Hassler 9827
— BD), also from the Sierra de Amambay is similar to the

preceding but is — * throughout and has smaller
associated with with H. pact is Fiebrig 6325 (G, BM, BD),
collected “‘in e fluminis Alto Parandé’” having small flowers and

a ‘Tn habit this last ee ont ecient

STUDIES IN THE BORAGINACEAE 65

in? these regards quite similar to H. Hasslerianum. I have not seen
the underground parts of Fiebrig 6325 nor of Hassler 4761. Helio-
tropium Hasslerianum, var. rigida Chodat & Hass., Bull. Herb.
Boiss. ser. 2, v. 483 (1905), based upon Hassler 7109 from Valenzuela
is unknown to me. It is said to be distinguished from H. Hass-
lerianum by having rigid, lanceolate and narrower leaves with blades
2.5-3 mm. broad and 12-15 mm. broad. Fiebrig’s collection men-
tioned is the only material from Paraguay which I have seen that fits
the description of the variety. The three collections which I have
associated with H. Hasslerianwm are so different from one another
and from the type of H. Hasslerianum that one naturally wonders if
they do not each represent a distinct species. Since, however, they
are intimately related and we have no idea as to the extent of vari-
ation which they exhibit in the field, I feel it is best to leave the col-
lections at least temporarily associated wi janum.
Certainly they are much closer related to that cheno than to any
other.

68. Heliotropium pallescens, sp. Nov., _erectum; cautibus igi
dm. altis solitariis vel plurik li
vel supra medium laxe ramosis herbaceis vel rariter basem versus
frutescentibus saepissime annuis; foliis caulinis alternis enervatis

sparse strigosis vel rariter glabratis, subtus strigosis pallidiori-
al spicis 1-4 dm. longis lateralibus et terminalibus cum bracteis
Gvato-oblongis vel ellipticis acutis 3-5 mm. longis distantibus ornatis;
ealyce tubo corollae vix breviori breviter pedicellato, lobis valde in-
aequalibus lanceolatis vel oblongo-lanceolatis strigosis; corolla alba
infundibuliformi ca. 7 mm. longa extus sparse minuteque strigosis,
lobis ovatis ascendentibus, intus in faucibus sparse villosis; antheris
medio tubi insertis apice coherentibus; ovario glabro; stylo 0.3-0.7
fm. longo; _stigmate 05-08 mm. longo, disco lato, columna cylin-
globosis ca. 2 mm. diametro minute strigosis;
nuculis dorso convexis is apice non rariter breviter acuminatis ventre
in utraque facie medio punctato-ocellatis.

PL Hasler. Par Loma Clavel, Gran Chaco, lat, 23° 20° 8. grag the

Maine 672 (BD 8 "Alto saeco a oo 21 8 Fiebrig

1 °o

1810 (ie, BD ; BD); Pu t. 1892, Kuntze ( ; Estancio
torzede de fe Mae Robert 841 (K, BM, BD); lower

-eourse
af the Bio Pileomao, Rojee 109 (BM BD); 1998 (NY. FM). near Asuncion,

Balansa 2035 (K); Pileomayo River, Morong 1528 (NY, ).

66 JOHNSTON

This is a very readily recognized species that is related to, and ths

elongate leaves, erect habit and apparently short-lived root.
plant is annual or at mest biennial. The collection made by Balansa,
however, is somewhat fruticose below. Unfortunately the root of
this collection has not been seen. Most of the specimens seen have
dried a very light color. None of them has become dark-colored as
is generally the case with H. polyphyllum. The plant i is very finely
and rather sparsely strigose throughout, though in consequence of
the light color of the plant this pubescence is not a Some
of — hairs on the leaves have minute bulbous ;
9. Heliotropium distantiflorum Hassler, Add. Pl. Hass. i. 17
( oe
PARAGUAY: vicinity of Caaguazt, Hassler 9023 (G ES);
iy de Amambay, _ Hassler 9777 (BM, BD), 9777a ‘BD Bong i094 9094 (BM.
The name H. detitilorass Hassler, is a nomen nudum published
by Hassler in a pamphlet entitled “Addenda ad Plantas Hasslerianas”
which a examined at the British Museum of Natural History. The
name is merely associated with two of Hassler’s eplicttiewanebnen:
viz, 9023 and 9777. The plant dries a light brownish and is erect
with one to several stems from a thickened perennial or biennial root
The stems and leaves are covered with short stiff very closely a
hairs, most of which — thickened bases. The neares'
relative of the species seems to H. pallescens, also from Paraguay)
from which it differs in its coarser = — revolute.
margined desivee, smaller flowers and less e
70. Heliotropium fruticosum L. Syst. ed. i. ii. 913 (1759) and
Amoen. ie v. 304 (1759); Lindm. Ark. Bot. ix. no. 6, 3 sees
(?) H. humile Lam. Tab. Encye. i. 393 (1791); Poir. Encye. S
agp un, Ee Neue Schr. Nat. Ges. Halle iii. pt. 2, 11 (1817)
and Asperif. i. 43 (1818); R. & S. Syst. iv. 38 (1819). _H. amanaik
-anum Lehm. Nov. Act. Acad. Caes. Leop. Nat. Cur. ix. 137 (1818)
_ and Asperif. i. 69 (1818); HBK. Nov. Gen. et Sp. iii. 86 (1818); Urb,

_ Symb. Ant. iv. 527 (1910) and 1. ec. viii. 589 (1921). _H. campechiense

Willd. exR. &S. Syst. iv. 731 (1819). Sehleidenia ovalifolia Fresen. in

l. e. 97; not Forsk. (1775). H. phyllostachyum Torr. Bot. Mex. Bound.
__ 137(1859). H. myosotoides Chapm. FI. So. U.S. 330 (1860), not Lehm.
(4817). fe ene & P. Nat. Pflanzenf. iv. Abt. 3a.
s sccmagen :

STUDIES IN THE BORAGINACEAE 67

Colombia and Venezuela northward in the West Indies and Central
America and Mexico to southernmost United States; local in eastern
Brazil.

BRA ZIL. Pravuy: dry sandy a Esperanca, 1839, Gardner
2261 ee BM;; tsoryPes of > Sati) "2250 “ay,
Hacienda de near San Joaquin,
gh 797 7 Ws). Se vietity a Dividive, Pittior 1 108.
te NY, US). Zura: Veras Altas, road from Maracaibo to eo: Markintes:

Piitier 10501 (G, NY. US); Maracaibo, Moritz 1312 (K, BM, BD).
COLOMBIA. To : Honda, 27. : 20a (NY); Mari-
Ae ta, de Ibaque, 1750 m., Triana 3751 (BM); spate 1200 m.,
(BM); betw. Toiaima and Ibaque, Stibel 194 in pt. (BD :
, oa. m., man K). VALLE:
Cali, ¢ m., Pennell 5849 (G, NY, US); Epinal belo : ;
Killip & Hazen 11062 (G, NY); La Paila, Holton (G, NY, K). Cauca:
p , 1876, André 2920 (NY, K). 1 .
April 27, 1876, André 2920 (NY, K). CunpinaMaRcA 700 m.,
853, Trian ); poi m., Triana (BM). Macpa.ena:
1926 (US); Rio Frio betw. the Cienaga ta ‘
m., Piitier 1613 (US); Santa 1 ., Smith 543 (G, NY, US,
ee a ; co do, Magdalena V;: near Lake Allen
196 and 422 TLANTICO: quilla, Bertero Seg’ , ca. 5 m.,
Pennell 1 NY, US, K, BD). Inperiire:

2063 (G, ! i BT.
685 (NY, K); Patico, Rio Ags ea. 1200 m., Lehmann Isa? (K, BD); in-
definite, Lehmann 4044 (BM).

‘Although quite inappropriate, the specific name H. fruticosum L.
obviously mgs to the present species. The name was founded
upon a description of a collection made by Browne in Jamaica. This

Tks en teat a occ “ Heliotrop um minus,
Lithospermi foliis”’ is a mixture consisting two plants of. the present
species and one of the perennial, H. ternatum Vahl. Sloane’s illus-
tration is obviously drawn from one of the two plants representing
the species here treated. Since both the specimens and the illus-
tration studied by Linnaeus unquestionably belong to the weedy
annual herb and not to the coarse perennial plant t (H. ternatum
Vahl), as supposed by past writers, it seems clear that the name
H. fruticosum should be taken up for th B
& Wilson, Bot. Porto Rico ii. 136 (1925), have recently done.

The name H. humile Lam., the subject of very diverse in’
tions, I doubtfully refer here. It has been frequently assumed th
be is based upon Tournefortia humilis in, but this I believe to have

which H.humile
is established, Lam. Tab. sere i. 393 and 417 azn, the name,
egeaess lmaaoccamre, valid species. The plant from _

68 JOHNSTON

the Royal Gardens upon which Lamarck, Encye. iii. 94 (1789),
apparently based his H. humile was first described two years earlier
as representing H. fruticosum. A consideration of rie description
makes me suspect that the plant is the herbaceous annual which I am
calling H. fruticosum

The material from the interior of Colombia differs from the typical
Caribbean form in usually drying darker, being slightly coarser in
habit, in having a loosely appressed or somewhat spreading pubescence
and in having less conspicuous bracts. Eventually it will be given
some sort of taxonomic recognition, but first its relations to #1.
strictum HBK. must be determined. In many details it seems to be
intermediate between H. fruticosum and the form of H. ternatum
called H. strictum.

71. Heliotropium salicoides Cham. Linnaea viii. 117 (1833).
Schleidenia salicoides Fresen. in Mart. Fl. Bras. viii. pt. 1, 38 (1857).
H. Clausseni A. DC. Prodr. ix. 543 (1845). S. Clausseni Fresen.
lc. 38 t. 11. H. rigidifolium A. DC. 1. ¢. 543. H. strictissimum
Moric. ex DC. |. c. 544; Moric. Pl. Nouv. Amer. 146, t. 87 (1846).

sohemncote acer heh H. glomeratum A. . c. 550.
s. glomerata Fresen. |. c. 39. S. macrantha Fresen. co H.
macranthum Girke in -— ‘& F. de Pflanzenf. iv. Abt. 3a, 97 (1893).
S. incana Fresen. |. c. 38. incanum Fresen. ex Giirke, |. c. 97;
not R. & P. (1799). S. gi at te iia a Kjoeb. Vidensk. Meddel.
1867: 16 (1868). H. Fresenii Giirke |. c. 97. H. Clausseni, f.
robustior Chodat, Bull. Herb. Boiss. ser. 2, ii. 816 (1902). H
maximum Chodat & Hass. Bull. Herb. Boiss. ser. ser. 2, v. 482 (1905).
H. hispidum, var. maximum Hass. Add. Pl. Hass. i. 17 (1917).

Bolivia,

Eastern and northeastern Argentina to Paraguay,
and Colombia. a
ARGENTINA. Saladas, Feb. 25, 1917, Hauman be
ee Entre Rios indefinite, Tweedie
1
TA. Sanr. Kuntze (NY, US); Buena-
vista, 500 m.,. Steinbach #178 ne 2 EM) ar ese oo BD); Campo de Buena-
wea : Steinbach 1 998 (BD); Canton de vista, 450 m., Steinbach
PARAGUAY: Sierra de ger es Hassler F wath (K) and 5300 (G, K,
BD); Rio Sse 9g fk Sierra de Maracayi, H 5954 (G, K, Bi BD):
; oo Marscayt, Hassler 3617 Gk. BM, PES of

BD;
. hisp. v. ee Yerbales, Sierra de Maracayd, Hassler 5295 (BM):
(K, BM. BD); Ry seus tae Gs hae a ee Wis Anaitaban, Pub
BD); ae Ghochi, Cord. de Altos, Fiebrig 432 (FM, K, —
}; north of the Cordilleras, Fiebrig 88. "BD); upper ‘Rio
7988 (BM) and 7980 (BM, BD; 1soryres of H.’ mazi
Comet, Hiemalor #418 (EMA). sete une

STUDIES IN THE BORAGINACEAE 69

BRAZIL. Hoag his newest ee near Cujabaé, Meyer 815 (BD).
Rio GRANDE Alegre, 1 Tweedie (K). fe ) Pauto: indefinite,
Regnell IIT Off (US, BD). Rio picid el indefinite u 13033 (

BM, BD), 13034 (K, BD). Mrnas GERAgs: hentia Regnell Til 911 (Us);
iamantina guia 8201 (BD); betw. Valle fondo and Melancias, Schwacke
tw. Barreiras eon
ta, hag aoeet ae ubracemosa); indefini

Claussen 229 (G, NY, F io, ytd 19678 and 20423 (K, BD), St Hilaire

MS ae Sener 1278 ‘ayy, B. 2138 (K). Goyaz: betw Palmital and Amorei
(NY); definite, Gardner 3911 (NY, US, K, BM), a (K),

NY, EM. K, BM, BD; ISOTYPES of H. strictissimum), 3606 (FM, BM);
Tamandis betw. Jacobina and St. _—_ Blanchet 3843 _ — of
ag esa igidifolium). Pra
4(K). Inp given, Burchell 5613 3(K), 5910-2"
. 6664 (G, K), 3051 "K): § alee Ul 3 K BD; isotype of H. s

Hes. Vind. 1583, 1586, 1589 (K, BD), 1590 (1 BD; isorye EOFS. caine’.

P sgt BIA. SanTANDER: Mesa de los tos’ 1500 n “ng Killip & Smith
15228

This i is a variable assemblage, which can not be sharply separated

hirsute or strigose-hirsute abr, This is catiads the case with
the material from the southern parts of the range. Typical H. sali-
coides is a form with a dense plush-like somewhat aureate pubescence.
Heliotropium Clausseni has a coarse, not dense, hirsute pubescence
and thickish leaves. It is a common form. Heliotropium strictis-
simum, H. rigidifolium, H. glomeratum and H. subracemosum are all

tion by Gardner, no. 4295a, is also canescent but it has a soft
felt-like pubescence and may te only a broad-leaved phase of H.
Fumana. Heliotropium Clausseni f. robusta and H. maximum
ee Clausseni.

Heliotropium Vahl, Symb. Bot. iii. 21 (1794);
ig Bice Suppl. ec 25 (1815); Lake, Sapo ae (1818). Pioe-
tonon ternatum Raf. Sylva Tellur. 88 (1838). H. fruticosum, var.
ternatum DC. Prodr. ix. 542 (1845). Tournjotia tunthe 1. Sp.

PL ed. 1, 141 (1753) and I. c. ed. 2, 202 (1762); Lam. Tab. Encye. i.
417 (1791); R. Br. Prodr. 497 (1810). H. humile R. Br. ace. R. & S.
> eatin 37 (1819), not Lam. (1791). H. hirtwm Lehm. Neue Sehr.

Naturf. Ges. Halle iii. ghia 10 (1817), Nov. Act. Acad. Caes. Leop.
Nat. Cur. 2 a tae OD a legisla R. &S. —_ we

70 JOHNSTON

38 (1819). H. Ottoni Lehm. Nov. Act. Acad. Caes. Leop. Nat. Cur.
ix. a (1818) and — i. 59 (1818). H. Ottonis DC. Prodr. | ix.
: m HBK. Nov. Gen. et oP. ii. 87 (1818);
C. Saree ix. 545 sis). H. hispidum HBK. 1. ec. 87 (1818) and
lL. c. 451 (1820). H. fruticosum, var. hispidum DC. 1. c. 543. Schilei-
denia hispida Fresen. in Mart. Fl. Bras. viii. pt. 1, 37 (1857). (?) \H.
demissum R. & S. Syst. iv. 37 and 733 (1819). TT. monostachya Willd.
ex R. & S. 1. c. 540; HBK. |. c. 452 (1820). H. Ottonianum R. & S.
Syst. iv. 733 (1819). P. antillanum Raf. Sylva Tellur. 88 (1838).
H. fruticosum, var. confertum DC. 1. ¢. 542. H. fruticosum, var. ed a
tilobum DC. 1. ¢. 543.
Northern South America northward to Mexico and the Greater
Antilles.
UELA. Méripa: Laderas de San Pablo, 600 m., Jahn _ 0 (US)
——, 1000 m., Jake: “grt (G, US). Lara: north of
Saer 6 (US).. Caranoso: vicinity of Sega ca. 600 m., Pittier 8890 290 (G.
NY, Us, — a beeing San J — fhe 800 o hae 8229
( ‘A: Coloni ar, Fendler 907 909 K). Dist. F
La Guaira, 1874, Farsi (NY XY) "Cabo Blanes, Collen: gr Playa de Tube
co, Pittier 10189 be pil , US); Gamboa savannas, 1000 m.,
Pitter Povet (US); Mi iddie Cotiza near Cardcas. cas, eager m, P
(US); Savanna Grande near Cardeas, 1855, Birachel (K); Cardeas, 950 m.,
Bailey 111 (US); — Humboldt (Herb. Willd.; typE of amet 8
H. hispidum). “Clednscnti no locality gi ven, Funcke 543 (BM).
BRITISH GUIANA: near Kamakot, soegtonea 1884-5, Jenman 3 (US).
This is * common fruticose species of the West Indies which has
passed as H. fruticosum. In its typical form it is known from
South America only in Venezuela and British Guiana.. It is distin-
and commonly ill-defined relatives by its-
range, pale eoloeasid | in being somewhat scabrid and closely though
not densely strigose. The plant from the region about Cumana,

dispidun is typical of the Caribbean form. The form common a shout:
a to be also the West Indian plant. Material collected
‘there by Humboldt, however, differs — from the typical form

and shows H. salicoides. It also
evaluat material. ius the interior of Colcnken which I have ten-
; referred to true H. fr names H. . o.

‘o "Humbolde’s colletion from Cares. The most striking variation

t

rather than
leaves. This is known from Venezuela and comes from

thee sow Caracas. dom -cmemenass by Fendler 909, Pittier

Peet aaye ee Onae a CRP RR Ny le. Ws Vey AOU OR eT) page tk, ma

STUDIES IN THE BORAGINACEAE 71
8890, 10189 and by the collections by Kuntze and Gollmer cited above.

\73. Heliotropium Fumana (Fresen.) Giirke in E. & P. Nat.
PHanzenf. iv. sige 3a, 97 (1893). Schleidenia Fumana Fresen. in
Mart. FI. Bras. viii. pt. 1, 40 (1857).

Brazil and aa British Guiana

‘BRAZIL. Min. Regnell ITI 912 ag , BD);
indefinite, Glaziou miei (KB BD) ot St. andaire B. 1952 bis (K), St. Hilavre (K).
Bawia: Jacobina, Blanchet 3903 (BM ); Joazeiro, Chase 7944 ‘oy gr Sve
indefinite, Gardner 3360 (K). Pravny: betw. Canabrava and d Boa Baperance,
Gardner 2260 (G, NY, US, K); betw. Cachoeira and Marmaliero,

2426 (K, BM, BD). Inverrrre: no locality given, Herb. Vind. 1588 (K,

“BRITISH GUIANA: Ireng pore fas raima, Quelch & M
and 302 Ag Rupurunie, Appon 2203 (K); soe a Nov. 16, ony er rhe
3 (US, K , BM); no locality, Schomburgk 189 (K).

The plants referred here appear to be intermediate between H.
ternatum and H. salicoides, though more closely related to the former
and perhaps constituting only a Brazilian phase of that species. The
concept as here taken is highly unsatisfactory and will no doubt be
eventually much changed or entirely done away with. The plants

have leaves which are linear to lanceolate and densely — especi-

ally béneath. It is pale and is not scabrous. The mature calyx is
ca. 2.5 mm. long and pale-strigose.

EXCLUDED OR Usineeneh SPECIES.

HYTUM CRISPULUM Fresen. in Mart. Fl. Bras. vii. pt. 1, 45,
t. 1; f 3 (1857). Heliotropium crispulum Giirke in E. & P. Nat.
Pflanzenf. iv. Abt. 3a, 96 (1893).—Near Rio San Francisco between
‘Tapera and yi cas Minas Geraes, Brazil, Martius.—Possibly a phase
of H. angiospermum
Hewopnyrum oporum Fresen. in Mart. Fl. Bras. viii. pt. 1, ind
(1857). Heliotropium odorum Giirke in E. & P. Nat. Pflanzenf. i

. Abt. 3a, 96 (1893); not Balf. f. (1884). — . ‘ragrans
‘Macbr. Proc. , Brazil,

. Am. Acad. li. 541 (1916)—Minas Geraes
gern a phase of H. angios permum
- PASSERINOIDES Klotzsch in By aah Fauna & Fl.

OPHYTUM

: Ney rere 1152 (1848), nomen nudum.—Auf der Savanne in der

‘Umgebung des s sprepsnees Strauch, British Guians, -caeieenis ale a

-Pessibly H. Fumana.
HELIOTROPIUM

Aisweecanine Si Gard. Diet. ed 8, no. 11. (1768)—
Unrecognized.

fe Seat en by the laie De. Houstoun fromVera Cruz.
i. not find it.

oe ws is = Ssi eI et othe fly. a

V2 JOHNSTON

HELIOTROPIUM BRASILIANUM Roth, Nov. Sp. Pl. 103 (1821)+—
Said to have come from Brazil. It belongs to the section Orthostachys.
_ HELrorropiuM cirriroLium Lehm. Géttingische gelehrte inne
1817: 1515 (1817); Nov. Act. Acad. Caes. Leop. Nat. Cur. ix. 127
(1818); Asperif. 1. 32 (1818); R. & S. Syst. iv. 729 (1819). —* Habitat
in America meridionali (v. s.).”—Apparently a Tournefortia and
probably Brazilian

Heliotropium eectein Vell. Fl. Flum. 69 (1825) and Icones ii. t. 42
(1827)—TournEFoRTIA sp.

Herorropium GraciteE R. Br. Prodr. 493 (1810).—This Ola
World species was reported from Brazil by Don, Gen. Syst. iv. —
(1838), probably upon a misdetermination.

Heliotropium humifusum of Walp. Nov. Act. Acad. Caes. ‘oan
Nat. Cur. xix. Suppl. 1, 371 (1843); DC. Prodr. ix. 542 (1845). —This
West Indian species is reported by Walpers as occurring in South

Meyen’ s collection, the basis of this report is cited thus,—
= ® Persia! in planitie circa Tissalomam. Fl. mense Aprili.” At Berlin
there is a ccllection by Meyen determined as H. humifusum and given
as collected April 1831 at “ Pipiloma” at “15000” ft. This collection
Js Coldenia paronychioides Ph. A study of Meyen’s account of his
journey to Lake Titicaca makes it seem clear that the locality men-
tioned by Walpers and on Meyen’s label both were intended for the
locality in extreme southern Peru called Piscacoma, cf. Meyen, Reise
i. 463-472 (1834).

Heliotropium lithospermoides Willd. ex Lehm. Nov. Acad. Caes.
Leop. 2 bd ix. ~~ (1818), nomen; R. & S. Syst. iv. 337 (1819).—

iment tila Bull. Soe. Bot. France lvii. Mem. 3e}
179 (io, sc helaeeinl Sitio and Barbacena, Minas Geraes, Brazil)
Glaziou 15272—The above is a nomen s stdmactiains. the only descriptive
rt g that the plant is subfrutescent and has pale yellowish

flowers.
. ee FI. Peruv. ii. 3, t. 109b (1799).—
Vaeikasoncien ip. .
Heliotropium oppositifolvwm Rts P. Fi. Peruv. i ii. 2, t. 108b (1799).—
ToURNEFORTIA sp. \
- _Heliotr tum Vahl, Symb. Bot. iii. 25 (1794). —PHACELIA
| MAGELLANICA Gan) Coe,
Hen C. Prodr. ix. 540 (1845)—“in Mexico
crea urbe Maton jlo for. legit el. Berlandier! (pl. exs. p. ees

bs a

= 2 ye : The ee a ti . Di. ZL o2ee
oo ‘pl exs. 2666), e ae 1 B 2 234, ok het

7
1
}
}
|
\ STUDIES IN THE BORAGINACEAE is

is clearly H. procumbens, and I suspect that H. rigidulum will eventu-
ally drop into the synonymy of that species.

-HELIOTROPIUM RIVINOIDES Hassler, Add. Pl. Hass. i. 17 (1917).—
This nomen nudum is associated with Fiebrig 1439 from Paraguay.
I have not seen this collection.

HELIOTROPIUM ROTUNDIFOLIUM Sieb. ex Lehm. Icones 19, t. 30
(1821).—This is a species of the Levant. At the New York Botanical
Garden, however, there is a plant evidently representing it which is
associated with one of Linden’s original labels indicating that the
plant was collected in New Grenada in 1842-3. It bears Linden’s
number 1594. I suspect that there has been some confusion in
labeling the collection.

Heliotropium scandens Vell. F\. Flum. 69 (1825) and Icones ii. t. 41
(1827).—TouRNEFORTIA sp

Heliotropium scorpioides Willd. ex Lehm. Nov. Act. Acad. Caes.

. Nat. Cur. ix. 140 (1818), nomen; R. & S. Syst. iv. 737 (1819);
not HBK (1818).—LiTHOsPERMUM DISTICHUM Ortega.

Heliotropium stylosum Ph. Bot. Zeit. xxviii. 500, (1870) .—NEso-
CARYUM STYLOsUM (Ph.) Johnston.

HELIOTROPIUM THYMIFOLIUM Vahl. ex Lehm. Neue Schr. Naturf.

Ges. Halle iii. pt. 2, 17 (1817); Asperif. i. 67 (1818); DC. Prodr. ix.
543 (1845)——The source of the type is unknown, but is probably
West Indian

Vilicietpiom tarmense Krause, Bot. Jahrb. xxxvii. 632 (1906).—
TOURNEFORTIA sp.
ENIA DASYCARPA Fresen. in Mart. Fl. Bras. viii. pt. 1, 37
(1857). Heliotropium Marti Giirke in E. & P. Nat. Pflanzenf. iv
Abt. 3a, 97 (1893). Heliotropium dasycarpum Fresen. ex Giirke, |. —
Rio San Francisco near Joazeiro, Bahia, Martius; state of Piauhy,

2262; state of Bahia, Blanchet 3903.—I have only seen
Gardner 2262. This is a phase of H. paradoxrum. t at
least some of the other collections t H. polyphyllum Lehm

2. Nores on vartous BoraGINomeEae.

Cryptantha Werdermanniana, sp. nov., annua 5-15 em. alta
hirsuta ascendenter ramosa robusta; foliis li ineari-clavatis eneuae
hirsutis minute pustulatis 1-3 cm. longis | 24 _ bt onde ;

Ceres otngP

Sasa thege we ahcia mina V9 os

|
74 JOHNSTON |

costa cum. setis fulvescentibus horridis et in marginibus hispido-v il-
losis; corolla inconspicua ad 4 mm. longa alba, lobis elliptico-obovatis
ascendentibus ad 1 mm. longis; ovulis 4; nuculis 4 homomorphis 2.7-

clausis basem versus divaricato-furcatis in areolam triangularem
= abrupte pra epsom stylo ca. 1 mm. longo quam nuculae con-
icue longiori—CutLe: Cerro de la Copa, Dept. Taltal, 2300 m.,
Sc 1925, Ilse Francks | in Pl. Chile. Werdermann. no. 1043 (TYPE,
Gray Herb.). —A remarkably distinct new species which keys out in
my revision, Contr. Gray Herb. -Ixxviii. 33 (1927), with C. globulifera
(Clos) Reiche and C. peruviana Johnston. It is probably mest
y related to C. globulifera from which it differs in the size, form
aad roughenings of the nutlets, long style, much larger calyx, gener-
ally more robust habit and very detached northern ran,
back of the nutlet is tuberculate with the tubercles very eucrendia:
_ There is no evident transverse arrangement of the roughenings as in
C. globulifera or C. diffusa. The new species comes from the salitre-
ee ee in about
lat. 25° 4’ S. and long. 69° 50’ W
CrypranTHa Ponpim Greene. Eeectly branched rather slender
strigose herb 1-3 dm. tall; stems tan-colored, antrorsely strigose;
leaves not numerous, linear, obtusish, 1- veges ace. eer em. emo. oan
1-2.5 mm. broad, more or
or ternate (or quadrinate acc. Greene), | 1-2 em. i rather « crowded
and frequently glomerate, irregularly inconspicuously and shortly
. _leafy-bracted throughout; Core ev sebepnie _— spreading, 2-3 mm.

- deciduous, appressed silky-hirsute; matare calyx-lobes Siscee.
Locos connivent, midrib thickened, unarmed; ovules 4; nutlets 4
_ or not agaamitad a subhomomorphous with the axial (?) one

y

ae deliattely angled: grobee nearly dosed: medial, aleapGy bot
arro’ ly forked at base; gyn sptobase slate, 24-34 height of nutlets;
deci 4]

Giana Greene, L e ii 232 ): Johnston, Contr. Gray

—Pittonia i. 291 (1889).

STUDIES IN THE BORAGINACEAE 75

Herb. Ixxiv. 103 (1925).—Lower Catirornia: Bay of San Bartolomé,
1889, Lieut. Pond 22 (Greene Herb., ryre of C. Pondii; US, 1soTyPe);
Bay of San Bartolomé, 1889, Lieut. Pond (Greene Herb., tyre of
C. Bartolomaei).—During the summer of 1925 I was able to visit the
Greene Herbarium now at the University of Notre Dame and took
the Oppor tunit. y to examine the types of C. Pondii and C. Bartolomaei
These plants are obviously the same species. They were collected
at the same locality, on the same date and by the same collector and
probably represent different portions of one collection. In 1925,
1. ce. 62, I incorrectly cited C. Pondii as a synonym of C. patula Greene,
since the plants in the Gray Herbarium, received from Greene and
labeled by him as part of the type-collection of C. Pondii, are clearly
representative of his other species, C. patula, and, as seems probable
now, part of the type-collection from San Benito Island. The plant
in the National Herbarium at Washington, labeled as C. Pondii, has
not been subjected to imislabeling and clearly represents C. Pondit
Greene. The species is a very distinct one of uncertain affinities.
It is probably best placed in a special Series between Graciles and

CRYPTANTHA MOHAVENSIS Greene, Pittonia i. 120 (1887). C. fallax
Greene, |. c. v. 54 (1902); Johnston, Contr. Gray Herb. Ixxiv. 104
(1925) —The type of C. fallax has been uneed. The species is
obviously a synonym of C. mohavensis.

CRYPTANTHA crrcumscissa (H. & A.) Johnston, Contr. Gray Herb.
Ixvili. 55 (1923). Piptocalyx circumscissus Torr. in Wats. Bot. t. King
Exped. 240 (1871).—ARGENTINA: annual with white flowers, in
near Me chat cheel, sb aig Nov. i, 1925, H. F. Comber 135 (G,

western United States. The list of indigenous borages which occur

rge and > and reappear in United States now consists of
Fotis linearis (R. & P.) DC., Pectocarya pusilla (A. DC.) Gray,
Cryptantha albida (HBK.) Johnston, C. cireumscissa (H. & A).
Johnston, Plagiobothrys fulous (H. & A.) Johnston, Lappula Redowskii
(Hornem.) Greene, Lappula texana (Scheele) ae. caeateaial vir-
ginica (L.) BSP. and Coldenia Nuttallii Hook.

Plagiobothrys (§ ? Allocarya) plurisepalus (F. v. Muell.), ak
nov. Annual; stems prostrate or ascending, 3-15(-20) em. long,
Darcy wale ye onaae es eninge mee strigose or some-

erous to the base; leaves sublinear, very

Fl tapered ova the ae pockn cae am 2

76 JOHNSTON

obtuse, strigose or appressed hispid-villous, occasionally somewhat
pustulate, lowest ones crowded into an evanescent rosette; flowers
numerous, loosely racemose; bracts linear, foliaceous, remote, extra-
y or even oppositiflorous; calyx ca. 2 mm. long in flower, more
or less villous, divided into 5-8(-9 “fide Muell.) linear lobes; fruiting
ealyces strongly accrescent and asymmetrical, 5-9 mm. long, their
lobes conspicuously indurated and thickened at the base and closely
investing the fruit, upper parts curved or much contorted; mature
pedicels ascending, 1-2 mm. long, coarse; corolla inconspicuous, white
but drying brownish, ca. 2 mm. long, subtubular; corolla-lobes 4-6,
ovate-orbicular, ascending, small; corolla-throat with weakly in-
puberulent areas alternating with the lobes; stamens 2-5,
inserted just below ie middle of the tube; ovules 2 or less commonly 4;
nutlets 2 or less commonly 4, oblong-ovate, heteromorphic, frequently
asymmetrical, pale or somewhat fuscous, the back and sides rounded
and roughened with a system of more or less transverse loosely
reticulate (frequently high narrow) ridges which anastomose with the
medial dorsal keel and below the middle frequently develop at the
expense of it, apex frequently cornute, ventral surface keeled promi-
nently to just below the middle; areola cuneate in outline, submedial,
en below the crest of the keel; axial nutlet usually slightly smaller
and broader, dulled with very minute papillae or spicules, most
persistent; style reaching to about -% height of nutlets.—Maccoy
a F. v. Muell. Frag. Aust. i. 127 (1859). Rochelia lars
lb ae Rep. Bot. Exch. Cl. Brit. Isles iv. 644 (1917). R. Mac-

ae — Nov. stesis R. wd me — — Aug.

STUDIES IN THE BORAFINACEAE . rye

able variation, while perhaps characteristic for the species, yields no
generic characters, inasmuch as the range of variation exhibited
includes that which is normal for Plagiobothrys. This may be appre-
ciated by a study of the tabulation given below. The. collections
dissected are indicated by the letter on the left. The collections
cited above have been lettered in the order that they are arranged,
thus, for example, “G” refers to Mr. Black’s collection from Quorn.
The material studied is scanty, but opportunity was usually found
for making at least three dissections of each collection.

Collection Ovules Anthers Corolla-lobes Calyx-lobes
A 2 4 4 5
B = 2 4 5
C 2 3-5 5 5
D z 4 4? or 5? 7
E 4 3 5 5
F 4 3 5 5
G = 4-5 5 cf
H 2 3-4 5-6 7-8
i 4 a 5 5
J 2 3 5 7
K 2 4-5 6 7

Z2or4 2-5 4-6 5-8

The number of floral parts furnishing no differentiae it is evident
that the plant must be distinguished by structural differences. The
most important of these is the thickened and indurated ates
which much suggest those of Cryptantha texana (DC.) Greene and
relatives and to a less extent P. glaber (Gray) Johnston and its
relatives.

In P. plurisepalus the great bulk of the plant is inflorescence, con-
sisting of elongate loosely flowered false racemes (helicoid uniparous ©
cymes). The plant is floriferous almost to the very base, and the
_ stem proper very much reduced. Th wd
into a loose rosette. I am unable to state definitely that the lowest
leaves are opposite, though I have some indication that they are so.
(aceon Cle aoe ee
settling of this point. As all the sp y pposit
lower leaves, though the p Rona ne £ iL. Pt ay 2S the
shortening of the stem and REIN ee aa
it is of particular interest that the presence or absence of them be

Cees “baniuingaenet gata cao

78 JOHNSTON

Some of the collections I have cited have been determined as

thickened calyx-lobes and elongate somewhat heteromorphous
nutlets. With the calyx-lobes, corolla-lobes and stamens showing
erratic departures from the normal number in the group, it is not at
all surprising to find the ovules behaving in a similar though less
extreme manner. Personally I am surprised that they do not vary
more, and that plants with 3 ovules or merely a single one have not
n fi In any event the number of nutlets is at best a
character since even in forms normally producing 4 ovules one to
several are frequently aborted and a reduced number of nutlets
matured.

For the opportunity of examining a series of specimens representing
this and the following species I am particularly indebted to Mr. J. M.
Black of North Adelaide, South Australia, who not only shared his
specimens with me, but also won the interest of Mr. J. W. Audas of
Victoria, who also generously provided me with material. To both
of these gentlemen I would here express my cordial thanks for their
important assistance without ae the notes on these Australian
pees could not have been prepare:

bothrys (§ Allocarya) Ss (F. v. Muell.), com
nov. Annual; stems one to several, erect or prostrate, Gacloding “e
inflorescence) 5-15 cm. long, simple or with a very few long simple

linear ilvase lobes: trait lyx 2-3(-5) mm. long, lobes herbaceous,
linear; pedicels thickened, ‘ca. 16 i. long, ascending; corolla incon-

ae spicuous, white but drying brownish, 1.8-2 mm. long, only shortly

the calyx; limb narrow, 0.8—1 mm. broad, with 5 ascending
obovate lobes; throat bearing 5 weakly intruded puberulent areas;
stamens 5, borne below the middle of the tube; ovules 4; nutlets 4,

ay oe NSPIcuouUsly develope
eee i ee res ss eg ie - Gn wee ;

STUDIES IN THE BORAGINACEAE 79

tomose with the keel ——? below the middle; gynobase pyra-
midal; style reaching to /s/s height of nutlets Heliotropium
elachanthum F. v. Muell. Linnaea xxv. 424 (1852).—Victorta:

Wattville near Dookie, Oct. 1922, A. B. 0’Dowd; Wimmera, 1890,
Mueller; Wimmera, Dallachy; Lara, 1923, A. F. Gates; Dimboola,
F. Reader. SoutH Austra: Pt. Lincoln, Oct. 1909, H. Griffith;

Rocky Creek, Oct. 1851, Mueller—A well marked species previously
confused with P. eusiedlavicns from which it is readily separated by the
characters of the nutlets. The body of the nutlet above the base is
somewhat drawn out obliquely downward. On the crest of this 3 puck-
ering is found the areola. In this character th

ships with the section Echidiocarya, as does also the non-compressed
body of the nutlet, its reticulate ridging and somewhat vitreous peri-

carp. The fruit, however, is by no means so definitely stiped as in
Echidiocarya.

Thanks to the courtesy and interest of Mr. J. F. Rae of the National
Herbarium of Victoria, I have had the privilege of examining a portion
of the type of H. sleclumnthie: Like the other collection from South
Australia, i i. e. Griffith’s from Pt. Lincoln, Baron Mueller’s type is
coarser in habit and has slightly larger fruit than the plants I have
seen from Victoria.

Praciopoturys (§ Allocarya) austraLasicus (A. DC.) Johnston.
Annual; stems one to several, erect or prostrate, (including the in-
flarescence) 5-12 cm. long, usually with several strictly ascending
simple laterals from below the middle, sparsely strigose; leaves linear,
1-3 cm. long, 1-2(-2.8) mm. broad, obtuse, spreading, glabrous or

sparsely appressed short-hispid, lower ones opposite, margins ciliate
and frequently pustulate; racemes 3-8 cm. long, remotely flowered,
interruptedly bracted, simple; calyx in flower ca. 1,5 mm. long,
appressed villous-hispid, divided into 5 erect lance-oblong lobes;
fruiting calyx 2-2.5 mm. long, the lobes lanceolate, herbaceous and
ascending; pedicels thickened, ca. 0.5 mm. long. , ascending; corolla. -
inconspicuous, white but drying brownish, ca. 1 3 mm. long, shortly
surpassing the calyx; limb 0.7-1 mm. broad, with 5 ascending obovate
lobes; stamens borne just below middle of tube; corolla-throat with
5 weekly 4; nutlets 4, pale or fuscous,
opaque, ca. 1.3 mm. long, compressed-ovoid, angulate, roughened
on the back and sides with loosely reticulate ridges hed anastomose
with the well developed medio-dorsal keel, ventral face with a large
areola sunken below the crest of the keel, ventral keel extending down
to about middle of ng nutlet and LE and crowded on either
side by wrin in , more

+ eS

80 JOHNSTON

or less excavated, about ?/; length of nutlet, not oblique; gynobase
pyramidal; style reaching to about */; height of nutlets. ee

Gray Herb. Ixviii. 75 (1923). Eritrichitum australasicum A. DC.
Prodr. x. 134 (1846). oe australasica Greene, Erythea iii. 57
(1895) —WeEstTERN Aust : Swan River, 1839, Drummond.

This species is a ate rete to Western Australia, the forms
from eastern Australia passing ing as this species being in large part at
least P. elachanthus. It is evidently quite distinct from that species
differing in its opaque, more compressed nutlets and very different

what excavated and is located on the lower half of the nutlet evidently
sunken below the level of the ventral keel. It appears to be rather
closely related to P. sacpoes? (Colla) Gray of Chile.

Plagiobo talis (L.), comb. nov. Heliotropium ori-
entale L. a 7. isi (1753); Houttuyn, Linn. Pflanzensyst. v. 402
(1779); Lehm. Asperif. i. 73 (1818). Lithospermum javanicum Steud.
Syst. i. 547 (1825). Eritrichium plebeyum, var. tenue Herder, Act.
Hort. Petrop. i. 542 (1872), excl. pl. amer. Allocarya asiatica Kom.
in Fedde, Repert. xiii. 236 (1914). P. astaticus Johnston, Contr.
Gray Herb. Ixxiii. 68 (1924)—The type of Heliotropium orientale L.,
since its publication unidentified, is clearly the Kamtchatkan species

tic
the plant was examined by the late i Daydon Jackson, who informed
me that the plant was presumably collected by Steller in Kamtchatka.
This would agree with what we know of the distribution of the species
_ represented. More authentic material of the species appears to be
Sika Lindon , Ark. Bot. ix. no. 6, pg. 3 (1909),
_ ives the source of this material as “leg. Konig in Asia ” The accur-

Mba ie Mise ot pe eS. ruc ON ee a eee POSE Tae a theta Sire heap ees st eg Ot [= ai St ee ASF ee it TaN ea

STUDIES IN THE BORAGINACEAE 81

arious authors after Linneus considered the plant to have come
from Java. Steudel even went so far as to rename it Lithospermum
javanicum. I can find no mention of the plant in any of the literature
on the fiora of Java. The basis for attributing the plant to Java
appears to be the statement, “ Diese ist . . . in Ostindien, vornimlich
in Java zu Hause,” which is found in the German edition of the
Systema commonly attributed to Houttuyn (for complete citation

. XXV).

Plagiobothrys gracilis (R. & P.), comb. nov. Myosotis gracilis
R. & P. Fl. Peruv. ii. 5 (1799). Echinospermum gr
Asperif. i. 129 (1818). Rochelia gracilis R. & S. Syst. iv. 111 (1819).
Pectocarya gracilis Johnston, Contr..Gray Herb. Ixx. 36 (1924).—
The type of Myosotis gracilis is not a meee which all authors for
almost a century have assumed, but is instead definitely a species of

Plagiobothrys of the section Meas The original collection is
said to have come from Concepcion, Chile. The specimen from Ruiz

of four consimilar nutlets all armed with evident glochidiate subulate
appendages. In P. procumbens the nutlets are all unarmed or only
the axial nutlet is armed with short carrie appendages. Just
how P. gracilis is to be distinguished from P. Greenei, if indeed it can
be distinguished, has not been determined.

Pectocarya platycarpa Munz & Johnston, comb. nov. P. gracilis,
var. platycarpa Munz & Johnston, Contr. Gray Herb. Ixx. 36 (1924).—
It having been found that the name Pectocarya gracilis (R. & P.)
Johnston belongs in the synonymy of a species of Soe asa the

‘distinct from Pectocarya linearis (R. & P.) DC. =o ee

— ee stiffer habit and different range.

minutus (Wernh.), — nov. Lithospermum mi-
eet Wernh. Trans. Linn. Soe. 2, Bot. ix. 118 (1916).
Plagiobothrys minutus Johnston, Cos 6 Gray Herb. Ixxiii. 68 mah
I have examined the type of this species at the British M
The material i sant but seems clearly t represent» species of
a ee

82 JOHNSTON

HAVILANDIA BORNEENSIS Stapf, Trans. Linn. Soc. ser. 2, Bot. iv.
209, t. 16a (1894). Lithospermum borneensis Boerl. Handl. FI.
Nederl. Ind. ii. pt. 2, 488 (1899). Bs aecaa borneensis Johnston,
Contr. Gray Herb. Ixxiii. 68 (192:

HavILANDIA PAPUANA Hemsl. su Bull. 1899: 107 (1899).—The
genus Havilandia was incorrectly referred to Plagiobothrys, its rela-
tiens, rather, appear to be with 7rigonotis, a genus which has a number

ever, in having nutlets with a definitely basal or very slightly supra-
basal attachment. The individual nutlets suggest those of Zoelleria.
ZoELLERIA PROCUMBENS Warb. Bot. Jahrb. xvi. 28 (1893).—I have
been able to study the type of this monotypic genus at Berlin.
entirely agree with Warburg, who stated that Zoelleria was aberrant
only as to the number of nutlets. The genus is related to Trigonotis
and Havilandia, which it clcsely resembles in habit and vegetative as
well as floral characters. The nutlets are smooth and obscurely

or > ace ,

Nat. Pflanzenf. iv. Abt. * 131 (1893). It is simply an aberrant
relative of Havilandia and Trigonotis and like them evidently be-
longing to the tribe every

_ Micro Rockii, sp. nov., perennis; caulibus erectis vel decum-

glabratis supra sparse adpresseque villcsis, inferioribus oblanceolatis
vel ol longo-oblanceolatis 2-3 cm. longis 0.7-1 cm. latis basem versus
gradatim attenuatis, caulinis 5-18 mm. longis ellipticis vel obovato-
ellipticis distantibus ascendentibus; floribus in glomerulos 5-8-floros

ecllectis ad apices ramulorum gracilium paucifoliatorum ei

ea. 1 mm. longis, meteeene paullo accrescentibus ovato-lanceolatis
: ad 2 mm. longis; pes Se ee
_ vel ascendentibus; ‘cache Hutisine caerulea conspicua 6-10 mm.
- diametro, lobis obovato-orbicularibus iii Ga peace iay tubo
_ subeylindrico apicem versus paullo gradatim ampliato 2-2.5 mm.
longo 15-220 mm. crasso fusco calycem — meperet fornicibus

STUDIES IN THE BORAGINACEAE 83

affixis; antheris oblongis ca. 0.7 mm. longis quam filamentis subulatis
2-3-plo longioribus; ovario glaberrimo 4-ovulato; stigmate capitato;
nuculis 1-4 horizontalibus vel ascendentibus pallidis plus minusve
spiculiferis 2.5-3.5 mm. longis basem versus sparse rugosis vel cristatis
dorso medio-longitudinali cum areola 2-3 mm. longa angusta promi-
nenti ornatis (marginis areolae obtuse dentatis) ventre per areolam
medialem vel paullo submedialem parvam ad gynobasem convexam
adfixis—T1BET: moist meadows of Wanchen nira, between Labrang
and Yellow River, alt. 3300 m., July 29, 1926, J. F. Rock 14511 (TPs,
Gray Herb.); wet meadows of Dzomo la, in alpine region between
Radja and Jupar range, alt. 3300 m., July 1926, Rock 14384 (G).
Kansv: alpine meadows in mountains west of Adjiian, east end of
inshan, T’ao River Basin, alt. 3750 m. and lower, July 5, 1925,
Rock 12605 (G).—An extremely well marked species characterized
by its large very pale-blue corollas, very fine appressed and incon-
spicuous pubescence, glabrate lower leaf-faces and very elongate
dorsal areola of the nutlets. In lateral view the nutlets much suggest
those of a true Eritrichium, but the surface of the nutlets and the
nature of the dorsal areola point unmistakably to Microwla. The
species is known only from eastern Tibet and southwestern Kansu.
Microula trichocarpa (Maxim.), comb. nov. Omphalodes tricho-
carpa Maxim. Bull. Acad. St. Petersb. xxvi. 500 (1880) and Mél.
Biol. x. 681 (1880); Brand, Pflanzenr. iv. Fam. 252, 105 (1921).
—I have had the opportunity to study isotypic material of this
species at Kew. It is obviously a Microula and appears to be most
related to M. myosotidea. A collection from Gargannar in south-
western Kansu collected by R. C. Ching, no. 919, is arch to the
species.
Microula Forrestii (Diels), comb. nov. Omphalodes Forrestii
Diels, Notes Royal Bot. Gard. Edinburgh v. 169 (1912); Brand,
iv. Fam. 252, 105 (1921). M. hirsuta Johnston, Contr.
Gray Herb. lxxv. 48 (1925).—The type of M. Forrestii cad 2 Aavenl
the same mountain-mass and are obviously synonymous.

84 WEATHERBY

II. A NEW SPECIES OF JUNCUS FROM COLOMBIA
By C. A. WEATHERBY

Juncus Andreanus n. sp., perennis, rhizomate perbrevi, radicibus
fibrosis ; culmis dense caespitosis - gracilibus basi non_incrassatis

longum apice gerentibus, superioribus laminis subteretibus canalicu-
latis gracilibus culmo plus minusve brevioribus instructis; auriculis
brevibus subcoriaceis obtusis; inflorescentia pauciflora valde glomerata
circa 1.5 cm. lata vix ultra 1 cm. alta, a bractea culmo simili plus
minusve superata, ramis brevibus patentibus; floribus prophyllatis,
bracteis scariosis late ovatis acutis vel obtusiusculis; perianthii
segmentis consimilibus 3-4 mm. longis ovato-lanceolatis acumin ctis
medio rubro-castaneis actate ‘brunneis margine hyalinis pallidis
lentit perianthio duplo
brevioribus, antheris circa 1 mm. longis filamentis subaequantibus;
ovario fere uniloculare, placentis in parietibus subsessilibus; capsulis
ovoideis truncato-obtusis stramineo-brunneis vel superne castaneis
textura ex comparatione tenui perianthio subaequantibus vel paullo
brevioribus; seminibus oblongo-ovoideis ecaudatis circa 0.3 mm.
is.—CoLomptra: Facatativié, Dept. Cundinamarca, alt. 2680 m.,

ja 16, 1875, André, no. 624, type in Hb. Gray; meadow in snbuaut,

t. 2600 m., Zipaquira, Cundinamarca, Oct. 20-24, 1917, Pennell,
no. 2558 (bcictbe a 4 as Pe ‘aeboliediond: prarien slope of Paramo de
las Coloradas, eons Santander, alt. 3900 m., Jan. 27, 1927,
Killip & Smith, no. 18483

Juncus i

from it as follows. capillaceus Lam., of the southern Andes, has

the bases ofthe culms some thickened; a very long bract making
the inflorescence appear lateral; green, merely acute or obtusish

anager taomrig the inner broader with narow median band and

o | seements a strongly a sre idee | 2 _ glob pic, apiculate, woody

SOME UNDESCRIBED AMERICAN SPERMATOPHYTES 85

Ill. SOME UNDESCRIBED AMERICAN SPERMATO-
PHYTES

By I. M. Jounston.

igostyla, gen. nov. Iridacearum. Perianthii segmenta basi

erecta et breviter connata deinde patentia multo inaequalia, 3 exteriora

oblongo-obovata conspicua unguiculata, 3 interiora inconspicua
Seccat

silibus erectis cum styli ramis ut videtur alternantibus. Ovarium
oblongum glabrum 3-loculare, loculis multi-ovulatis. Stylus intra
tubum stamineum filiformis, ramis 3 furcatis membranaceo-mar-
ginatis apice obtusis stigmatosis latere interiori sub apice flagellum
longissimum gerentibus. Capsula oblonga apice truncata loculicide
3-valva. Semina angulata rugosa opaca.—Bulbus tunicatus, Folia
pauca anguste linearia. Spathae oblongae pedunculatae terminales
vel laterales vix acuminatae. Flores in spatha pauci a
bracteis exterioribus spathae subsimilibus, satesiuatbes i
tenuioribus scariosis—Nomine a wdstryos, flagellum, et otdAos,

A very well marked genus of uncertain affinities. In the Natiir-
lichen Pflanzenfamilien, ii. Abt. 5, 144 (1887), it keys out as a member
of the Iridoideae-Tigridinae. Of the three genera of that group it

long contorted subapical fla . The stigmatic surface
apparently tips each lobe of the branches.

ir membranously so appear to be
very strongly compressed radially to the7axts.” is

breadth, and perpendicular to the two lateral wings of the lobe.
The stigmas reach to about the tips of the anthers and the flagella,
of course, much surpass them. I have been unable to determine
definitely whether the style-branches were opposite or alternate with
the stamens. fags latter condition, however, seems to be the case.

simplicibus vel furcatis nodis saepe 2; foliis 3-4 remotis 1040
peeents #9 rm nai 2 5 a te Senden soem oni

86 JOHNSTON

nervatis; bracteis spathae exterioribus 2 ca. 3 cm. longis acutis siccatis
plumbaceis vel purpurascentibus; perianthio azureo vel violaceo, lobis
imam ad basem breviter connatis in parte (5-7 mm. longa) infima plus
minusve erectis supra patentibus, lobis exterioribus 2.7-3 em. longis
1—1.2 cm. latis lamina oblongo-obovatis apice rotundis subulate 8-10
mm. longe unguiculatis, lobis interioribus lanceolato-linearibus 12-14
mm. longis 1—-1.5 mm. latis acutis; staminibus monodelphis, filamentis
in tubo 8 mm. longo confluentibus; antheris 5-6 mm. longis rectis; styli
ramis 3 infra medium bilobatis 5 mm. longis membranaceo-marginatis
1-1.3 mm. latis tenuibus, lobis ca. 3 mm. longis apicem versus flagel-
lum gracile 8 mm. longum contortum gerentibus, flagello ad latus
axillare styli 2-3 mm. longe decurrenti; capsulis 10-12 mm. longis 4-5
mm. crassis; seminibus brunneis 2-2.5 mm. longis undulato-rugosis
opacis sub lente densissime minutissimeque tuberculatis.—Prrv:
sandy pampa, 2400-3600 m. alt. on southern slopes of Chichani
north of Arequipa, March 1920, Hinkley 16 (type, Gray Herb.);
pen sandy soil above Arequipa, 2500-2600 m., April 1925, Pennell
13172 and 13173 (G).—A slender plant producing few narrow grace-
y recurved plicate leaves. According to Pennell the corolla is
violet and is either unspotted or spotted with a darker color. The

— gave the color as blue.

loxensis, sp. nov., frutex vel aie ramulis ri
juv

latis integris apice obtusis basi rotundis vel subcordatis supra lucidis
glaberrimis, subtus specie cum nervis ‘is perlaxe reticulatis prominulis

bris; petiolis
ascendentibus 1-3.5 cm. longis velutinis betes = racemis axil-
laribus conspicue ferrugineo-velutinis 15-21 em. longis sessilibus folia
evidenter superantibus; pedicellis 1-2 mm. longis basem versus con-
natis; aoeeriaes 7-10 mm. longis | subclavatis; lo lobis peends intus
_albis sicea’ i

t+ hor:

- lascastiae 3-3 case: Weis style plabro iar atiweiate sension subdiacras-
gato; ovario globoso dense ferrugineo-velutino; folliculis 2-25 em.
longis 10-12 mm. latis obliquis oblongi:
-Ecvuapor: between La Toma and Loja, 1800-2600 m., 5 oe 4, 1923,

Hitchcock 21414 (rvee, Gray Herb.); between Loja and Portovelo,
ae — esse Rose, Pachano “elagealesenioa Boggart ta dandiiedes
= Ecuador. It differs from that

2 Ae ge

SOME UNDESCRIBED AMERICAN SPERMATOPHYTES 87

maturitate glabratis et cinerascentibus, juventate brunneo-strigosis
pilis minutis et arcte adpressis; foliis glabris coriaceis alternis
4-6 em. longis 1.5-3 cm. lati ellipticis vel nae se bicolori-
bus, apice acutis vel obtusiusculis, basi acutis in petiolum canalicu-
latum gracilem 7-18 mm. ‘bau shane margine integris vel
irregulariter remoteque obtuso-dentatis, supra lucidis, subtus opacis
cum venis saepe distinctis tenuibus leviter ahnemnger es spicis
axillaribus multifloris densis 2-3.5 em. longis 1.5 em. erassis foliis
evidenter brevioribus, rachibus dense brunneo-strigosis; pedicellis
gracilibus distinctis 1-2 mm. longis strigosis; perianthiis albis
pea nigris) glaberrimis clavatis 6-7 mm. longis, lobis linearibus
versus paullo ampliatis; antheris angustis 1.5 mm. longis;
sph tenuibus glabris incluso stigmate clavato 5-6 mm. longis;
vario brunneo-strigoso; folliculis ignotis—-Ecuapor: between El
Tambo and La Toma, Prov. Loja, 1000-2200 m., Sept. 3, 1923,
Hitchcock 21366 bates Sand Herb.).—Related to R. complicata
HBK. but differing in its smaller leaves, short racemes and small
glabrous flowers.
Muhlenbeckia dumosa, sp. Nov., fruticosa 5-10 dm. alta vix
ramis ¢€

volubilis glaberrima ramosissima; rectis sulcato-angulatis
co brunneo obtectis, internodiis 1-2 em. longis; ocreis laxis diu
persistentibus; foliis coriaceis concoloribus e nervatis

15-35 mm. longa 6-15 mm. lata acuta basi cuneata et inde in jekilass
gracilem 5-8 mm. longum transmutata ; floribus fasciculatis vel soli-

ill
— 5-lobi lobis ellipticis ca. 2 mm. . longis rotundis patentibus;

b

oo
policnsitat nudis,. stylo brevissimo; pueTes: te. 2. 5 mm. - longo x

E, Gray .); open rocky
April 1925, Pennell phan 5 (G). Bouivia: La Paz, 3000 m., Bang 132
and ree ng (G). Tihs | getter confused with M. rupestris
Wedd. M. chilensis Meisn. The former is a very close relative
ee Sa a ee
angulate at the base. ee
twigs and its flowers are borne in well developed racemose

‘correct combination for it as not been published, and may be oS

88 JOHNSTON

Muhlenbeckia hastulata (Sm.) comb. nov. Rumezx hastulatus
Sm. in Rees Cyclop. no. 29 (1819). M. chilensis Meisn. in DC.
gm xiv. 148 (1856).

; sp. nov., annua herbacea 5-15 cm. alta laxe
ue ramosa; ramis laxe ascendentibus vel decumbentibus
exuéscennibes: infimis oppositis ceteris alternis, cum pilis as
tricis crassis subulatis vel conicis ca. 1 mm. longis villosis; —
ovatis vel oblongo-ovatis 2.5-6 cm. longis 1.5-3 em. latis flav
virentibus — patentibus subcarnosis trinerviis sata
papillosis et saepe sparse villosis apice obtusis basi subsessilibus
deltoideis; ‘floribus axillaribus solitariis; calyce monophyllo 4-5 mm.
longo quadrifido, lobis petaliformibus latis obtusis 3-3.5 mm. longis
ad 3 mm. latis externe villosis interne glabris luteis; corolla nulla;
staminibus 15-20; filamentis capillaribus flavis; antheris oblongis
flavis; stigmatibus 5-7; ovario conspicue villoso 5-7-loculato et
-ovulato; fructibus ad 5 mm. longe pedicellatis — obscure
quadricestatis ca. 6 mm. longis; seminibus 5-7.—Perv: frequent in
dryish places in and just below the fertile belt in the hills directly
back of Mollendo, Dept. of Arequipa, Oct. 16, 1925, Johnston 3553
(TYPE, spe Herb.).—A very distinct species readily recognized by
its villous pubescence and its obscurely angled 5-7-celled fruit. The
subulate or conical single-celled hairs are peculiar and are very
abundant on the younger parts of the stem and on the ovary.
Portulaca

sp. nov., elongata crassa
valida; caulibus gracilibus decumbentibus vel laxe ascendentibus 3-9
em. longis saepe ramosis ad axillas foliorum m perbreviter dense Pilosis

aliter glabratis, internodiis 2—5 mm. |

oblongis vel anguste — vel etiam spathulatis 5-10 mm.
longis 1-2 mm. latis carnosis pressis apice acutis vel rotundis;
capitulis bracteis = cuphylloideid oblongis vel lanceolatis 4-6 mm.

oe ik longioribus; capsulis sivas Gobads vel
; viter glo it —3 mm : - -

dehiscentibus parte superiore nitentibus cum calyce et corolla pe AE
rentibus emarcidis coronatis, parte i inferiore stramineis subdiciform-
di tuberculis

ih cum

ad 0.5 mm diametro dep is stellat diantit tik

notatis—PERv: open rock y slopes, Tingo, Dept. Arequipa,

Bein, Foul cae aly Gray Herb.); sandy pampa, so tes
. slopes of of Chichani above ve Arequipa, Hinckley t (G); — 1892,

SOME UNDESCRIBED AMERICAN SPERMATOPHYTES 89

A. E. Douglas (G).—This species belongs to the group of P. pilosa L.
and is characterized by its perennial habit, purple or red flowers and
basally circumscissile capsule. Though the minute stellate scales
which cover the seed are not prominent, they seem to stand out
individually more plainly than in those species in whick the crowded
scales are drawn up into a strong boss or tuberculation. The whole
scale in P. peruviana seems to be thickened rather than merely
umbonate as in most species.
Calandrina adenosperma, sp. nov., perennis glaberrima glau-
robusta ad 4.5 dm. alta; coulibes erectis subsimplicibus
pluribus sabieetiinn basem versus 4—5 mm. crassis e caudice prostrato
fruticoso crasso ramoso erumpentibus; foliis carnosis sed non cras-
sissimis ad basem caulis laxe rosulatis, supra rosulam valde reductis
remotis saepissime 2; foliis rosulatis obovatis vel oblongo-ovatis 5-10
em. longis 3.5-6 cm. latis integerrimis apice obtusis vel breviter
eens a parte supra m medium basem versus in petiolum latum
grada ; foliis supra rosulam ovatis vel orbiculari-ovatis
subamplexicaulibus 2. 5-5 cm. longis 2-3.5 cm. latis apice obtusis;
inflorescentia racemosa 5—10-flora simplici vel furcata conspicue brac-
teata; bracteis oppositis scariosis orbicularibus vel ovatis amplexi-
caulibus ca. 1 cm. —_— pedicellis elongatis ascendentibus quam
bracteis 3-7-plo longioribus infimis ad 6 cm. longis, superioribus
brevioribus ; sepalis pe esas 8-10 mm. longis obscure delicateque
nervosis viridibus vel basem versus plus minusve scariosis apice
rotundis vel latissime obtusis; petalis elliptico-obovatis rubris ca.
8-10 cm. |
capsula subglobosa sepalis paullo breviori 3 (vel rariter 4)-valva ca.

of ravines on southern slopes of Chackaadk aheth of Arequipa, 2700
m., March 1920, Hinckley 35 (rype, Gray Herb.)—A member of the
group centering around C. grandiflora Lindl. It is characterized by
its small corolla, conspicuous bracts, glandular echinate seeds,
peculiar habit and isolated northern range.

collina, sp. nov.,.annua herbacea 5-10 mm. alta

abundanter breviter glan villosa laxe d.

laxe ascendentibus teretibus cum internodiis 1-2 em. longis; foliis

linearibus patentibus pressis carnosulis 1-2 cm. longis 1-2 mm.

latis cuspidulatis; stipulis ovatis scariosis acuminatis et saepe plus

minusve laceratis basem versus connatis ca. 1 mm. longis; floribus
; pedicellis gracilibus 1-2 em. longis vel reflexis

[say need ? i pgeened | 1, 4A mm

90 JOHNSTON

longis I- = nan latis obtusis viridibus margine scariosis; petalis
albis elliptic brevioribus ca. 1.7 mm. latis apice
rotundis. et integris; staminibus 10 paullo inaequilongis ovarium

tibus; filamentis quam antheris 6—-7-plo longioribus; ovario
ovoideo-ellipsoideo breviter stipitato; stylis 3 distinctis ad 0.8 mm.
longis; capsula ovoidea sepalis paullo breviori ca. 2.5 mm. crassa;

exalatis vel saepe cum margine acuto — tuberculatis vel rariter
glanduloso-muriculatis—PEru: sandy open place at lower edge of
the fertile belt in the hills back of Mollendo, Oct. 16, 1925, Johnston
3568 (type, Gray Herb.)—Apparently a member of the group of
S. diandra (Gurs.) Boiss., in which it is distinguished nd the length of
its petals, its 10 well developed stamens and roughened seeds. It is
the plant from Mollendo reported and figured by Weberbauer, Engler
& Drude, Veg. Erde xii. 144, fig. 9 (1911), as Drymaria molluginea.
This determination is most obviously incorrect for D. molluginea is
a glabrate glaucescent plant of Mexico and a true Drymaria having
bifid petals and 5 stamens.

r a congestifolia, sp. nov., perennis; caulibus prostratis
pluribus ramosis 2-8 em. longis e radice e valida erecta —
tibus a !

quam internodiis renee stipulis hyalinis albidis oblongis foliis
oe ilongis in pilis villosis longis profunde laciniatis basem versus
a —— 3-12 in cymam laxam parvibracteatam dispositis;
2 mm. longis ascendentibus; sepalis lanceolato-oblongis
eres rs glandulosis 4-5(-6) mm. longis 1-2 mm. latis apice
obtusis margine scariosis; petalis albis ca. 6 mm. longis et 4 mm. latis

oribus; stylis 3 basem versus coalescentibus, capsulis : sevalvis ca.

mm. longis ad 3 mm. crassis is ;
seminibus exalatis, maturis ignotis.—Penw +: prostrate o on epen places
oS sb Bk bi Mollend , Oct. 16, 1925, Johnston

as Be trees, Cay he es & related to S. fasciculata

Ph. and S. Stuebelii (Hieron.) Johnston ( Tissa Stuebelii Hieron. Bot.
. xxi. 308 (1895), but quickly distinguished by its prostrate

ve prs ofthe Hat ar onsen of hn
entacaena).

SOME UNDESCRIBED AMERICAN SPERMATOPHYTES 91

Alchemilla Parodii, sp. nov., annua multicaulis; caulibus pro-
stratis vel ascendentibus gracilibus villosis 3-15 cm. longis; foliis
flabellatis adpresse villosis, lamina 3-6 mm. longa 5-8 mm. lata basi
obtusa tripartita, segmentis 3—-4-fidis, icine oblongis vel linearibus;
stipulis inciso-dentatis amplexicaulibus villosis ad petiolum 3-5 mm.
longum adnatis; floribus oppositifoliis glomeratis parvulis 2-2.7 mm.

longis breviter pedicellatis, glomerulis 5—10-floris; hypanthio 1-1.3
mm. longo obovoideo basem versus attenuato extus 8-costato inter
costas glabras breviter villoso intus glaberrimo, 4 sepalis et 4 bracteo-
lis hypanthii consimilibus erectis vel ascendentibus 0.9-1.5 mm.

longis paullo inaequalibus villosis subulatis acutis; achaenio ovoideo
solitario; stamine ad fauces —— in discum angustissimum
affixa.—ARGENTINA: Manantiales, Prov. Buenos Aires, Nov. 12, 1925,
Parodi 6664 (type, Gray Herb.); Avellaneda, Prov. Buenos Aires,
Oct. 11, 1925, Parodi 6530 (G); Palermo, Buenos Aires, Oct. 17, 1926,
Parodi 7313 (G).—This interesting herb is a member of the section
Aphanes and much resembles A. arvensis L., with which it has been
confused. It is readily separated from A. arvensis, however, by having
subulate rather than deltoid or ovate sepals and hypanthium-bracte-
oles. In addition the sepals and bracteoles equal or surpass the

hypanthium in length and the pdangrayes averages slightly smaller
and more prominently ribbed than in arvensis. According to
Prof. Parodi the new species is widely pShesnue and frequently
very abundant locally on the virgin meadowy pampa in the Prov. of

urbed
peculiarities, give strong arguments for considering the shiek as
—< to Argentina.

With the species I associate the name of Dr. Lorenzo R. Parodi,
the well known and scholarly agrostologist of the University of Buenos
Aires. Prof. Parodi has been very generous and helpful in his relations
with the Gray Herbarium and hence it is with pleasure that I name
this plant after him and acknowledge sf appreciation for his many
kindnesses.

Wendtia miniata, sp. nov., fruticans : m. alta ramosissima; :

sis graci pubeacen
tibus; ramis vetustis lignosis tenuibus feiss | foliis numerosissimis-
: 7 mm 2. ce RM | . ie ae eS

. 4 *

=

t adpresso-villosis vetus- _

tate Gabrats 58 mum, logs 1-3 mm tie lt acts in e :

ea. 1 em. longe pedicellatis; invlcris 6-7 ma, Tn ners a

92 JOHNSTON

aliquid in lobos furcatis minute pubescentibus; sepalis oblongo-
lanceolatis ca. 7 mm. longis 2-2.5 mm. latis graciliter acuminatis;
petalis obovatis 1.4-1.5 mm. longis ca. 8 mm. latis apice rotundis vel
obtusis miniatis; staminibus 10, 5 cum filamentis 3 mm. longis alternis
cum filamentis 2 mm. longis; antheris ovoideis; ovario globoso dense
argenteo-strigoso; stigmatibus 3 (? vel rariter 4) ca. 3 mm. longis
lin tis; capsula ad 5 mm. longa 3- vel rariter 4-loculata
glabrata loculicidaliter dehiscenti, seminibus in loculis 1 (? vel 2).—
ARGENTINA: in gulches, Cuesta de Miranda near Chilecito, La Rioja,
ca. 2300 m., Jan. 30, 1927, L. R. Parodi 7821 (type, Gray Herb.).—
A very well-marked species differing from all known congeners in
the color and large size of its corolla. According to Prof. Parodi its
petals are vermilion. species seems to be most closely related
to W. calycina Griseb., but differs from it, in addition to the corolla-
characters mentioned, in its biseriate stamens, simple or merely
fureate involucral bracts, scantier pubescence, and smaller differently

Geranium Parodii, sp. nov., acaule perenne caespitosum humile;
radice verticali 5-8 mm. crassa apice breviter furcata paucicipite;
coe omnibus basalibos, petiolis 1.5-8 cm. longis minute apicem

saepe densissime subsericeo-strigulosis, laminis 1.2-1.8 cm.
tae cua strigulosis ambitu reniformi-rotundis palmate 5-7-
partitis lobis plus minusve lato- vel ovato-cuneatis et varie incisis,
lobo medio: saepe usque ad medium palmate * Stoveate, lobulis

ahlonoic coed Pack weeebe ul
€TIOLO

ad ad 3 a adnatis ores medio-longitudinaliter cdcauies parte Shc
lanceolata acu a ta; pedunculis solitariis basilibus unifloris ebracteatis

longis 2.5 mm. latis conspicue trinervatis; petalis albis ad 12 mm.
longis oblongo-obovatis —— ca. 2.5-plo longioribus; staminibus
pistillisque calyce brevioribus ; fructibus maturis ignotis—ArGEN-
TINA: Pampa de Achala, Sierra de Achala, Prov.
Cérdoba, ca. 2200 m., Dee. 4, 1926, Parodi 7514 (type, Gray
Herb.) —This species belongs to the section Andina of of Knuth,
- Pflanzenr. iv. Fam. 129, 78 ( (1912), d to b st 5 related
. to G. ntl Knuth of Ber from whch fers in tly salle

lea’

argentina, sp. nov., suffrutescens 2-4 24 dm. aha i intra
4; Tamis gracilibus erectis cum pilis curvatis strigosis.
tis; foliis lineari-lanceolatis 2-4 cm. longis 24 mm. latis
Sanaa : : - oe me %

SOME UNDESCRIBED AMERICAN SPERMATOPHYTES 93

gracilem ca. 2 mm. longum exstipulatum contractis sparse p' ubescen-
tibus vetustate glabratis, superioribus reductis; racemis elongatis ca.
1 dm. longis; bracteis minutis caducis; floribus ad 4 mm. longis;
pedicellis pergracilibus ca. 1 mm. longis; sepalis lanceolato-ovatis
acutis ca. 0.9 mm. longis; alis oblique lateque obovatis ca. 4 mm.
longis et 2 mm. latis flavis; carino plicato aureo ad 4 mm. longo
galeato apice undulato; androecii appendice parva ovata brevi inter
phalanges staminum prolongata subtus supra medium dense longeque
pilosa ad 2.5 mm. longa; antheris 8 duabus solitariis, aliis (6) in
phalanges duas plus minusve aggregatis et basi connatis; stylo M.
Lorentzianae persimili; ovario dense pubescenti; samaris magnis late
ovatis apice profunde emarginatis basi cordatis ad 11 mm. longis et
8 mm. latis pains breviterque pubescentibus.—ARGENTINA: Las
Minas, Dept. Andalgal4, Prov. Catamarca, April 10, 1917, Jérgensen
1703 (G); dee Cuesta de Miranda near Chilecito, Prov. La
Rioja, 2200 m., Parodi 7805 (rype, Gray Herb.)—A member of the
subgenus Pterocarya. In Chodat’s conspectus, Bull. Herb. Boiss. i.
252 (1896), it works out to M. leptostachya Benth. to which it is ob-
viously not related. Its closest relatives appear to be M. dictyocarpa
Griseb. and M. a Chodat, both of which came from
Catamarca. From the masa it differs in its pubescent larger sama-
ras, much larger leaves, and more robust habit. Monnina Lorentziana,
to judge from the illustration, Chodat in E. & P. Nat. Pflanzenf. iu.
Abt. 4, 341, fig. 184 (1896), is quite different in its larger flowers, large
terminal-appendage of the androecium, glabrous fruit, distinct
anthers, etc. salle pena piste , stigma and upper

ne ten 3

nd 6 en: Nas anda Ga disco brevi vix carnoso m
undulato;

eee Pere sunny open space st se level

94 JOHNSTON

near Tela, Prov. Atlantida, Apr. 10, 1926, Elizabeth R. Mitchell 74
(TYPE, Gray Herb.).—Apparently belonging to the subgenus Pris-
vmera and most nearly related to H. Miersii pee (= Pristimera
apiculata Miers) of the Lesser Antilles and Guiana from which it
differs in its broader entire-margined white petals, shorter filaments
and less developed stigmas. According to the collector it grows
among other bushes on a sunny open place and becomes 3 m. tall.

e flowers were fragrant and the petals white. The leaves were
glossy like those of a Citrus. It is a pleasure to name this species for
its discoverer, Mrs. John Mitchell, vis during her residence in
northern Honduras, prepared excellent specimens of the plants of
that country. This material has been generously given to the Gray
Herbarium.

Fuchsia Killipii, sp. nov., fruticosa subseandens; ramulis sub-
teretibus brunnescentibus olacure -o parseque puberulentis; foliis
oppositis vel teretibus ellipticis 5-9 cm. longis 2-3.5 ecm. latis
integris obscure undulatis juventute obscure puberulentis mox

stipulis ede liatinie: Satin lea in axillis s supremis solitariis
vel geminatis vel corymbulatis conjunctim racemos vel thyrsos foliatos

cm. longis, fructiferis 2-4 em . longis; ovario crasse ellipsoideo incon-
spicue puberulento ca. 4 mm. longo; hypanthio 3-4 cm. longo basi
nodoso deinde contracto apicem versus gradatim ampliato 5 5-7 mm.

iis subaequilongis; staminibus inaequalibus quam petalis paulo
brevioribus; pate PA stylo ad medium vel supra villoso
stigmate subintegro coronato; bacca globosa ca. 6 mm. crassa.—
Cotomsra: Rio Suraté ety Suraté, Dept. Santander, 2000-2300
m., Jan. 5-6, 1927, Killip & Smith 16695 (rxpe, Gray Herb.) and
_-- 16604 (G); vicinity of La Baja, Dept. Santander, 2200-2600 m., :
Jan. 14-28, 1927, Kilip & Smith 16776 (G); vicinity of Charta,
__Dept. Santander, 2600 m., Feb. 1-11, 1927, Killip & Smith
ae Sige and Ltrsa ©. Giesay to F. venusta HBK., but — in

SOME UNDESCRIBED AMERICAN SPERMATOPHYTES 95

Killipit are much more numerous and more intensely scarlet than in

either of its relatives. It is known only from northeastern Colombia

in the Dept. of Santander and is described by its collectors as a
woody vine” and as a shrub 4~12 ft. tall.

Ochroma peruviana, sp. nov., arborea laxe ramosa; trunco 2-4.5
dm. crasso pallido; partibus juv vaiiies omnibus densissime tomen-
tosis saepe cum pilis stellatis dimorphis cinnamomeis obtectis; foliis
maturitate firmis ca. 34 em. longis et 30 cm. latis cordatis ——
vatis obsolete trilobatis b supra glabrati
subtus cum pilis minutis homomorphis stellatis — eae
cinnamomeo-tomentosis, margine obscure undulatis; petiolis maturi-
tate 25-30 cm. longis glabratis, juventute cum pilis dimorphis dense
obtectis; calycibus 8 em. longis coriaceis extus cum trichomatibus
dimorphis dense tomentosis, lobis exterioribus 3 em. lo:

sem ca. 2 cm. latis triangularibus ecostatis acutis, lobis interioribus
iia 3 cm. longis, tubo ca. 5 em. longo cylin

ulato; petalis ad 12 cm. longis clavatis 4—5 cm. latis extus apicem
versus tomenthais aliter glabris; capsulis ca. 14 em. longis
infra medium ca. 2.8 cm. crassis ca. 6 cm. longe pedicellatis apice
truncatis, valvis extus tomentosis marginem versus bicostatis, intus

aberrimis; seminibus ellipsoideis 2.6-3 mm. longis 1.5-2 mm.
cas stipitatis minute glandulosis, lana seminum fumea.—Perv:
Pampayacu, Rio Huallaga, Dept. Huanuco, Jan. 23, 1927, Ryozo
Kanehira 354 (type, Gray Herb.)—In range lying between 0.
boliviana Rusby, of Bolivia, and O. grandiflora Rowlee, of southern
Ecuador. It appears to be most closely related to O. grandiftora,
but differs in its plane rather than carinate outer corolla-lobes, and
in its more copious and different pubescence. The new species seems
to be unique in the genus because of the very dense brown tomentum
which covers the young branches and leaves as well as the outer
surface of the mature calyx. At first glance this tomentum appears
ramentaceous, though close examination | shows it to consist of a

= It is locally known as “ es ‘and its wood is used for rafts and
shelters.

96 JOHNSTON

Ardisia Mitchellae, sp. nov., fruticosa 1.5-3 m. alta; foliis ellip-
ticis 10-13 em. longis 4-5.5 em. latin glabris apice late acuminatis basi
acutis eign late et t supra medium crenulatis basem ° versus reste
supra | entib plie

pinnato-nervatis; petiolis 5-10 mm. longis; ihesssndatin terminal
laxe subcorymbosa ad 25-flora foliis multo breviori ad 6 em tro;
pedicellis (1.5-2 em. longis) et pedunculis ferrugineis aalalas minutis
eylindricis brevi-stipitatis obtectis; floribus pulcherrimi

1 em. longis; sepalis basi breviter coalitis triangulari-ov. a ca.
2.5 mm. longis glandulis stipitatis obtectis; petalis dextrorsum tegenti-
bus oe = late _acutis albis. vel _apicem versus

Tosa erat OW
WALT IS

ae antheris angustis linearibus apice cuspidu-
latis 6-7.5 mm. longis quam petalis evidenter brevioribus; ovario
ovoideo glabro; stylo gracili antheris vix superanti; bacca ignota.—
Honpuras: in good soil in unde yth near river, vicinity of Tela,
Prov. Atlantida, alt. 15 m., April 4, 1926, Elizabeth R. Mitchell 66
(rypr, Gray Herb.); rich soil in dense jungle, vicinity of Tela, 150
m., May 6, 1926, Elizabeth R. Mitchell 98 (G).—A species belonging
to the section Icocorea and apparently most clésely related to A.
Donnell. Smithii Mez, from which it differs in its go flowers,
cuspidulate anthers and glabrous lower leaf-surfaces
Verbena j Lag., var. gTisea, var. nov., a varietate

densos formantibus.—PERU: ravines, southern slope of Chachani,
north of Arequipa, 3350 m., March 1920, Hinkley 76 (vypE, Gray

Herb.); open rocky slopes on Chachani, ea, 2600 m., April 1925,
Posiecll 13259 (G); open mixed formation, Cuajones Mine, Torata,
ea. 3250 m., Feb. 1925, Weberbauer 7461 (FM).—Obviously related
to the Argentine V. juniperina Lag., but readily separable by the
ashy rather than light-green herbage and its reduced branchlets. In
the typical form of V. juniperina the branchlets develop, becoming
4-9 em. long, but in the var. grisea they are represented by leaf-
fascicl i very rarely b than 10mm.long. The variety

~~

y

oy is a low, 3-5 dm. tall, very dense shrub and is known only from
Peru whence it has been reported by Weberbauer, Veg.
Erde xii. 130 (1911) as V. juniperina.

sp. nov., pam pace cote Uns tos

pslonge Sy ; i

SOME UNDESCRIBED AMERICAN SPERMATOPHYTES 97

alternis linearibus obtusis glabris 2-6 cm. longis 1-1.5 mm. latis,
superioribus plus minusve reductis; floribus ad apices ramulorum in
cymam simplicem irregularem aggregatis resupinatis; calycibus glabris
irregulariter 5-lobatis cum lobis superioribus longioribus, ad anthesin
4-7 mm. longis, fructiferis 8-10 mm. longis quam capsula subduplo
—— pedicellis 3-20 mm. longis rectis ascendentibus; corolla
t plus minusve aurantiaca tubulata 17-20 mm. longa;
tubo 11- 15 mm. longo, parte inferiori ad 2 mm. longa gracillima ca.
0.5 mm. crassa aurantiaca in calyce occulta, parte superiori et
media a 1.8-2 mm. crassa aurantiaca vel purpurea, eat
inco! e glandulari-villosa, intus basem versus retro:
villosa ; ache paullo inflatis et obliquis brevibus; limba eiiak
cum venis purpureo-marginatis reticulatis; labia inferiori breviori
bilobata ad 2.5 mm. longa oblongo-lanceolata recurvata; labia super-
iori longiori trilobata, lobis anguste ovato-triangularibus 2 mm. longis
in alabastro exterioribus; staminibus 5 inclusis, duobus superioribus
perfectis cum filamentis linearibus infra medium corollae affixis
decurrentibus basem versus sparse villosis hi antheris orbicularibus
1 mim. longis: encakcomae: cum loculis app et apic 3
;

oo. a

ality supra mednim

corollae lineartbus
imperfectis apicem versus paullo dilatatis glabris, stamine infimo
perfecto sed valde reducto in faucibus affixo minus quam 0.5 mm.
longo; stigmate dilatato ca. 2.3 mm. lato 1.2 mm. longo transverse
pentagono-oblongo; stylo lineari glabro; disco miniato; capsula ellip-
soidea ad 4 mm. longa; seminibus numerosis angulatis——Pervu:
white dunes, Tiabaya, Dept. Arequipa, ca. 2150 m., April 8, 1925,
Pennell 13063 (rype, Gray Herb.); open sandy slope, Tiabaya, ca
2150 m., Pennell 13081 (G); open sandy soil, Tingo, Dept. Arequipa,
ea. 2200 m., April 8, 1925, Pennell 13119 (G). This striking plant
appears to be without close relatives. It seems to fit best into the
section Leptoglossis of Wettstein, E. & P. Nat. Pflanzenf. iv. Abt. 3b,
36 (1895), but is atypical there since it produces aborted
stamen. The species is characterized by its entire leaves, of which
the cauline are glabrous and decidedly linear, and by its elongated
tubular corolla which is abruptly contracted (inside the calyx) above ©
the base. The detailed notes made by Dr. Pennell in the field on
numbers 13081 and 13119 are as Sie Co near apex

98 JOHNSTON

Parodii, sp. nov., herbacea annua glaberrima erecta
ca. 15 em. alta supra sparse stricteque ramosa; foliis radicalibus
ignotis, caulinis alternis pmnatis vel bipinnatis 3-6 cm. longis oblongo~
ovatis, lpeag aiianar, distantibus Inearibus RG CNAE Ae pels ad
basem dilatatis mm. diametro
eradiatis terminalibus a foliis subtendentibus aes superatis,
pedunculo 2-10 mm. longo ebracteato sulcato apice sub capitulo
comapaene | incrassato; involueris. biseriatis rigidis conniventibus,
ca. 5 mm. longis quam interioribus

paullo brevioribus; receptaculo conico nudo; achaeniis subteretibus
2.5 mm. longis multicostatis dense adpresseque sericeo-villosis; pappi
paleis 5-6 hyalinis ellipticis vel oblongis erosis acuminatis paullo in-
aequalibus achaeniis subaequilongis quam corollis paullo brevioribus;
corollis 2.5 mm. longis flavis, tubo 1 mm. longo brunnescenti sub-
cylindrico extus ad apicem puberulento, faucibus cylindro-cam-
panulatis, dentibus limbi erectis triangularibus extus puberulentis.—
ARGENTINA: soil near. Bafado de Flores, near Buenos Aires,
Nov. 13, 1927, L. R. Parodi 8170a (TxPx, Gray Herb.).—Related to
H. anthemoides (Juss.) Cass. but differing in habit, size of heads,
of peduncles and length of pappus. The plant is erect, simple

below but with a few strict branches above. The heads are large and

surpass the adjoining foliage. The acuminate tips of the pappus-
seales surpass the corolla and appear as bristles protruding from
between the florets in an undissected head. Although the present

: W. a BOTANICAL ACTIVITIES OF THOMAS BRIDGES
By Ivan M. Jounston

Tuomas Brinces (1807-1865) was one of the botanical collectors of
ae ee Oe et es he eeeoreny ae
ee ee eet fom ie eae Bein. His collections are

THE BOTANICAL ACTIVITIES OF THOMAS BRIDGES 99

important through the study given them, and the species and records
based upon them, by such botanists as Hooker, Lindley, Bentham and
Miers in Great Britain and by DeCandolle and Turczaninow on the
Continent. Many of Bridges’s plants being critical ones it is an
unfortunate fact that the geographical data accompanying them are
very meagre and frequently misleading if not actually incorrect.
Suspecting these facts I have on several occasions desired information
concerning the routes traveled “as Bridges with the hope of gaining
therefrom some suggestion as to the precise source and probable
identity of some critical collection. Laségue, Mus. Bot. Delessert
259-260 (1845), has provided the most extended and reliable account
of the collecting activities of Bridges. His account, however, is
necessarily brief and is not especially detailed.

The present sketch of Bridges’s botanical activities has been worked
out from a study of the letters from Bridges, H. Cuming and A. Cald-
cleugh! in the Hooker Correspondence at Kew, from Bridges’s plant-
lists at Kew and the British Museum of Natural History, and from
such pertinent published works as I could discover. A determined

Ro
Society, Royal Geographic Society, Horticultural Society and the
British Museum at Bloomsbury. Cuming was a friend and later
became London agent for Bridges. Were it possible to locate
Cuming’s correspondence and manuscripts much information con-
cerning Bridges might be found. Except, however, for some letters
at a. nothing of this sort was discovered in any of the institutions

“Lately we have received from England a collector of the is name of
Bridges, but as he has come out on his own speculation, he has been
forced to take a trade for his existence, and is now a brewer of small
beer in this place. Mr. Lambert and the Linnean Society were
instrumental in his coming out.” Thus wrote Caldcleugh to Hooker
from Valparaiso, Chile, on May 10, 1829 concerning Bridges who, in
his twenty-second year, had landed at Valparaiso nine months earlier _

on Aug. 8, 1828.

Bridges's activities el oe
wing him as earlier ha sach connections in. Brazil. ee ene
| foamy. He collected about ¥: and Coquimbo, but his great service to

— ae oe fuming in gore gl ga that,
as yet na but later famous =a oo Sir ‘Wiliam Hooker

1c0 JOHNSTON

- During the first year in Chile Bridges does not appear to have been
able to collect extensively, although he did make the acquaintance ~
and become an admirer of Carlo Bertero, the Italian botanist who was
then actively solicctioe 3 in Central Chile. In Dec. 1829 Bridges sent
Hooker his “collections of 1828,” that is presumably those obtained
in the sprmg and summer of 1828-29. These specimens were un-
numbered and apparently all came from the region about Valpariso.

In Oct. 1830 Bridges wrote that he had recently collected in the
Department of Quillota and that he was preparing for an expedition
to the Cordilleras, having “received some assistance from Mr
Barclay and Mr. Bevan.” I have found no letters giving the details
of this trip to the Cordilleras. In May 1832 Bridges shipped from
Hacienda de San Isidro, Quillota, 497 specimens to Hooker, 280 to
Arnott and some to Bentham and Lindley. These are obviously the
plants, numbered 1497, listed in a catalogue preserved in the library
at Kew. These plants were collected in the Cordilleras, in the Acon-
cagua River Valley and in the region centering around Valparaiso.
They obviously contain the results of his first expedition to the
Cordilleras, which probably took place early in 1831.

A study of the catalogue which Bridges prepared for his numbers
1-497 shows that the localities mentioned can be roughly grouped as
follows: 1. localities near Valparaiso, viz., Via del Mar, Placillas,
Playa Ancha, Queb. Lacumas and La Hacienda de la Merced;
2. localities along the Casa Blanca road to Santiago, viz., Casa Blanca,
Cuesta de Zapata and Bustamante; 3. localities along the coast north
of V: , Viz, Refiaca, “oncon and Quintero; 4. localities center-
ing around Licdlistn ta, viz., Limache, Hacienda de la Palma, Sierra de
San Isidro, Cajon de San Pecins and Cuesta de Pachacuma; 5. localities
along the old Mendoza road up the Aconcagua Valley, viz., Liaillai,
Tinajas, Hacienda de San aig San Felipe feud north of that town,
Los Lores and Sierra Bella Vista), Salto del Soldado, Guardia Vieja,
- Ojos de Aguas and La Laguna [del Inca]; and finally, 6. a
along the road between Los Andes and Santiago, viz., Cuesta d
‘Chacabuco and Colina. Two stations given as “plains near cas

and “ Cuesta ce Chile Cauquen Wn. Quillota” I have ea unable to
locate.

Le Rasy i Sep 1832 Bridges wrote that he had just arrived in San-
_ tiago from eS ee year laying out a
arm for : . Waddington. He wrote that he had had little

THE BOTANICAL ACTIVITIES OF THOMAS BRIDGES 101

In a letter from Valparaiso, dated Oct. 25, 1832, Bridges stated that
he now planned to devote himself to botany and had booked a
to Valdivia. Writing from Valdivia on Feb. 26, 1833? he told of
having just returned from a “very long journey to the interior,”
having traveled with a party from the “ Commissary of the Indians”
whose object was to stop one of the passes in the Cordilleras and
prevent the incursions into Chile of the Pehuelche Indians. He also
mentioned visiting Lake Ranco and concluded his letter by stating
that he was going directly to Chiloé in a “few days.” On Aug. 27,
1833 Bridges wrote that he had returned to Valparaiso after an absence
of nine months and announced the sending to Hooker of 283, to Arnott
of 268 and to Greville of 248 plants of southern Chile. He also sent
to Hooker a catalogue of his numbers 558-857 which covered all of
his collections from southern Chile. This catalogue is now preserved
in the library at Kew. I have seen no catalogue of his numbers
498-557. These numbers probably — to the “se which
Bridges collected about Valparaiso and Quillota in 183:

It does not seem possible to determine Bridges’s aa in Valdivia
from a study of the localities given in his catalogue. About Corral,
whence many of his plants were obtained, he mentions such localities
as Castello del Corral, Castello de Amargos, Isla Mansera and Castello
Niebla. In the region about the town of Valdivia he mentions Arique,
Pufude, “Las Animas,” “Chumpulla,”’ “Los Canellos” and “El
Caneagual,”’ the four latter of which I have not been able to locate.
Also mentioned in his list and the sources given for many of his col-
lections are Lago Ranco, “Los Andes between Osorno and Rio
Maullin” and “Los Uanos between Valdivia and Osorno,” the last
two I have been unable to locate. The only localities which indicate
that he actually visited Chiloé, where he apparently collected very
little, are “ Puguenun River” on the northern extremity of the island
and “La Punta del Carelmapu” on the mainland just north of the
island across the channel from Puguenun.

Bridges, —— did not continue active botanical work as ie
had planned. We learn from Cuming, in a letter dated a 1834,
that Bridges had agreed to superintend for two years an “estate
near Talca.” Bridges is next heard of through Caldeleugh who
wrote from Santiago in — 1836 that “Mr. Bridges came up lately
Se . he says [collecting] is quite out of ae

at present, as his employer edicatin

any porn of his ine to other pst, Tt seems that he is in :

: of this letter was summarized and _— ok
a Bot. {. ies. * dai 250 ey fants

102 JOHNSTON

receipt of a good salary, is saving money and is fearful of risking his
situation.”

After a lapse of six or seven years Bridges resumed correspondence
with Hooker, writing from Valparaiso on June Ist, 1841, that since
he had last written he had been “occupied with agricultural affairs,
with little benefit to himself and too busy for Natural History.”
That = however, he had resumed botanical work. He had

excursion over the Andes by Pass of Planchon, lat. 34-35°,
to the esate valleys on the eastern slope” and in a “few days”
was sailing for Copiap6é to begin exploration there. Bridges also
announced that he was sending Hooker, some plants from the “ Andes
of Colchagua” and some from the base of the Andes in the Province
of Colchagua. He stated that no catalogue of these collections had
been made. These specimens are no doubt those from Bridges which
Cuming, i in his letters of Oct. 27 and Nov. 26, 1841, indicated as con-
sisting of 150 from “the Andes” and 220 from “ various localities.”
specimens, it is certain, carry numbers falling between 858
and 1278. It is to be noted, however, that, besides the 370 plants
mentioned by Cuming, 51 additional ones would be co

base early in 1841, some of this gamut came from elsewhere, probably
from near Val;

I have very little precise information regarding the —
details of Bridges’s work in the Province of — It is
important fact, however, that the “Province of Colchagua” is aaa
smaller now than formerly, for in 1865 it was divided and a new

Province of Caz, To reach Paso Hl Plnchon, lt. 25 Hence
there can be little doubt that plants of middle and high altitudes,
ee ee ey ee ee Coy oF

Bridges mentioned visiting “El Valle de las

i Cuevas, about 10 leagues to the east of Volcano of Petorca,” ’ a valley
So the Mendoza side.

oon is a reference by Bridges, Proc. Zool. Soe.

a London xi. 7 (846, n which he mentions the Teno and seme a

THE BOTANICAL ACTIVITIES OF THOMAS BRIDGES 103

to infer a thorough acquaintance with the parts of that stream at the
foot of the mountains. I suspect that the hacienda at which he was
employed was situated on or near the Rio Teno and that the plants
from the “base of the Andes in the Prov. Colchagua” were obtained
in the region just north of the city of Curico

Bridges was next heard from on Nov. 20, 1841, announcing that
he had arrived in Coquimbo a few days before, having been occupied
since the end of June in collecting between Copiapé6 and Coquimbo.
He next planned to ascend the Rio Elqui “to snowline” and subse-
quently to go to Valparaiso by way of Illapel, Petorca and Quillota
and to arrive there late in January. The plants collected on this
expedition to northern Chile bear the numbers 1279-1424. The
original catalogue which gives rather full locality- and habitat-
data, is preserved in the library of the British Museum of Natural
History. The localities mentioned by Bridges, arranged in the
probable order in which he visited them are as follows,—Port of
Copiapé, Copiapé6, Andes of the Valley of Copiapé, near Chafiarcillo
El Totoral, Los Pozos, Vallenar, Freirina, Huasco, Pefia Dianen:
Coquimbo, Mina Arqueros, Valle Elqui, Vicufia, Paice, *Valle
of Borasa, Prov. Coquimbo” (unlocated), Andacolla and Illapel.
Although the first ten localities listed lie within the Province of
Atacama and about 50% of the collections came from that province,
all of Bridges’s collections of late 1841 are found in herbaria and are
almost universally cited as from the Province of Coquimbo or merely
as from Coquimbo. In Hooker’s herbarium by some blunder the
collection is all labeled as from Concepcion! There also appears

confusioi of
rege Tiust. S. Am. Pl. ii. 25 (1857). According to Bridges, in
: ‘The best one of ‘ane Sei to Hooker.

tee a a

We next hear from ce 5 ee oe Cochabamba.’ dated
+ This was published in large part by Hooker, Jour. Bot. iv. 571-577 (1848).

104 JOHNSTON

April 3, 1845. We learn that he had landed, a second time, at Cobija
on Sept. 13, 1844. He remained there a “ few days” and was delighted
to find a variety of plants on the fog-bathed slopes above the town.
Obtaining mules and men he went to Calama, thence northeastward
through Tapaquilcha to Potosi, where he spent “only a few days.”

Cochabamba on Dec. 24, 1844. After about three months, early in
April 1844, he crossed the mountains northeast of Cochabamba and
entered the Amazon Basin following down the Mamoré River north-
ward towards the Brazilian frontier. In July 1845, when he en-
countered Victoria,‘ he was at Santa Ana on the Yacuma River,

tributary of the Mamoré. Loreto and Trinidad are the only at Si
localities in the Mamoré Basin mentioned as having been visited by
him. Bridges also visited Santa Cruz de la Sierra, going there prob-
ably up the Rio Piray as he had planned. In March, 1846, Cuming
wrote Hooker that a letter from Bridges, dated Sept. 11, 1845, had
been received from Santa Cruz de la Sierra. Bridges reported that
his excursion into the Amazon Basin had net been productive of a
variety of plants, in the “thousand miles” of travel he had not col-
lected 100 species. I have no definite information concerning the

to return to Cochabamba. Since he reports a bird, Proc.
Zool. Soc. London xv. 29 (1847), from eNeetl of La Paz” I suspect
that he left the country by way of La Paz. He must have returned
promptly to England, for from London on June 21, 1846, apparently
some time after his return, he wrote Hooker that the seeds of Victoria
were to be sold at 2 shillings each and the herbarium specimens of the
plant at 30 shillings. The first and best set of Bridges’s Bolivian col-
_ lections were sent to Hooker; according to its collector it consisted
of about “550 species.” The specimens were unnumbered and were
apparently unprovided with definite enor? data. According to

Bolivia. ‘He wrote Hooker on Dee. 8, 1846 from Bristol that i
Le rag a ™ ~> = “been expectorating blood from the lungs for

"to Dall, Pre. Calta 236 (180 he

THE BOTANICAL ACTIVITIES OF THOMAS BRIDGES 105

married Mary Benson, niece’ of Hugh Cuming. Shortly after his
marriage, _ the sake of his health, he returned to Chile going there
via Panam

The last ee we have concerning Bridges 1 in South America is in
Caldcleugh’s letter from Valparaiso which is dated Sept. 30, 1851.
We hear that “ Mr. Thomas Bridges is now in this place having formed
a kind of nursery ground for the sale of every description of fruit and
other trees and plants. He now collects little and is very sore upon
the subject of Victoria Regia which he asserts was introduced (at
least the subject which flowered) by him and that he has received no
medal and scarcely ‘mention honorable’ for having done so.”

According to Dall, |. c., Bridges “ visited and explored the island of
Juan Fernandez” in 1851. If he did so he made no botanical col-
lections there or at least these did not reach European herbaria, for
neither Hemsley nor Skottsberg, who h tudied and published upo:
the flora of the island, mentions his collections nor notes him as acing
contributed to our knowledge of the flora of the island.

In 1855, Dall states that Bridges proceeded to Panama remaining
there some six months; and from thence to England, subsequently
to France, and finally to California where he arrived in Nov. 1856.
About 1857 he went to British Columbia, and remained there nearly
two years collecting and exploring. He then returned to California

and made his home in San Francisco until his illfated journey in 1865.
Very little is known concerning his collecting-itinerary in California.
We learn from a letter dated May 5, 1858, that he had collected i in

lections, except those distributed by the Smithsonian Institution after

his death, lack numbers, and invariably, it seems, have no ——

eation of collection-localit

In April 1865 Bridges went to Nicaragua for biological exploration,

ne there until September. While returning to to San Francisco
with malaria, four days later on Sept. 9,

1865. He was 59 years of age at “his death and was survived by: ee

ca = hve children.

ee ccoantae i re

o

—* L ce ee Sir Williaan Hooke . ie

eS
PAE OU RU.

an a ed es de poder econ La

at least, appear oF

: _ spondence one infers that the second set was retained for him i
A What became of this material I do not know. ‘At Kew with the

106 JOHNSTON

Bridges plants from the Hooker Herbarium are those received from
Bentham. Many of the collections by Bridges in the Bentham
Herbarium have “Lord Colchester, 1832” written on the labels.
The significance of this annotation is not entirely clear. Mr. S. A.
Skan, librarian at Kew, has, however, called my attention to the fact
that, according to the Dictionary of National Biography, the mother
of the parliamentarian, Charles Abbot, First Lord Colchester, by a
second marriage became the wife of Jeremy Bentham, by his first
marriage grandfather of George Bentham the botanist. Th

Lord Colchester was an officer in the British Navy. Bentham, hence,
probably received the specimens in question from his relative the
First or Second Lord Colchester and consequently the annotation
“Lord Colchester, aay ” so promineat on Bridges’s labels has only
historical si

The Gray ieee contains an incomplete set of Bridges’s col-
lections from Chile and a few of his plants from Bolivia. During the
course of some work on the Chilean flora I have become convinced
that the numerous collections in the Gray Herbarium which are
unaccompanied by labels and associated only with the data, “Am.
Aust. F.” or “Am. Aust. Fielding” in Asa Gray’s handwriting, are
in large part also collections of Bridges. These are probably from
the herbarium of Henry B. Fielding and in some manner became
divorced from their original data.

The salient facts cor ing Bridges’s collections in South America
may be summarized as follows:—
Year Locality Number Collection
« colleeted numbers
oe a y — Valparaiso unknown cen
1830 (&? 1831 cg region, Aconcagua
alley and Cordilleras. 497! 1-497
1832 ? Prov. V: ‘aiso Set 498-557
1832-1833 Prov. Valdivia 299! 558-857
_ Jan.—June, 1841 uri ed 858-1278
1841 ? '